WO2010108287A1 - A wearable intelligent healthcare system and method - Google Patents

Abstract

A system and method for the wearable intelligent healthcare system for monitoring a subject and providing feedback, comprising of physiological sensors, activity sensors, a processor, a real-time detection and analyzing module for continuous health and activity monitoring, adjustable user setting mode with the adaptive optimization, data-collecting capability to record important health information, audio outputs to the user through audio path and audio interface, preset and user confirmable alarm conditions via wireless communications network to the appropriate individual for prompt and necessary assistance. The system uses noninvasive monitoring technology for continuous, painless and bloodless health state monitoring. The system works through the short range wireless link with carry-on mobile unit for displaying health information, making urgent contact to support center, doctor or individual, and for information transmission with a healthcare center.

Description

A WEARABLE INTELLIGENT HEALTHCARE SYSTEM AND METHOD

[001] This application relates generally to a wearable intelligent healthcare system for monitoring a subject and providing feedback. More particularly, the first application relates to a system and method wherein the healthcare system monitors a subject's activity and health condition to provide feedback regarding health risks, exercise programs and healthier living. The second application relates to a system and method for fall detection, wherein the healthcare system monitors a subject's activity and health condition to enable the early detection of falls and other adverse health conditions. The detection of a fall also allows urgent contact or information transmission through the mobile device over wireless communication network to alert a monitoring center to the fall. The third application relates to an ear hook assembly and method for securing a device or components of a system to an ear. More particularly, this application relates to an ear hook assembly for a healthcare system, wherein the ear hook assembly facilitates support, data acquisition, processing, communication and the like.

[002] As is well-known, people are generally becoming more and more aware of health issues. Further, an aging baby boomer population has added to interest in healthy living and better medical care. Although there are many instruments or devices available for monitoring an individual's activities and evaluating their health states they are typically not very portable/mobile and a person with health issue needs to periodically visit a medical facility to obtain the proper diagnosis and medical treatment. The health information obtained during the visit only represents a small portion of the subject's physiological information at the time of the examination, which usually does not reflect the actual health problem occurring in the daily life. In order to obtain more complete medical information, doctors would need to observe a subject's health condition over a longer duration. Because of time and costs associated with these tests and observations, it is usually impractical to conduct the required long-term observation and full evaluation for most people who may need them.

[003] Another major issue for many individuals is getting prompt medical instruction and care as soon as a health problem occurs. For example, a heart attack victim will have a significantly greater chance of full recovery if medical care is received as soon as a heart attack is detected. As another example, an early detection of the sleep apnea can give an individual good opportunity to take necessary actions to prevent the serious sleep-disordered breathing problem from developing.

[004] Unfortunately, an individual usually does not recognize the early signs which indicate an occurring risk. Quite often, by the time the individual does realize an occurring risk, they might be incapable of seeking for medical assistance. This often occurs with sudden falls, which are a common occurrence and can be both an indication of and a cause of serious health problems that require prompt medical attention. A person that has suffered a fall may suffer injury or unconsciousness, that renders them incapable of recovering from the fall or seeking assistance.

[005] Conventional health monitoring systems typically have issues with high power consumption, larger size, which can limit a user's freedoms and mobility, the use of wireless communications among sensors, which can result in interference and the like. For example, conventional wearable healthcare systems may provide a belt or wrist mounted central unit that uses a wireless sensor network such as a Personal Area Network (PAN) or Body Area Network (BAN) for the sensor data transmission between sensors and central unit. Although wireless technology for local area network has made it possible to communicate the constant data streams of sensors to the central processing unit, it has serious limitations such as system complexity, device size, power consumption, reliability of the wireless body area network, interference from environment and user health affection possibly induced by the constant wireless signals around body all the times.

[006] Various physiological and activity monitors are known in the art. In some cases, there have been rudimentary efforts at correlating physiological signals and activity, for example, correlating between physiological data and musculoskeletal loading, however this has been limited to very specific tests, including gait-related loading. Other health/activity monitoring systems rely on the user to enter data directly, similar to the way that a patient may tell a doctor a medical history or activity history. As is well-known in the industry, this information is not typically very accurate and it is easy for a person to neglect to enter/discuss a specific occurrence or the like because of forgetting or embarrassment or the like. [007] In general, it is well known that activity may affect health. Some known systems address how to detect activity level or amount or how to detect health condition. However, all the activity detection systems or methods have not included personalized health information as a part of the detection knowledge. Known systems do not appear to have included intelligent processing for individual activity detection.

[008] Some approaches to fall detection and notification rely on monitoring the orientation and acceleration of the person and analyzing the monitored data to determine whether a fall has occurred. Even in these situations, however, a person that experiences a fall may not require assistance or medical attention and sending a fall notification would result in an unnecessary use of resources. These approaches generally do not provide information about whether a detected fall represents a health risk to the person and whether they require assistance or medical attention.

[009] Accordingly, there is a need for a wearable intelligent healthcare system and service that can provide a small size, low power consumption, low cost, high intelligence with minimized use of wireless communication. In particular, there is a need for a system and method for health/activity monitoring and feedback that is reliable, accurate, and capable of providing feedback that can help a user improve their health.

[0010] In particular, there is a need for a system and method for fall detection and notification that is reliable, accurate, and capable of determining whether a detected fall is critical to the user and whether or not the person needs assistance or medical attention.

[0011] In view of the above, it is an object of embodiments of the system and method herein to address at least some of the problems associated with monitoring health and activity reliably and accurately, determining whether the user requires changes to health or activity variable, and providing feedback based on that information.

[0012] The present application relates to real-time monitoring of a subject's health condition with intelligent detection and analysis capability. The monitoring may be continuous in some cases. In a particular embodiment, the unique advantages of having the healthcare monitor/activity detector worn around the ear are used to advantage. In general, the embodiments herein are intended to have a smaller size, lower power consumption, less complicated design, more reliable performance, easier to wear and lower cost compared with the conventional healthcare and activity monitoring systems.

[0013] The present application also relates to systems and methods for fall detection. An accurate detection of a fall is intended to allow for a warning or alarm to be sent with better accuracy. In a particular embodiment, the unique advantages of having the healthcare monitor fall detector worn around the ear are used to advantage. In general, the embodiments herein are intended to have a smaller size, lower power consumption, less complicated design, more reliable performance, easier to wear and lower cost compared with the conventional healthcare and fall detection systems.

[0014] Once the healthcare system detects a concerned health condition in the subject, which may correlate with the subject's activity detection, it can alert the subject or notify the appropriate people so that the subject can take necessary action accordingly. In addition to issue some types of alarms such as a loud beep sound to alert the subject, the device can issue audio messages for specific health condition as smart warning, advice or reminder. If a serious or dangerous health condition is identified in the subject, the device may issue smart audio warning to the subject and automatically use the integrated short range wireless link between the device and the mobile unit to request the mobile unit to make a contact with medical center, doctor or family member through the available wireless network. The device may be programmed such that a call to 911 is immediately made and the subject's name and medical history are provided therewith. At the same time, the device may also provide the 911 operator with the subject's location, by sending them a global positioning coordinate if the positioning capability has been included in the mobile device. It is possible to use mobile device via the short range wireless link to display the health state and dynamic health signal so that the subject or other person around can observe them.

[0015] Accordingly, the intelligent healthcare system may be worn on a subject over the ear and carried anywhere while using noninvasive monitoring technology. The intelligent healthcare system may be setup to store current medical information and detect any predefined alarm conditions, such as heart attack. Upon an occurrence of such alarm conditions, the device may provide smart audio outputs such as warning, advice or reminder to the subject for a concerned situation or contact the healthcare center, doctor or family member with health information for the necessary healthcare or medical assistance for serious situation.

[0016] It is another object to combine the advantages of the global positioning system for locating the subject at the time of the health crisis. The communications capabilities of a mobile device would provide the most prompt emergency assistance.

[0017] It is still another object to provide a service for health information collection and long term storage of a remote subject's medical data via wireless communications technology.

[0018] The device may also provide an audio interface with media player, phone receiver through the short range wireless link, or even hearing enhancement capability over the basic healthcare functionalities.

[0019] The device may also provide for other factor detections such as environment detection, weather detection, acoustic signal detection or even subject's emotion detection.

[0020] For a better understanding of the embodiments herein and to show more clearly how they may be carried into effect, reference will now be made, by way of example only, to the accompanying drawings in which:

[0021] FIG. 1 is a system overview illustrating the healthcare system.

[0022] FIG. 2 is a system application overview illustrating the healthcare system.

[0023] FIG. 3 is a system diagram for the healthcare system.

[0024] FIG. 4 is a block diagram illustrating an example of a physiological sensor unit with intelligent signal processing capability.

[0025] FIG. 5 is a block diagram illustrating an example of an activity sensor unit with intelligent signal processing capability.

[0026] FIG. 6 is a block diagram illustrating the health monitoring principle of an embodiment of the intelligent healthcare system. [0027] FlG. 1 is a system overview illustrating an embodiment of the healthcare system. The system consists of physiological sensors (S1 ) and body temperature sensor seamlessly contacting the skin behind the ear, activity sensors (S2), a central processing module (CPM), a speaker for smart audio outputs, an audio delivery path with the audio interface, a contact sensor touching the skin behind the ear, a battery as system power supply, a short range wireless communication unit (RF) and a shell to contain the system. The FIG. 1 also contains the adjustable user setting for system optimization, user cancel for self-confirmation to eliminate possible false alarm and user request for user to check current health state or issue a necessary urgent request. The invented system is a mini-size device designed to be worn on the ear by a subject, providing the subject with great mobility and comfort.

[0028] FIG. 2 is a system application overview illustrating the healthcare system connected to the healthcare center, doctor or family member through cellular network or any wireless network via a mobile unit available in the art. The urgent contact, health information or location can be transmitted as either user request or automatically generated by the healthcare system. The healthcare system can also receive instructions or other information from the healthcare center, doctor or family member. In the meantime, the mobile unit can be used to display the health information or save the medical data from the healthcare system when necessary or required by the user. The healthcare system is able to communicate with the mobile unit within the short distance of arm coverage such as 1.5 m via the short range wireless link. On the other side, the mobile unit can connect to the healthcare center, doctor or family member without distance limitation as long as commercial wireless communication coverage is available.

[0029] FIG. 3 is a system diagram for the healthcare system. In this embodiment, the monitoring system 1 continuously monitors a subject's physiological signals and/or activity signals as it receives them continuously from the physiological sensors and physical activity sensors. The system consists of a central processing module CPM 11 , physiological sign sensors (S1) 21 , activity sensors (S2) 22, a contact sensor 23, a speaker 41 for smart audio outputs, an audio path 42 with audio interface 43 to the ear canal without affecting normal acoustic signal access to the eardrum, a RF communication unit 44, an I/O interface 45, a battery 51 to power the system and a shell 52 to contain the system. The FIG. 3 also contains the user controls including user setting unit 31 , user cancel 32 and user request 33. In the invented system, one or multiple vital life sign sensors 21 for detecting the subject's physiological condition such as SpO2, glucose or other signals. In one embodiment, activity sensors 22 are for detecting the subject's physical activity. The unit CPM 11 is typically comprised of a central processing unit (CPU) and memory with intelligent signal processing algorithm running in real-time.

[0030] FIG. 4 is a physiological monitoring unit, associated with physiological sensors 21 , which can continuously monitor physiological condition such as oxyhemoglobin saturation (SpO2), body temperature or even glucose. It is preferable to use noninvasive monitoring technology for continuous, painless and bloodless measurements for physiological signal monitoring. In the example of using physiological sensors for oxygen saturation detection, the red light (with 660 nm wavelengths) and infrared light (with 910 nm wavelengths) are emitted through the earlobe by light sources of sensor unit (S1) and to use optoelectronic sensors to detect the amount of light reflected back from the reflection plate, in which lights have gone through the earlobe twice by reflection. In addition to obtaining real-time blood oxygen level and plethysmography signal, the intelligent detection algorithm extracts heart rate, blood flow information, sleep apnea when the subject is in sleep, and the like. Another example of such physiological sensor is to use near-infrared light (with wavelengths between 1000 nm and 2500 nm) to detect the glucose in the similar principle. The real-time physiological detection algorithm continuously monitors the subject's physiological signals, extracts its pattern, predicts the trend of the physiological condition and analyze the physiological condition according to the medical expert knowledge and the subject's own health history. In the physiological monitoring unit, the body temperature may also be monitored since it offers basic physiological information of a subject, which can be used to help to analyze the subject's health condition.

[0031] FIG. 5 is an activity monitoring unit, associated with activity sensors 22, which can continuously monitor the subject's physical activity in XYZ dimensions for motion detection including fall detection. There are many types of activity sensors available and the example of the smallest activity sensors are piezo-resistive 3-axis acceleration sensors. The real-time activity detection algorithm continuously detects the subject's activity information such as rest, walk or run, and amount of the activity over time. The extracted activity information such as activity state, activity strength and duration can offer important correlation information for health condition evaluation in addition to be used for analyzing the subject's life style, exercise pattern and health plan. A fall detection capability is included in the activity monitoring, which is especially valuable for the elder people since a fall can be very dangerous and the person may need urgent attention.

[0032] FIG. 6 is a block diagram illustrating the health monitoring principle of the healthcare system, in which either the physiological information detected from the physiological monitoring unit or the activity information detected from the activity monitoring unit are analyzed, or both of them are analyzed accordingly with the correlation of these signals. The health diagnosis is conducted with the use of expert knowledge and subject's health reference. The health state is determined and updated as a function of time. If a health state of concern is detected, the system will emit smart audio outputs to alert or remind the subject of the health condition. If a serious or dangerous health state is detected, the intelligent healthcare system will both emit the smart audio outputs to the subject and request, via the short range wireless link, the mobile device to contact the health center, doctor or family member through the available wireless communication network. Preferably, the subject has the opportunity to cancel such urgent contact if the subject feels he/she can handle the serious situation or can get the help nearby. In this case, only if the subject feels necessary to make such an urgent contact or he/she is incapable of canceling such an urgent contact, the mobile device will make the urgent contact and translate the necessary information.

[0033] With the integration of the physiological signal monitoring and physical activity monitoring, the present monitoring system can make more intelligent and more reliable health detection decisions since the health condition can be highly associated with the user's physical activity condition. For example, at normal resting condition, a heart rate of 60 to 100 beats per minute for a subject can be treated as normal. A jump to 120 or higher at the same activity condition for the same subject can imply a health condition change. However, if the subject is going through an activity change from the resting condition to run condition, such a heart rate jump can be considered as normal because the intense activity usually results in a heart rate jump within a certain range. If the heart rate jumps much higher than the normal range, it may still be necessary to be detected as the health problem. In the case that the heart rate becomes very low, it is another important health condition to identify. In another case, if the heart rate becomes irregular, such as missing heart beat or irregular beat duration along time, it can also imply a heart issue.

[0034] The contact sensor 23 is included to ensure that the healthcare system has been properly installed on the designed position to obtain the physiological and/or activity signals. Any improper position or installation of the device has adverse impact on the signal quality and monitoring reliability. Once an improper installation of the device is detected, the device can issue an audio warning signal such as long beep or voice warning (e.g. "Please check the device position") so that the user can make sure the device works properly.

[0035] Depending on the health condition detected by the monitoring system, various output actions can take place. Examples of output actions that may be triggered are an emergency call/transmission (page or phone call) through RF 44 for a very serious condition, activation of the smart audio outputs such as beep, advice, reminding or warning through speaker 41 , audio path 42 and audio interface 43 to the ear canal for a concerned health condition, data storage on the CPM or transmission through RF 44 for the future analysis or review purpose. The medical expert knowledge can then be applied to the obtained information with the subject's health data, the pre-determined alarm setting and urgent contact requirement.

[0036] The monitoring system 1 may include an short-range wireless unit 44 to communicate with a mobile device, which consists of an short-range wireless transmitter for one way communication to send out the subject's health urgent condition that may include the detailed health information or subject's personal information; or an RF transceiver for dual way communication to send out health information and to receive the necessary medical or action instruction. The RF unit of the intelligent healthcare system is designed to communicate with the PDA or cell phone for the short distance coverage such as 1.5 m to save system power consumption as the subject will carry the PDA or cell phone all the time within the coverage,

[0037] The monitoring system 1 may also communicate, through the RF unit 44, with a mobile unit that may have included the global positioning or navigation system capability so that the user's current geographical location can be identified by the clinic center, doctor or family member.

[0038] The user setting 31 of the monitoring device can be adjusted by the subject for regular monitoring over a long duration such as 30 minutes, 5 minutes or 1 minute for power saving purpose or for continuous monitoring. Even working in the different user setting modes, the system can adapt to the health situation by, for example, adaptively switching to real-time continuous mode in case of health issue detected. Therefore, the system can achieve both power saving purpose and full-on engagement monitoring when necessary.

[0039] The user cancel 32 is to cancel an automatic emergent call when an urgent and serious health condition is detected by the healthcare system. Only if the user thinks it is necessary to send this request or the user is incapable to cancel the emergent request, the user can cancel such a request to reduce the false alarm.

[0040] The user request 33 is for the user to request a current health state update or an urgent call. The button of user request can be pushed shortly for the current medical state update as smart audio outputs or displayed over the available mobile unit. The same button of user request can be pushed with hold for a certain time such as 2 seconds as an urgent call. In this case, the critical health information of the user may be sent to clinic center, doctor or family member to determine the subject's health condition and the necessary help. This user control enables the user to be able to check his/her health state or manually seek necessary assistance for a variety of conditions, including injuries from a fall or an automobile malfunction. It is also beneficial to provide the geographical coordinate locations with the emergency call if the global positioning capability is included in the mobile unit. Another example of user controls is to request a data-save action in conjunction with the intelligent signal processing so that the user or doctor can obtain the necessary medical information for the time being the user feels or wants to save.

[0041] The healthcare system may include a Device ID, which comprises a unique identifier for each monitoring device and its user. This identifier may be included with data transmission, and is used by the receiving end (e.g., 911 call center or clinic center) to identify the source device of each transmission. Each device ID is mapped to a particular subject, so that the receiving center can identify the subject and take the necessary action to response the request or inform the user's family member.

[0042] The healthcare system may also include a basic subject profile such as name and contact phone number in the data transmission for the particular subject wearing the device. The subject profile may include more subject information such as medical history and current medical conditions. This is useful for situations in which a Subject Profile Database is not available. For example, if the device transmitter is a cell phone, and a call is triggered to a 911 call center which does not have access to the Subject Profile Database, the device may transmit the subject identifier, name, address, medical history, current medical conditions, current geographical coordinate locations and other information as necessary to the call center.

[0043] Upon the detection of an urgent health condition, the healthcare system may start a transmission sequence that includes dialing sequences for issuing a page or phone call. A device may have more than one transmission sequence. For example, one sequence may be used to call a 911 call center for an emergency condition, and another sequence may be used to call the clinic center for status reporting. Another sequence may be used to call a family doctor or the family member for help.

[0044] Historical and current health information can be collected from the monitoring device for a specified period of time, or for a specified number of data collections. The health information is extracted and saved on the device, or it is sent out in an emergency transmission. For information only purpose, the health information such as heart rates or sleep apnea collected over certain time duration such as every 15 minutes for the past week or month may be analyzed and then updated. The information may be extracted and downloaded to a computer on a periodic basis for observation or evaluation purpose.

[0045] I/O Interface 45 can be a standard communication interface such as Universal Serial Bus (USB) port between the system and the external computer or device. The health information can be downloaded to the external computer or device for further analysis; the new system code or the new parameters can be uploaded into the system to upgrade the system or performance. [0046] Battery 51 is preferably a low voltage power supply such as 3V or lower for the whole system. The battery may be a one-time battery, rechargeable battery or any new type of power supply. The system has one or multiple internal battery level thresholds to trigger preset low battery warning or the system continuously checks the battery level with the pre-set thresholds. Once a low battery level is reached, the system will emit a corresponding low battery reminder or warning signal to inform the user to exchange a new battery or recharge the battery. In the meantime, the system will make the necessary update or save the most recent health information.

[0047] The monitoring device is usually worn by a user on the specified position around the ear. That is, people with health concerns or health history can use the monitoring device for health assistant device, or people with no known medical history can use the monitoring device as a safeguard or simply a self-health check/survey purpose; athletes may employ the present devices to monitor their own physical condition during competition, practice or training; parents may use the healthcare system to monitor and care for their children or infants, and the most importantly, the elder people can use the device to monitor their physical activity and health condition during their daily life.

[0048] The healthcare system can provide many types of medical monitoring device. With the medical progress, new medical sensors with new detecting technology can be integrated into the healthcare system. Examples of detection include: blood oxygen level, heart rate or pulse, blood flow information, body temperature, sleep apnea, glucose, exercise amount, unexpected fall or any type of health sign or activity that may be detected by the monitoring device.

APPLICATION 1 : SYSTEM AND METHOD FOR HEALTH AND ACTIVITY MONITORING AND FEEDBACK

[0049] FIG. 1-7 illustrates, for an intelligent health system, the relations between activity level, health condition, and expert systems or medical experts, according to a particular embodiment. [0050] FIG. 1-8 illustrates the use of remote expert systems or medical experts to provide further monitoring or feedback according to a particular embodiment.

[0051] FIG. 1 -10 illustrates the positioning of the healthcare monitor on an individual's head. [0052] FIG. 1-11 illustrates activity thresholds according to a particular embodiment.

[0053] FIG. 1-12 illustrates activity detection based on acceleration according to a particular embodiment.

[0054] FIG. 1 -13 illustrates activity dynamic status and activity index according to a particular embodiment.

[0055] FIG. 1 -14 illustrates the effect of activity on health according to a particular embodiment.

[0056] FIG. 1 -15 illustrates the effect of activity on heart rate according to a particular embodiment.

[0057] FIG. 1-16 illustrates the effect of activity on respiration rate according to a particular embodiment.

[0058] FIG. 1-17 illustrates the effect of activity on SpO2 according to a particular embodiment.

[0059] FIG. 1-18 illustrates the effect of activity on blood pressure according to a particular embodiment.

[0060] FIG. 1-19 illustrates the effect of rehabilitation activity on health improvement according to a particular embodiment.

[0061] FIG. 1-20 the threshold detection of an individual's health condition based on a life parameter according to a particular embodiment.

[0062] FIG. 1-21 is a graph illustrating health state detection based on a life parameter according to a particular embodiment. [0063] FIG. 1 -22 the effect of activity on a subject's health state and detection thresholds according to a particular embodiment.

[0064] FIG. 1-23 is a block diagram for Overall Health Analysis and Feedback according to a particular embodiment.

[0065] FIG. 1-24 is a flowchart for Overall Health Analysis and Feedback according to a particular embodiment

[0066] FIG. 1-25 is a Schedule of Activity and Level for Rehab according to a particular embodiment.

[0067] The following paragraphs provide further detail on the use of a healthcare system such as the one described above to monitor activity level and health condition in order to provide feedback to a user according to each individual's personalized health condition. As shown in FIG. 1-7, Activity and Health can have an impact on each other and can be adjusted based on feedback to provide beneficial effects:

Activity-> Health; Health-^Activity; Activity & health -> expert analysis; Expert system/medical expert^activity styles health

[0068] The monitoring of activity and health condition at the individual level can also be part of a system of healthcare with further monitoring and feedback conducted via a wireless connection as illustrated in FIG. 1-8.

[0069] The systems and methods described are intended to provide at least some of the following:

1. Define or suggest a most favourable activity style and amount for each individual user.

2. Monitor an individual's activity according to the personalized activity style and amount;

3. Manage individual activity and health in reference to a recommendation; 4. Work independently without sending activity and health information through a communication network to get further analysis, to provide an intelligent and real-time system.

5. Warn or alarm with regard to any harmful activity according to an individual's real-time health condition, ideally before it causes any potential risks;

6. Collect both activity patterns and their effect on health patterns, which helps to modify or optimize the health activity management;

[0070] Using the systems and methods is intended to allow an expert system or medical expert to review and modify the effect of an individual's activity on health relationship to improve real-time activity monitoring. In this way, a single real-time system provides information on activity's affect on health and health feedback on activity in an intelligently integrated way. The systems and methods herein are expected to be particularly helpful in rehab management and the like.

[0071] The healthcare system herein is intended to present a unique way to measure the activity level and activity type while taking account of personalized factors relating to an individual user's health condition.

[0072] In this document, four different activity types are used to describe a person's activity types: quiet, light, medium and strong. In this definition, "light" and "medium" are recommended activity styles for each individual and "strong" means some level over the recommended or preferable activity level. For each individual, these thresholds are usually different considering individual age, health condition, life style, and purpose of activity. They may start from an average threshold based on the same kind of person with similar age and health condition, but the thresholds can be updated according to individual's actual health condition, activity style or purpose of activity.

[0073] In order to start with an optimized activity threshold for any individual, a pre- examination of health is preferred, which should include the person's health condition, activity style and some information regarding the effect of activity on health through inquiry and a few typical activity tests. Based on this pre-knowledge, a suggested or recommended activity type and amount of activity can be applied according to the user's activity style and health condition. Thus, personalized activity detection and monitoring are configured into the system.

[0074] An overall activity index can be measured by considering an individual's activity type and duration, which reflects the overall activity status. As we know, different persons have different activity styles for different health conditions; the same amount of physical activity may be treated as light activity style for a healthy person but may become a strong activity type for an elderly person or a person with a health problem such as a heart or lung problem.

[0075] For each person, a health condition check can be done first and activity type is identified according to the person's actual activity style and health condition. In this way, a set of pre-initialized activity detection thresholds are produced, for example, by a health inspector, doctor, and authorized service person or by self-checking. (Fig. 1 -11 Activity Thresholds).

[0076] Once pre-initialized or self-checked activity thresholds are set, they can be used for activity detection (Fig. 1 -12: Activity detection within T). A time unit T such as 3s is set to the detect the activity amount and this detection result will be used to determine the activity status within this time duration T, and the result in T will be used to calculate its contribution to the current activity status and long-term activity affection. (Fig. 1-11. Activity Detection Thresholds; Fig. 1 -12. Activity stats detection).

1. Detect activity amount based on 3-D acceleration sensors: x, y and z (Fig. 1-9)s)

2. Detect human's activity by positioning the sensors on head, especially positioning on ear (Fig. 1-10);

3. Determine the activity information based on change of acceleration (Fig. 1-12) with the pre-defined Activity Thresholds (Fig. 1-11 );

4. Activity amount at time t: Activityft] = |dx|+|dy|+|dz|

[0077] The thresholds of "strong", "medium" and "light" have now been pre-set according to person's health condition and activity style. These thresholds can be varied by a medical expert through data analysis or self-learning in real-time according to the correlation between activity and health. [0078] Within each time TO, such as, for example, 3s, it is possible to calculate the approximation activity degree:

N_S is the number of the activity falls into the Strong class. N_M is the number of the activity falls into the Medium class. N_L is the number of the activity falls into the Light class. N_Q is the rest of all other lower activity level.

[0079] As we know, within a certain time unit T, if we only consider the physical activity level and summing them together, it doesn't necessarily reflect the mean activity level. Such an example is: if a person keeps walking during T, it has the physical activity level X. If he jogs for 1/2T and rest for 1/2T, he/she may have the same amount of activity level, but the actual activity result may be higher than the "walk". In the same way, if a person runs for a certain time such as 1/4T, his overall activity level may still be in similar level as "walk", but the actual contribution to effects on health can be higher than "walk" or "jog". Therefore, a weight factor is introduced to integrate all activity types' contribution to the overall activity index.

A_ S = w_LA_L + w_M A_M + w_sA_s

And the decision of Activity type in this time duration T can be decided by: 1 ). If A_s > A_S_Mm , activity is "Strong", else:

2). If A_s > A_M_Mm , activity is "Medium", else 3). If A_s > A_L_Mm , activity is "Light", else 4). Activity is "Quiet".

[0080] In Fig. 1-13, when a person starts activity, it will go through a warm-up phase and then follow by a continue phase depending whether the activity is consistent and continue for a certain time. Then the person slow-down or stop the activity, which will have a hold phase for a short period. Then the release phase will start and its last time will depend on an individual's activity strength and health condition.

[0081] In FIG. 1 -13, the activity dynamic status (Fig.1 -13: Solid Curve) reflects the actual activity amount over time: Quick attack and slow release are shown, including an Activity accumulation phase, activity stable phase and activity release phase, generally matching with Warm up phase, stable phase and relief phase; [0082] The overall accumulation of activity is described as Activityjndex (Fig.1-13: Dash

Curve), which reflects the physical status over time on a human body.

for warm-up phase. After a certain time T with a certain activity level, it will reach a stable phase with the constant activity level. Once activity stops, it will go through a physical release phase

^ = ^⁽'-ⁿ when ' > 7\

[0083] FIG. 1 -14 shows the activity affection on health of a given level of activity (solid line): For a health individual, the affection on health will be rather lower and will warm-up in a certain speed and release in a short time after activity stops as shown in the dotted line curve for person A. However, for an aged person or individual with poor health, affection will start quickly with big affection and then remains at a higher level. After activity stops, the effect may hold for a long while and then release slowly over a long time as shown in the dotted line curve for person B.

[0084] The effect of activity on health in general can also be broken down into an effect on specific physiological indicators, at least some of which may be used in the formulation of the health variable shown in, for example, FIG. 1 -14.

[0085] Activity affection on Heart (Fig. 1 -15): According to medical knowledge and individual experience, in general, stronger activity will cause higher heart rate raises as the heart needs to pump blood faster than normal to supply required energy and oxygen. The activity affection on heart rate also relates to duration of activity. In general, in certain activity level, the longer activity, the heart rate will rise to a certain high level and then it may remains the certain level for certain constant activity. Once activity stops, the heart rate will remain at higher level for a short time and then gradually slow-down to the normal level.

[0086] Activity affection on RR (Fig. 1 -16): The effect of activity on RR may be similar to the effect of activity on HR. [0087] Activity affection on SpO2 (Fig. 1 -17): The affection on SpO2 may be different from HR and RR. Activity may consume more oxygen than normal, which causes SpO2 to drop when activity level increases. Therefore, the curve of SpO2 may change quite differently compared with HR or RR. In general, stronger activity may cause more of a drop of SpO2 than in normal case.

[0088] Activity affection on Blood pressure (Fig. 1 -18): Blood pressure may rise when activity level increases or duration of activity increases. However, some individuals can have blood pressure drop when activity level increases. There might be some difference in activity affection on blood pressure.

[0089] FIG. 1 -19 shows the effect of Rehabilitation Activity on Health improvement. Continuous monitoring and feedback may be used in rehabilitation therapy, enabling an individual to receive efficient rehabilitation management to recover his/her health to normal or healthier level, or from a current level to a target level.

[0090] FIG. 1-20 shows, for a particular embodiment, the threshold detection of an individual's health condition based on a life parameter, such as HR, RR, or Blood pressure, in this case when the individual is not engaged in activity. For the case of SpO2, the detection of the individual's health condition is typically low-limit detection, as the upper limit is typically 100% for normal health. When considering the correlation between the individual's activity level and the life parameter, the health detection thresholds for both lowjimit and highjimit can be affected by a certain correlation factor. A life parameter may have different correlation factors for different thresholds, and different life parameters may have different sets of correlation factors.

[0091] An example formula that may be used to describe detection thresholds for a life parameter with correlation of activity level is given below:

High _ Limit - High _ Limit _ Static + λ_H * Activity _ Index

Low _ Limit - Low _ Limit _ Static + λ_L * Activity _ Index [0092] FIG. 1-21 shows, for a particular embodiment, health state detection based on a life parameter as a function of time. The life parameter may be Heart Rate, SpO2, Respiration Rate, Blood Pressure or any other parameter. For an individual, a set of detection thresholds for high limit and low limit may be determined according to the individual's health history, current health condition, recommended health expectation for rehabilitation, or other criteria.

[0093] For example, when an individual's Heart Rate (HR) is beyond the normal range between High Limit for Attention and Low Limit for Attention, a feedback message may be issued based on the relationship between the current HR and thresholds for health states. In this example, four health states may be defined as "Normal", "Attention", "Warning" and "Alarm". If HR is higher than the High Limit for Attention and lower than the High Limit for Warning, the health state is detected as "Attention" and a message for Attention may be issued. If HR is higher than the High Limit for Warning and lower than the High Limit for Alarm, the health state is detected as "Warning" and a warning message may be issued. If HR is higher than the High Limit for Alarm, the health state is detected as "Alarm" and an alarm message may be issued. Similarly, HR is compared to the Low Limits for the health states. If HR is within the High Limit for Attention and the Low Limit for Attention, the health state is detected as "Normal".

[0094] Referring to Fig. 1 -21 , the health state is detected as "Normal" before time t1. As the life parameter changes over time, the health state is detected as "Attention" after t1 and then becomes "Warning" after t2.

[0095] FIG. 1 -22 shows Effect of Activity on Health State according to a particular embodiment, and illustrates the life parameter curve of FIG. 1 -21 , as well as a detected activity factor (Activityjndex) plotted as a function of time. Life parameters such as HR, SpO2, RR, Blood Pressure, or other parameters may be correlated with the Activityjndex. Changing activity levels may result in corresponding changes in detection thresholds. An example formula that may be used to express the correlation relationship is given below:

High _ Limit = High __ Limit _ Static + λ_H * Activity _ Index

Low _ Limit = Low _ Limit _ Static + λ_L * Activity _ Index [0096] In the above expressions, the HighJJmit includes a limit and a correlation factor λ_H for each of the health states, Attention, Warning and Alarm. The correlation factor λ_H may be same or different according to the individual's health condition and activity affection on health state. The correlation factors may be constant or variable with time or another factor. The same principles apply to the LowJJmit, and its correlation factor λ_L may be different for each respective limit and different from the correlation factors for the HighJJmit. In this figure, the detection thresholds have been adjusted according to the activity level, and the life parameter curve remains within the range for the Normal health state.

[0097] FIG. 1-23 is a block diagram for Overall Health Analysis and Feedback according to particular embodiment. A human being is a complicated system and its overall health state may be determined by considering the contribution of various life parameters of the individual. After detecting individual life parameters and determining their contribution to the individual's health state including activity correlation, the life parameters may be integrated to determine an overall health state.

[0098] As illustrated in Fig. 1 -22, a particular level of activity may have different effects on an individual's life parameter. This is shown in FIG. 1-23 as the block Health Detection with Activity Correlation, which incorporates correlation factors λ_H and λ_L . The block Overall Health Analysis and Feedback in Fig. 1-23 represents a logical analysis process and is further described in Fig. 1 -24.

[0099] FIG. 1 -24 shows a flowchart for Overall Health Analysis and Feedback according to a particular embodiment. The analysis process and logic conditions may be adjusted according to individual's health condition and expert knowledge.

[00100] FIG. 1 -25 shows a Schedule of Activity and Level for Rehab according to a particular embodiment. An activity schedule may be created based on individual medical history, current health condition and expected rehab projections. The activity schedule may include recommended start times for activity, recommended durations for activity, recommended activity levels, and expected effects on the individual's health state. A friendly reminder message may be delivered to the user at recommended start times and stop times. A encourage or discourage message may be delivered to the user based on real-time health state detection, activity detection and correlation analysis. In this way, favorable times for activity, favorable activity levels and favorable amounts of activity may be achieved to obtain the efficient health recovery.

[00101] Individuals may have similar or very different effects of activity on his/her health parameters such as HR, RR, SpO2 or blood pressure. Detecting an individual's personal activity level and amount may be helpful in maintaining and improving one's health. For example, the same degree of physical activity can be treated as light activity for a person or strong activity for a different person. Therefore, a concept of personalized activity level and duration are provided here, which includes individual's health information and the potential activity affection on his/her health. The personalized activity types and amount can be used to indicate whether an individual is doing suitable activity level and duration at the preferred time in a preferred style.

[00102] The system and method can detect a person's actual activity in real-time and determine whether his/her activity level is favourable and activity duration is favorable by comparing the health affect with either historical data for that individual or with average data for other individuals with a similar age, physique, health condition or the like. The system can then provide feedback to the individual with regard to raising/lowering their activity level and/or the duration of the activity in order to adjust the impact on their health. As a simple example, the system can monitor a user's heart rate during activity and provide feedback to the user if the heart rate is outside of the optimum range for increasing cardio-vascular health or for fat burning or the like. In another example, the system monitors a user's SpO2 during activity and provides feedback to reduce activity when SpO2 falls too low.

[00109] The system and method can then continue to detect a person's activity and the effects of their activity on their health in real-time, and compare with previous results to determine/diagnose whether a level of activity has helped to improve one's health or decrease one's health (for example if the activity has caused issues with blood pressure or the like). [00110] When integrated with a wireless system or when attended by a healthcare provider, a person's activity level, duration and its effect on health can be detected in realtime or downloaded and can be analyzed, for example, by a healthcare provider or medical expert. As a result, a user can get prompt recommendations, attention, warning or alarm related to his/her activity and health condition. This can help a person to improve his/her health through activity management. This also can prevent any health damage from harmful activity, under or overdoing an activity or the like.

[00111] The aspects described herein can be adapted to provide more or less physiological signal monitoring or an alternate activity monitoring system without departing from the spirit or essential attributes thereof. On the other hand, the aspects described herein can also be expanded to include more signal detection such as environment detection, weather detection, acoustic signal detection or even subject's emotion detection, in addition to the described health monitoring, without departing from the spirit or essential attributes thereof. It will also be understood that the processes and apparatuses may be implemented using hardware or software components or an appropriate combination thereof. Software may be provided as instructions on a physical computer medium or the like for execution on a processor of a computing device to perform the functions described.

APPLICATION 2: SYSTEM AND METHOD FOR INTELLIGENT FALL DETECTION

[00112] The following description describes a fall detection system and method in further detail. It will be understood that the fall detection system may include the elements of the healthcare system described above or may include an appropriate subset of those elements. The fall detection system is intended to reliably detect a fall by monitoring both static and active states of body activity and by confirming a detected fall by checking the health state of the person and the activity state of the person around the time of the physical fall action. As the fall detection system involves the use of many variables rather than just an accelerometer or the like, it may be referred to as an intelligent fall detection system.

[00113] The fall detection system is intended to meet at least some of the following conditions: 1. The monitoring device records data and detects fall only when it is correctly positioned on the ear

2. Detect falls only when the user starts from an upright position, such as standing up or sitting up (static)

3. Apply active fall detection process during a fall (active)

4. Only ends with lay-down (static)

5. Integrated health information for judgment

6. An alarm can be cancelled by the subject if not necessary

[00114] The intelligent fall detection system is intended to result in fewer false alarms and to reduce the resources needed to respond appropriately. Incorrect positioning of the monitoring device on the subject's ear may result in data being incorrectly recorded and falls incorrectly detected. This may be addressed by recording data and detecting falls only when the monitoring device is correctly positioned on the ear and by notifying the user if the monitoring device is incorrectly placed.

[00115] The position of the subject when the monitoring device is initially placed on the ear may also contribute to incorrect detection of falls. For example, if a reclining subject were to put on the monitoring device and then quickly sit up or stand up, a fall might be incorrectly detected. This may be addressed by detecting falls only when the subject is starting from an upright position, e.g. sitting up or standing.

[00116] When the subject is in a vehicle, the motion of the vehicle may contribute to inaccuracies in the detection of falls. This may be addressed by eliminating the effect of the vehicular motion using signal processing, thereby allowing the detection of falls to be primarily based on the motion of the subject. Digital signal processing can be used to determine and remove these types of acceleration from the signal.

[00117] FIG. 2-7 illustrates the state of the person before a fall. In this case the person starts in a standing-up position and the fall detection system is worn on the body in the correct position. The sensor's x, y, and z axes are respectively aligned with the reference frame's X, Y, and Z axes. The fall detection system then provides a working model in which the sensor's z axis is perpendicular to the plane defined by the X and Y axes. [00118] FIG. 2-8 illustrates the state of the person during a fall. During a fall, there is a high acceleration that changes above a predetermined threshold and lasts over a predetermined time duration (based on tests of a physical fall process of a human body). The result is that the body orientation is changed so that the previous z axis of the sensor is now on plane defined by the X and Y axes of the reference frame. The use of acceleration and time thresholds will differentiate a fall from the normal lay-down action or normal body activity such as walking, running, cough or other non-fall action.

[00119] FIG. 2-9 illustrates the state of the person after a fall. In the case of a true fall, the body will generally remain in a state for a certain time on the plane defined by the X and Y axes, which means that the subject loses its capability to stand up from the fall. The sensor's z axis will be on the plane defined by the reference's X and Y axes. So far a fall is detected.

[00120] The fall detection system may be set such that only a detected fall that meets the following conditions will be alarmed:

-The person remains in lay-down state without standing up within a certain time: this means that the person either have difficulty to stand up or losing consciousness because of fall.

-The person's critical health parameters such as HR₁ SpO2 or RR has a change that is outside of the normal variation range for that person, which means a sudden fall may have caused the person's health condition to decline or the fall happens because of the person's declined health condition.

[00121] In some embodiments, a user may cancel a fall alarm within a predetermined period of time such as, for example, 15s or the like before help is actually dispatched. In this case, even although a fall alarm condition has been met, the user may still decide that he/she can handle such a fall without needing help. Only if the user either feels needing such alarm or he/she cannot react to cancel the alarm, the alarm will be activated and sent out for help. [00122] It will be understood that the fall detection system or healthcare system may also provide the capability for an SOS to be issued even if the fall alarm condition is not met, in case the person feels necessary to get help;

[00123] In some embodiments, the fall detection system may include GPS position capability or GPRS localization for the identified fall alarm in order to provide help to the user in the quickest time by pinpointing their location for care providers and the like.

[00124] The following paragraphs provide some additional details relating to an example embodiment of the fall detection system and method.

[00125] In detecting a fall, the angle change is about 90 degree for a person falls from standing or sitting position to floor, and this angle change doesn't matter how tall the person is. However, when the fall speed is detected by a sensor worn on the earlobe, the fall speed (through acceleration detection) detected by the healthcare device can vary depending on the height of the person. A typical formula can be: a(t) = x(t) + y(t) + z{t) where χ(t) is the acceleration along x-axis, y(t) is the acceleration along y-axis, z(t) is the acceleration along z-axis and a(t) is the acceleration vector expressed with acceleration sensor axis. The acceleration sensor z-axis has angle φ = 0 to the reference Z-axis, which is the vertical line of human with body standing up. The reference axis X, Y and Z is referred to 3-D as the X_Y to express the plane defined by the X and Y axes, i.e. the floor. The same acceleration a(t) = x(t) + y(t) + z(t) can be also expressed as a{t) - X{t) + Y(t) + Z(t) with the reference of X,Y and Z axis. In the stand up position, the sensor's x, y, and z axes are overlapping with the reference's X, Y, and Z axes as shown in Fig. 7.

[00126] During a sudden fall, the sensor will move with the user starting with the angle φ = 0\o a new φ , with angle acceleration a_φ as shown in Fig. 2-8. Here we can see the height d of a user will affect fall speed.

a_z = a_φ * d * sin(^J) [00127] When a user completely falls down on the floor, sensor's axis has changed 90 degree to the reference 3D floor as shown in Fig. 2-9.

[00128] The acceleration ait) at time t can be easily expressed as acceleration vector: a(t) = x(t) + y(t) + z(t) = X(t) + Y(t) + Z(t) . The actual fall can be detected along Z axis and X_Y plane. The Z-axis acceleration a_z = a_φ * d * sin(#>) will change at a limit level for normal activity, but it will change in unique way together with X_Y plane activity during a fall.

[00129] For example, consider a person A has height ^{di mά} ^^{€rsm B te άL} . FIG.

10A illustrates a taller person A and FIG. 1OB illustrates a smaller person B. Person A is taller than person B, and we have ^{** Gt*} * ^^βm' ^{f #} linear acceleration for person A is much higher than person B, i.e. ^{VA > VM} and a_z] > a_Z2

az_A = a_φ * dl * ύn(φ) a_ZB = a_φ * d2 * ήn(φ)

[00130] FIG. 2-11 illustrates acceleration profiles for walking, normal lying down and for falls of a taller person and a shorter person. By setting appropriate thresholds, based on, for example, height, it is possible to detect an action that appears to be a fall.

[00131] As illustrated in FIG. 2-12A, 2-12B and 2-13, similar considerations can be made with regard to differences in a fall from standing position to the floor or from a sitting position to the floor for the same person. In this case, if height d is the standing distance from the floor, h is about d/2 for the sitting position to floor.

[00132] The aspects described herein can be scaled down for physiological signal monitoring system only or activity monitoring system only without departing from the spirit or essential attributes thereof. On the other side, the aspects described herein can be expanded to include more signal detections such as environment detection, weather detection, acoustic signal detection or even subject's emotion detection, in addition to the described health monitoring, without departing from the spirit or essential attributes thereof. APPLICATION 3: EAR HOOK ASSEMBLY

[00133] In the particular case of devices to be worn on or around an ear, there are additional challenges associated with securing the device to the ear, including: accurately and reliably positioning the device relative to the ear; accommodating variation in the shapes of individuals' ears; maintaining contact between the device and the ear at necessary points; ensuring the user's comfort despite prolonged, continuous use of the device; and providing a form-factor that permits the integration of various audio outputs, sensors, processors, communication interfaces and other elements.

[00134] The present application is directed to a system for continuous real-time monitoring of a subject's health condition with intelligent detection and analysis capability, smart warning or reminder of an urgent health condition and storage of an individual's health information without interrupting an individual's daily life. The present application allows the system to load a doctor's voice as warning/ reminding/instruction message or the voice of a family member for reminding purposes, in which emotional factors may help users, especially elderly ones, feel warm and disposed to take the necessary action.

[00135] The intelligent healthcare system may be setup to store current medical information and detect any pre-defined alarm conditions, such as a heart attack. Upon an occurrence of such alarm conditions, the device may provide smart audio outputs such as warning, advice or reminder to the subject for a situation of concern or contact the healthcare center, doctor or family member with health information for the necessary healthcare or medical assistance for serious situation.

[00136] In view of the above, it is an object of the present application to address at least some of the problems associated with securing a device or components of a system to an ear.

[00137] It is another object of the present application to provide a healthcare system that includes an ear hook assembly, wherein the ear hook assembly may include one or more sensors and an interface for communicating data. [00138] According to an aspect of this application, there is provided an ear hook assembly, which includes a hook that is attached to a clamp assembly, wherein the clamp assembly includes an exterior clamp member, an interior clamp member, and a coupling.

[00139] In a particular case, the force exerted by the coupling is adjustable.

[00140] In yet another particular case, the exterior clamp member may further include a gripping surface, and the interior clamp member may further include a gripping surface.

[00141] In yet another particular case, the ear hook assembly includes an in-the-ear portion that is attached to the hook, wherein the in-the-ear portion rests at least partially in the intertragic notch of the subject's ear.

[00142] In still another particular case, the ear hook assembly includes an interface for communicating power, data, or acoustic signals.

[00143] In another particular case, the ear hook assembly includes an audio output.

[00144] In yet another particular case, the ear hook assembly includes a sensor. In this particular case, the sensor may be a physiological sensor, an activity sensor, an environmental sensor, or the like.

[00145] In still another particular case, the ear hook assembly may include a processor and appropriate power supply.

[00141] FIG. 3-7. is a cross-sectional view of an embodiment of the ear hook assembly.

[00142] Fig. 3-8 is a perspective view of an embodiment of the ear hook assembly as worn by a user.

[00143] FIG. 3-7 shows a cross-section of an embodiment of an ear hook assembly.

The ear hook assembly includes a hook 708 attached that is attached to a clamp assembly 700. The hook 708 engages an intertragic notch 810 without significantly blocking the acoustic path to the ear, and the end of the hook is shaped to prevent injury or significant discomfort to the user. The clamp assembly 700 includes an exterior clamp member 702 and an interior clamp member 704, which are urged toward each other by a coupling 706. Ideally, the lower limit of the force exerted by the coupling 706 is the force sufficient to induce contact between the exterior clamp member 702 and the earlobe, and the interior clamp member 704 and the earlobe. The upper limit of the force exerted by the coupling 706 is the greatest force that can be sustained for the intended duration of use against the earlobe without causing injury or significant discomfort to the user. In a further embodiment, the force exerted by coupling 706 may be adjusted to accommodate differences in earlobe proportions.

[00144] In a further embodiment, the exterior clamp member 702 and the interior clamp member 704 may be provided with gripping surfaces 730 and 732 to facilitate gripping. Gripping surfaces 730 and 732 may include a shallow depression, a textured surface, a high- friction material such as rubber, or any other gripping surface that would be apparent to one that has ordinary skill in the art.

[00145] In a further embodiment, the ear hook assembly includes an in-the-ear portion

712 that is attached to the hook 708. In this embodiment, the in-the-ear portion 712 rests in the intertragic notch and substantially bears the weight of the ear hook assembly. The hook 708 may contain an electronic path or an acoustic path between the in-the-ear portion 712 and the clamp assembly 700.

[00146] In yet a further embodiment, the ear hook assembly includes an interface 790 that may be housed in the clamp assembly 700 or the in-the-ear portion 712. The interface enables the transfer of one or more of power, data, or acoustic signals between the ear hook assembly and a device, and may include a physical connection or a wireless connection.

[00147] In yet a further embodiment, the ear hook assembly includes an audio output

740. The audio output may be housed in the in-the-ear portion 712 or the clamp assembly 700, and provides audio output that is audible to the user, such as alarms, voice reminders, or other audio data.

[00148] In yet a further embodiment, the ear hook assembly includes a sensor 720,

722, 724, 726 or 728, which may be housed in the clamp assembly 700 or the in-the-ear portion 712. Sensor 720, 722, 724, 726 or 728 may be used to observe physiological data, such as oxyhemoglobin saturation (SpO2), body temperature, or glucose levels, or other data such as acoustic data or environmental data, and may include any sensor now known or hereafter developed. The placement of a sensor on or in a particular location may facilitate optimal operation of the sensor. For example, housing a body temperature sensor in the in- the-ear portion 712 may reduce the effects of ambient temperature on the observed measurement.

[00149] In yet a further embodiment, the ear hook assembly includes a processor 780 or 782, which includes a low-power signal processor.

[00150] In yet a further embodiment, the ear hook assembly includes a power supply

770. The power supply 770 may include, a ceil battery, rechargeable battery or kinetic battery, and supplies power to any one or more elements of the ear hook assembly, or to a device connected to the ear hook assembly by interface 790.

[00151] FIG. 8 shows an embodiment of the ear hook assembly for securing a device to an ear worn on the user's ear. The in-the-ear portion 712 is engaged in the intertragic notch 810 of the ear without significantly blocking the acoustic path to the ear, and substantially bears the weight of the ear hook assembly. Clamp assembly 700 clamps the earlobe 800, with the exterior clamp member 702 engaging the outward face of the earlobe 800, and the interior clamp member 704 engaging the inward face of the earlobe 800. Clamp assembly 700 bears a portion of the weight of the ear hook assembly, and keeps the ear hook assembly substantially immobile relative to the ear lobe during user movement and environmental disturbances, such as high wind. The urging force exerted by coupling 706 induces contact between both the exterior clamp member 702 and the earlobe 800, and the interior clamp member 704 and the ear lobe, which may facilitate optimal operation of a sensor 726 and 728. In this particular embodiment, gripping surface 730 is provided to facilitate gripping of the clamp assembly 700.

[00152] It will be understood that other embodiments may extend to other devices worn on the ear, including headphones, hearing aids, telephony headsets, and the like.

[00153] The intelligent healthcare system can be many types of medical monitoring device. With the medical progress, many new medical sensors with new detecting technology can be integrated into the present system. Examples of detection include: blood oxygen level, heart rate or pulse, blood flow information, body temperature, sleep apnea, glucose, exercise amount, unexpected fall or any type of health sign or activity that may be detected by the monitoring device.

[00154] The aspects of the present system can be scaled down for physiological signal monitoring system only or activity monitoring system only without departing from the spirit or essential attributes thereof. On the other side, the aspects of the present system can be expanded to include more signal detections such as environment detection, weather detection, acoustic signal detection or even subject's emotion detection, in addition to the described health monitoring, without departing from the spirit or essential attributes thereof.

Claims

Claims:

1. A wearable mini-size intelligent healthcare system comprising: physiological sensors for detecting physiological condition of a subject; a processing module for system control and real-time signal processing to detect physiological condition and to extract health information based on medical expert knowledge and subject's health history, and to generate system output signals in response to pre-determined parameters and storing/transmitting said health information; a speaker for the system to generate smart audio outputs and a audio path to delivery the acoustic signal to the subject's ear canal through audio interface without blocking the normal acoustic signal access and reception; a short range RF link for communicating said health information to mobile unit such as a PDA or cell phone, which can communicate with a reporting system, doctor or family member in response to a detected adverse health condition;

2. The intelligent healthcare system as in claim 1 , wherein the physiological sensors include oximetry sensor (SpO2), temperature sensor, glucose sensor or any other physiological sensor which can collect valuable physiological signals around the ear with the small size and low power consumption.

3. The intelligent healthcare system as in claim 1 , wherein physiological signals from the physiological sensors are detected and analyzed by the physiological signal detection algorithm to determine subject's physiological condition including vital signals such as blood oxygen level, heart rate, pulse rate, blood flow information, body temperature, blood sugar level and other physiological signals.

4. The intelligent healthcare system as in claim 1 , wherein said health information is detected according to the said physiological information such as blood oxygen level, heart rate, pulse rate, blood flow information, body temperature or glucose.

5. The intelligent healthcare system as in claim 1 , further comprising activity sensors, such as the piezo-resistive 3-axis acceleration sensors, with small size and low power consumption.

6. The intelligent healthcare system as in claim 5, wherein health information is detected in associating with the physiological information such as blood oxygen level, heart rate, pulse rate, blood flow information, body temperature or glucose and the physical activity information including subject's activity state, activity amount and fall detection for intelligent health monitoring.

7. The intelligent healthcare system as in claim 5, wherein activity signals from the activity sensors are detected and analyzed by the real-time activity detection algorithm to determine the user's physical activity state, activity amount and/or unexpected fall.

8. The intelligent healthcare system as in claim 5, wherein the real-time activity detection algorithm to determine activity index along time and activity pattern along time.

9. The intelligent healthcare system as in claim 6, wherein an individual has own adaptation of activity pattern. Different individuals can have different activity patterns for the same/similar activity, and it is important to detect an individual activity pattern for health evaluation purpose or rehab benefit. The healthcare system herein is intended to present a unique way to measure the activity level and activity type while taking account of personalized factors relating to an individual user's health condition.

10. The intelligent healthcare system as in claim 6, wherein a model for establishing activity effect on individual physiological signals over time.

11. The intelligent healthcare system as in claim 6, wherein a model for establishing activity effect on individual overall health state over time.

12. The intelligent healthcare system as in claim 6, wherein a model for analysis and detection of overall health state with the correlation of physiological signal and activity signal.

13. The intelligent healthcare system as in claim 6, wherein a system for continuing to monitor health states through the continuous analysis of the correlation between physiological signals and dynamic activity signal including activity speed-up behavior, activity release behavior and continuous activity behavior.

14. The intelligent healthcare system as in claim 6, wherein a self-learning system allows the user or expert to adjust or optimize the expected correlation of physiological signal and activity signal.

15. The intelligent healthcare system as in claim 6, wherein a valuable system for health research, medicine research, healthcare and rehab guidance in real-time with a pre-set goal and schedule with active and real-time feedback. The system manages individual activity level and health through the schedule with activity instruction or suggestion in reference to a recommendation.

16. The intelligent healthcare system as in claim 6, wherein health information is detected in associating with the physiological information such as blood oxygen level, heart rate, pulse rate, blood flow information, body temperature or glucose and the physical activity information including subject's activity state, activity amount and fall detection for more intelligent health monitoring.

17. A method for a wearable mini-size intelligent healthcare system, comprising: using the specific body position around the ear to easily and comfortably hold the healthcare system; using the same specific body position around the ear to acquire the necessary physiological signals from the earlobe, the body position behind the ear and activity signals of body; using the same specific body position around the ear to easily delivery the smart audio outputs of the system to the subject; detecting physiological condition of a subject based on the physiological signals from one or multiple physiological sensors; using center processing module for system control and real-time signal processing to detect physiological condition, to extract health information based on medical expert knowledge and subject's health history, and to generate system output signals in response to pre-determined parameters and storing/transmitting said health information; communicating health information and/or possibly geographic coordinates to a reporting system, doctor or family member in response to a detected adverse health condition.

18. A method according to claim 17, wherein the health information is detected according to the physiological information such as blood oxygen level, heart rate, pulse rate, blood flow information, body temperature or glucose.

19. A method according to claim 17, further comprising of detecting subject's activity condition such as activity state, activity strength and activity amount along time as well fall detection of a subject based on the embedded activity sensors.

20. A method according to claim 19, comprising of the real-time activity detection to determine activity index along time and activity pattern along time.

21. A method according to claim 19, wherein the health information is detected in associating with the physiological information and the physical activity information including subject's activity index and activity pattern for intelligent health monitoring.

22. A method according to claim 21 , further comprising of an advanced algorithm to monitor subject's sleep quality and apnea, which can be extremely useful to analyze a subject's health condition and prevent the serious respiration problem from happening. The early alarm of apnea or degraded sleep quality can be used to obtain prompt diagnose or even treatment from doctor.

23. A method according to claim 21 , comprising of analyzing an individual own adaptation of activity pattern and detecting an individual activity pattern for health evaluation purpose or rehab benefit. The method herein is intended to present a unique way to measure the activity level and activity type while taking account of personalized factors relating to an individual user's health condition.

24. A method according to claim 21 , further comprising of detecting and analyzing activity effect on individual physiological signals.

25. A method according to claim 21 , further comprising of detecting and analyzing activity effect on overall health state.

26. A method according to claim 21 , further comprising of determining the correlation for each level of the health states such as Attention, Warning and Alarm, and activity effect.

27. A method according to claim 21 , further comprise of Overall Health Analysis and Feedback. The analysis process and logic conditions may be adjusted according to individual's health condition and expert knowledge.

28. A method according to claim 21 , for healthcare or rehab purpose that an activity schedule is created based on individual medical history, current health condition and expected rehab goal. The activity schedule may include recommended start times for activity, recommended durations for activity, recommended activity levels, and expected effects on the individual's health state. An instruction or reminder message may be delivered to the user according to the schedule with general or expert message.

Aplication2: Fall Detection

29. The intelligent healthcare system as in claim 5, comprising of 3-Dimension Fall- detector wherein activity signals from the activity sensors are detected and analyzed by the fall detector for reliable fall detection with three levels: Weak, Medium and High for various sudden falls such as falling from standing or sitting.

30. The intelligent healthcare system as in claim 29, comprising of detecting a sudden fall by considering the individual's body physical condition such as height an important factor for the fall detection.

31. The intelligent healthcare system as in claim 29, a system that detects an individual's fall pattern for health diagnose and healthcare purpose.

32. The intelligent healthcare system as in claim 5, wherein health information is detected in associating with the physiological information and the physical activity information including subject's activity index, activity pattern and fall detection for intelligent health monitoring.

33. The intelligent healthcare system as in claim 32, a model for establishing correlation between physiological signal and fall signal;

34. The intelligent healthcare system as in claim 32, a model for analysis and projection of health state with physiological signal and fall pattern;

35. The intelligent healthcare system as in claim 32, a system that avoids false alarm by allowing the user to cancel an active alarm. The default alarm condition is when the individual cannot cancel the active alarm or doesn't want to cancel the active alarm within a certain time window such as 5s while the individual is still lay-down after falling down.

36. The intelligent healthcare system as in claim 32, a system further integrating both health information and fall detection for reliable fall detection and preventing false fail alarm from happening. A fall alarm will only be issued when a fall is detected and when the user's health condition suddenly degrades to need attention or healthcare help.

37. A method according to claim 17, wherein a 3-Dimension Fall-detector wherein for reliable fall detection with three levels: Weak, Medium and High for various sudden falls such as falling from standing or sitting.

38. A method according to claim 37, comprising of detecting a sudden fall by considering the individual's body physical condition such as height an important factor for the fall detection.

39. A method according to claim 37, comprising of detecting an individual's fall pattern for health diagnose and healthcare purpose.

40. A method according to claim 37, wherein health information is detected in associating with the physiological information and the physical activity information including subject's activity index, activity pattern and fall detection for intelligent health monitoring.

41. A method according to claim 40, wherein a model for establishing correlation between physiological signal and fall signal;

42. A method according to claim 40, wherein a model for analysis and projection of health state with physiological signal and fall pattern;

43. A method according to claim 40, wherein avoiding false alarm by allowing the user to cancel an active alarm. The default alarm condition is when the individual cannot cancel the active alarm or doesn't want to cancel the active alarm within a certain time window such as 5s while the individual is still lay-down after falling down.

44. A method according to claim 40, comprising of further integrating both health information and fall detection for reliable fall detection and preventing false fall alarm from happening. A fall alarm will only be issued when a fall is detected and when the user's health condition suddenly degrades to need attention or healthcare help. APPLICATION 3: EAR HOOK ASSEMBLY

45. The intelligent healthcare system as in claim 5, wherein the system is integrated into an embodiment of an ear hook assembly to be worn in the ear.

46. The intelligent healthcare system as in claim 45, wherein the system is continuous real-time monitoring of a subject's health condition with intelligent detection and analysis capability, smart warning or reminder of an urgent health condition and storage of an individual's health information without interrupting an individual's daily life.

47. The intelligent healthcare system as in claim 45, wherein the device is secured to the ear, including: accurately and reliably positioning the device relative to the ear;

48. The intelligent healthcare system as in claim 45, wherein the device is accommodating variation in the shapes of individuals' ears; maintaining contact between the device and the ear at necessary points; ensuring the user's comfort despite prolonged, continuous use of the device;

49. The intelligent healthcare system as in claim 45, wherein the device is providing a form-factor that permits the integration of various audio outputs, sensors, processors, communication interfaces and other elements.

50. A method according to claim 17, wherein a healthcare system is integrated into an embodiment of an ear hook assembly to be worn in the ear.

51. A method according to claim 50, wherein continuous real-time monitoring of a subject's health condition with intelligent detection and analysis capability, smart warning or reminder of an urgent health condition and storage of an individual's health information without interrupting an individual's daily life.

52. A method according to claim 50, wherein securing a device to the ear, including: accurately and reliably positioning the device relative to the ear;

53. A method according to claim 50, wherein accommodating device variation in the shapes of individuals' ears; maintaining contact between the device and the ear at necessary points; ensuring the user's comfort despite prolonged, continuous use of the device;

54. A method according to claim 50, wherein providing a form-factor that permits the integration of various audio outputs, sensors, processors, communication interfaces and other elements.