DE102005008386A1 - Patient health preventive monitoring procedure has body position, movement or environmental sensors with processor using data mining and decision techniques to make recommendations - Google Patents

Abstract

A patient (5) health preventive monitoring procedure has body position, movement or environmental influence sensors (4a-d) with processor (3) examining their outputs using data mining (14) and decision (10) techniques to determine trends and the crossing of critical thresholds and provide recommendations to the patient. Independent claims are included for equipment using the procedure.

Description

The This invention relates to a preventive health monitoring method a patient and a device for performing the mentioned method.

The individual health care, in particular the conscious adaptation a healthy lifestyle, takes in modern society in importance steadily increasing rank. One reason for this is below the steadily increasing life expectancy, due to which late damage, the caused or promoted by unhealthy lifestyles, to a greater extent than previously to step. In addition, there is an increasing burden on the human organism by harmful Factors such as stress, environmental pollutants, allergenic substances, etc. This is promoted in particular by the fact that unhealthy ways of life, especially due to lack of exercise, unbalanced diet, smoking, etc. embossed are inherently widespread in modern society are.

On the other side falls it the ordinary one Individual increasingly difficult for an effective, individual To recognize and implement health care measures to be taken. This is particularly due to the fact that the interaction network of health-influencing factors As medical science becomes more sophisticated, it becomes increasingly complex represents, so without special medical knowledge only hardly the overview to keep. The abundance the available standing, partly over-subscribed advertising and certain fads subject information is doing here the rest, to give the individual the right balance of health care to disguise rather than facilitate.

In front This background is a method or a device performing this desirable, this finding process on as possible simple and personalized way supports and thus the individual without big additional time required for effective health care.

On the other hand, so-called mobile patient monitoring systems are known from outpatient and post-hospital therapy. Such monitoring systems are described for example in US 2002/0169584 A1, EP 1 226 782 A2 and US 2002/0099273 A1. In such a patient monitoring system, vital parameters of the patient, in particular blood pressure, heart rate, body temperature, etc., are detected by one or more sensors, in particular worn on the body of the patient, and with regard to symptoms of an acute health deterioration, in particular a stroke, a post-operative complication , etc. analyzed. If such a deterioration in health is detected, a medical service (doctor, emergency service, etc.) is automatically notified. In individual cases, feedback is sometimes provided to the patient. Such monitoring systems are so far aimed at detecting a given disease and are used to monitor already ill individuals. On the other hand, such monitoring systems are not designed for preventive health care.

Of the Invention is based on the object, in particular a patient side specify a procedure that is easy and flexible to handle, which is a particularly effective health care, i. preventive health monitoring, allows. The invention is further based on the object, one for carrying out the Specify method particularly suitable device.

Regarding the Method, this object is achieved by the features of the claim 1. Regarding the one to carry The apparatus according to the invention is achieved by the device formed by the method the features of claim 9.

According to the invention is at least a sensor, but preferably a plurality of sensors provided. The or each sensor is hereby at or in the vicinity of the patient in the course of the procedure (one each) for health Relevant parameter of the assigned patient. The or each characteristic is fed to an evaluation unit. The evaluation unit analyzes the or each characteristic to whether it is more critical on the basis of deposited decision criteria Condition exists. As a "critical" here is general every condition viewed for according to the filed Decision criteria a particular action is provided. In The decision criteria here are a number of specific states as critically defined, which preclude a healthy lifestyle.

If a critical condition is detected in accordance with the decision criteria, the evaluation unit issues a recommendation to the patient. The recommendation contains, in particular, a suggestion for the patient to change his lifestyle which is suitable for remedying the recognized critical condition. To transfer the Characteristic of the or each sensor to the evaluation and for retransmission and output of the recommendation to the patient, the device further comprises suitable transmission means, which are based in particular on mobile phone and / or Internet technology. Conveniently, in particular the integration of near-patient components of the device in a mobile phone and / or a web-enabled computer is provided. Optionally, it is further provided that when a critical condition is detected, a message is generated to a medical expert.

Of the Term "patient" stands for any individual Person, independent their state of health, and thus includes in particular the healthy People. The "patient" may be further also to act a healthy or diseased animal.

As a result continuous and automatic observation of the patient first to recognize unhealthy lifestyles very effectively and inexpensively, in particular even if they are not at all aware of the patient. As a result of the customized feedback to the patient receives this information about the for him relevant measures to improve his lifestyle without him having this information in time consuming Way out of the available self-procure and select sources. By in case of need also repeated transmission of recommendations to the patient the device as a "constant reminders" also indicates that the patient over the measures needed to improve their lifestyle Not only once informed, but also beyond its implementation remembered as often as necessary.

When Characteristics are for a convenient vital parameters such as e.g. Blood pressure, body temperature, Heart rate, the concentration of certain substances in the organism or its excretions. But unhealthy ways of life preferably early, especially before they cause a significant burden on the human Lead organism, to be able to recognize is in a preferred embodiment the invention additionally or as an alternative to vital measured variables, at least one parameter is detected, the one body position, a body movement and / or an environmental influence, and thus only indirectly is relevant to health.

When the one body position corresponding characteristic is in particular the geographical position of the body, i. the place where the patient stays, detected. Furthermore, optionally the relative position of certain Body Parts recorded, from which in the evaluation unit arithmetically on the attitude and / or Movement of the patient is inferred. Alternatively or in addition Optionally, at least one parameter is determined which is immediate characterized a movement of the patient. For example optionally provided by a computer keyboard associated Pressure sensor the strength and / or frequency of the key operation to determine the patient as a parameter. Latter Characteristics are especially for monitoring the stress state of the patient used in computer work. As the parameter to be assigned to an environmental influence, in particular the outside temperature, the noise effect (or ambient volume), the solar radiation, etc. recorded.

ever according to the type to be detected characteristic is the associated sensor optionally in the body implanted in or worn by the patient, in particular as part of a wristwatch, glasses or clothing. One or several sensors can continue in a working device (e.g., those mentioned above Computer keyboard), a sports device and / or a piece of furniture be integrated. The term furnishing includes here next to furniture also technical devices, in particular self-diagnostic devices such as. a personal scale, a mobile sphygmomanometer, etc.

The of each sensor detected characteristics are preferably in the evaluation unit as a function of time (hereinafter as a characteristic function ), so that the course of each characteristic over a sufficient past Time span available stands.

The Device preferably comprises a patient-related component (patient unit), which is directly associated with the individual patient. The Patient unit includes the sensors associated with the patient and suitable means for data transmission, Data input and data output. The latter funds will be in particular provided by a mobile phone or a PDA (handheld computer), in which parts of the device according to the invention preferably integrated are.

Conversely, the evaluation unit is expediently arranged centrally and therefore in particular remote from the patient. In this case, the central evaluation unit cooperates in particular with a plurality of patient units and serves for the evaluation and optionally recording of the characteristics of several patients. On the one hand, monitoring takes place on a patient-specific basis, ie a separate set of relevant decision-making criteria is kept available for each patient in the evaluation unit. On the other hand, the common archiving of the parameters of a large number of patients and their comparative evaluation allows a simple efficient and efficient extraction of normative values or findings on conspicuous social phenomena, etc.

In the simplest form of evaluation performed by the evaluation unit becomes the current value of a characteristic with one in the context of the decision criteria stored threshold value, whereby a "critical state" is then detected, if the parameter exceeds the threshold value or falls below. additionally or alternatively, in the course of evaluation of the parameters or characteristic features One or more methods used in information technology as "Data Mining and data warehousing.

in the Course of the "Data Mining "will be the recorded characteristic functions the presence of hidden trends and correlations analyzed. In particular, a trend that continues over a longer period of time is becoming a trend Increase or decrease of a characteristic, a (unusual for ordinary living conditions) Constancy or irregularity a characteristic (e.g. Respiratory arrest or cardiac arrhythmia) or a sig nificant periodicity referred to in the course of a characteristic. As a correlation in particular, a coherent, i.e. a certain temporal relationship underlying occurrence of certain patterns in different characteristic functions called. such Trends and correlations become especially in the course of the "Data Mining" without underlying assumptions about the presence and the characteristics of such trends and correlations searched. this makes possible especially the finding of hidden trends and correlations, their existence a priori neither the patient nor the procedure caring medical experts. Will such a trend and / or such a correlation found, this is based on the decision criteria for a possible health relevance. A a periodic rhythm of a parameter becomes a 24-hour rhythm for example, the natural one Assigned daily rhythm and classified as non-health-relevant, while Again, for example, a significant with a particular geographical Location correlated change the heart rate as possibly would be classified as relevant to health. Will a health-related trend or a health-related Correlation is recognized, then in accordance with the decision criteria of Patient warned. Alternatively or additionally, the deposited Decision criteria on the basis of the recognized trend or recognized Correlation automatically or mediated by the patient or medical experts adapted to the trend or the Correlation is counteracted.

For example is determined in the course of the procedure as a parameter Ambient noise of the patient in the context of "Data Mining" to peak values examined. Be repeated comparable peak values found, these are correlated with other recorded parameters. Here, for example, a correlation between the noise peak and a geographical position of the patient. For example is automatically detected in this way by the evaluation unit, that the patient during the lunch break is a shortcut through a very noisy machine shop takes to get into the faster Canteen and this one possibly in the long term harmful noise pollution exposes. This realization becomes, for example, adaptation the decision criteria implemented that the patient from now on Receiving recommendation as soon as he realized in the vicinity of the noise source (Machine hall) is coming.

in the Course of the "Data Warehousing " so-called one or more parameters Meta-data generated, which compared to the parameters increased information density or efficient processing, management or presentation of the Support characteristics. The Meta-data include in particular statistical information on the parameters, as well as Indices and other navigational aids.

For example becomes a geographical location of the patient's position (e.g., a GPS signal) in the course of the "Data Warehousing "in transferred an XML representation that easy with a common one Display program can be displayed and by means of which the traveled Path of the patient is graphically represented. In the course of "Data Warehousing" will be expediently also characteristics of different Patients combined, e.g. statistical cross-patient statistical averages for one specific characteristic type receive.

following is an embodiment of Invention explained in more detail with reference to a drawing. In it shows the only one Figure a device for preventive Health monitoring of a patient with a number of patients assigned Sensors for recording health relevant parameters and to send a recommendation upon detection of a decision criterion set according to the requirements defined critical condition.

The schematically illustrated device 1 includes a number of patient units 2 (from which for the sake of simplicity in the figure, only one is shown) and an evaluation unit 3 ,

The patient unit 2 includes a number of sensors 4a - 4d attached to a patient 5 or are arranged in the vicinity thereof and serve for detecting in each case a health-relevant parameter G _i (i = 1, 2, 3, 4). The sensors 4a - 4d are partly in the body of the patient 5 implanted, partly on the body of the patient 5 portable, partly in work and sports equipment or other furnishings of the patient 5 integrated. On the one hand, the recorded parameters G _i are vital measured variables of the patient 5 , such as blood pressure, heart rate, adrenaline and / or insulin concentration in the blood, etc., on the other hand by parameters G _i , the geographic position of the patient 5 or a body part position or body movement, or by characteristics G _i that are on the patient 5 reflect environmental influences (eg solar radiation, ambient temperature, pollution of ambient air, etc.).

For example, the sensor 4a in a patient's glasses 5 integrates and records the on the patient 5 acting solar radiation as parameter G ₁ . The sensor 4b is for example in the body of the patient 5 implants and records the heart rate as characteristic G ₂ . The sensor 4c is for example in a wristwatch of the patient 5 integrates and detects a GPS signal as parameter G ₃ . The sensor 4d is for example in a personal scale of the patient 5 integrates and records the body weight of the patient 5 as parameter G ₄ as soon as it uses the personal scale. Additionally or alternatively, a variety of other and / or different types of sensors can be used, which are not explicitly listed in the figure. In particular, different patients may be assigned different configurations of sensors.

The patient unit 2 further comprises a sensor interface 6 , Input / output means 7 for data such as display, alphanumeric keyboard, etc. as well as a transceiver unit 8th for communication with the evaluation unit 3 ,

The sensors 4a - 4d in this case transmit the respectively detected parameter G _i continuously or at discrete time intervals to the sensor interface 6 , Every sensor 4a - 4d is here via a wired or wireless transmission link with the sensor interface 6 connected. The wireless transmission of the parameters G _i takes place here in particular using Bluetooth, infrared data transmission or other wireless transmission technology. The sensor interface 6 transmits the parameters G _i to the transmitting / receiving unit 8th , the characteristic G _i to a corresponding transmitting / receiving unit 9 the evaluation unit 3 forwards. If necessary, the parameters G _i in the sensor interface 6 before forwarding to the evaluation unit 3 cached.

The data transfer between the patient unit 2 and the evaluation unit 3 takes place by means of a secure, ie access-protected, Internet or mobile telephone transmission technology. The sensor interface 6 , the input / output means 7 and the transceiver unit 8th are expediently integrated in a mobile telephone, a PDA (Personal Digital Assistant, ie handheld computer), a personal computer or other Internet-enabled computer system.

Within the evaluation unit 3 are those of the send / receive unit 9 received parameters G _i on the one hand a decision module 10 , on the other hand, a patient database 11 fed. In the patient database 11 the received characteristics G _{i are} time-dependent archived, ie there is a characteristic function F _i = G _i (t) of the time t in the patient database 11 deposited, which reflects the time course of the associated characteristic G _i . The decision module 10 is adapted to that of the transmitting / receiving unit 9 received parameters G _i then analyze whether a critical condition exists. Such critical states are within the evaluation unit 3 in the form of decision criteria R defined in the patient database 11 are patient-specific deposited and the decision module 10 from the patient database 11 be supplied. Represents the decision module 10 Based on the decision criteria R with respect to one or more parameters G _i fixed a critical state, it generates a recommendation W in accordance with the decision criteria R and transmits them via the transmitting / receiving unit 9 to the patient unit 2 , There is the recommendation W on the input / output means 7 issued to the patient. As recommendation W, in addition to acute warnings, suggestions are also issued that aim to improve the patient's lifestyle in the medium and long term 5 each strive. The decision module 10 optionally initiates a message N to an expert interface upon detection of a critical condition 12 who is a medical expert 13 , in particular a doctor, for example, the family doctor of the patient 5 , an emergency service, etc. is assigned. The expert interface 15 In particular, in turn, includes input and output means for data, such as screen, keyboard, etc. and is in particular in a working computer of the expert 13 integrated. The message N is in particular in the form of an e-mail or the like. transmitted. The decision module 10 decides case by case depending on the type of critical condition detected and on the basis of this case deposited decision criteria R, whether both a recommendation W to the patient 5 and a message N to the experts 13 should be generated or if only the recommendation W to the patient 5 should be sent. Optionally, the decision criteria R also include those states in which only one message N to the expert 13 is produced.

Furthermore, the evaluation unit comprises 3 a data mining module 14 , This is designed to be in the patient database 11 stored characteristic functions F _i for hidden trends T and correlations C to analyze. As a result of a recognized trend T or a recognized correlation C, the decision criteria R are automatically determined by the data mining module 14 customized. Alternatively or additionally, it is provided that recognized trends T and correlations C to the expert 13 in the form of a through the decision module 10 generated message N or the patient 5 be reported in the form of a recommendation W.

The evaluation unit 3 also includes a data warehousing module 15 , The data warehousing module 15 is intended to derive from the characteristic functions F _i so-called metadata M and to their archiving to a metadata memory 16 to convey. As metadata M data are generally referred to, which are related to the stored characteristic functions F _i , but compared to the latter have a higher information level. Such data include in particular statistical data, as well as data that allow faster access to the characteristic functions F _i or an improved representation of the latter. In particular, average values relating to specific parameters G _i , which were determined by cross-patient statistics, are also stored as metadata M. On the in the metadata store 16 stored metadata M have in particular the decision module 10 and the data mining module 14 Access. The metadata store 16 is optional in the patient database 11 integrated.

In the patient database 11 is also to every patient 5 , if applicable, to any authorized expert 13 a user profile P deposited. The user profile P contains individual settings as well as personal details. In the case of a patient 5 Here are in particular the type and number of the user 5 associated sensors and the nature of the detected characteristics G _i registered. Furthermore, the user profile P includes a list for the individual patient 5 relevant decision criteria R. The user 5 can this user profile P on the input / output means 7 In particular, also edit user-specific by sending appropriate instructions A to the patient database 11 emits. To avoid improper access to patient-related data is in the access path of the patient 5 on the patient database 11 an authentication module 17 switched, at which the patient 5 , identified by specifying a user name and password, for example.

The user 5 By editing the user profile P, it is thus possible to determine which situations should be classified as critical for him, that is to say under which situations he is to be classified by the evaluation unit 3 wants to be warned. Optionally, it is provided that the patient 5 by editing the user profile P can also choose under what conditions the expert 13 to be notified.

The expert 13 in turn has particular access to the decision criteria R and may modify or supplement them by appropriate instructions A 'if necessary. The access of the expert 13 on the patient database 11 Here again is an authentication module 17 interposed.

In the example shown, it is with respect to the user 5 For example, a decision rule R deposited, according to which of the decision module 10 the patient 5 warns when the sun exposure, the patient 5 is exposed exceeds a certain value. The decision module 10 monitors the parameter G ₁ and its stored characteristic function F ₁ . Similarly, by monitoring the characteristic G ₄ and the associated characteristic function F _4, the body weight of the patient 5 supervised. The patient 5 this is warned when its weight exceeds a designated threshold or through the data mining module 14 a creeping increase has been detected in a past period of time. In connection with the recommendation W be the patient 5 It submits proposals for a customized diet, which is the decision module 10 Based on the deposited decision criteria R and others, in the patient database 11 deposited medical background information I created. At the same time, all parameters G _{i are} generated by the data mining module 14 examined for hidden correlations C. In this way, particular circumstances are determined, among which the heart rate of the patient 5 takes on critical values. The data mining module 14 develops automatically adapted rules R, on the basis of which the patient 5 automatically be warned when a situation reoccurs.

A use case not described in detail of the device described above is in particular in the detection of poor posture. This will be particular detected by using sensors that measure the relative position of body parts of the patient 5 capture each other as parameters G _i . In the evaluation unit 3 becomes the body posture of the patient 5 recalculated. In an unhealthy posture (eg at a computer workstation) the patient becomes 5 For example, automatically warned to prevent long-term damage to the cervical spine.

Likewise, optional, the typing speed and the button pressure during the work of the patient 5 on the computer by suitable sensors, such as pressure sensors on the computer keyboard to monitor. The evaluation unit is determined by the keypress and the typing behavior 3 an unhealthy state of stress of the patient 5 detected, as it typically occurs in excessive computer use. The evaluation unit 3 sends upon detection of such a critical state, a corresponding warning signal on the computer screen of the patient 5 ,

Furthermore, the device is optionally provided 1 to use in the course of cancer screening. This is done by means of a in a vein of the patient 5 implanted blood sensor blood counts of the patient as measured variables. A critical state is in this case of the evaluation 3 detected when the concentration of platelets is found to be in the range of 150,000 to 350,000 cells per microliter of blood. Upon detection of abnormal blood levels, in the case described by the decision module 10 as part of Recommendation W, the proposal to the patient 5 issued to seek a doctor for the purpose of a check-up.

Yet another application of the device 1 It consists in recognizing and treating unconscious sleep disorders (parasomnias), snoring, etc. Such conditions are caused by the patient 5 often not consciously perceived and therefore not readily recognized. The detection of parasomnia by means of the device 1 is done in particular by sensors that detect the heart rate, breathing, the ambient volume. The recommendation W is here preferably on the one hand as a written message that the patient 5 after waking up, as well as in the form of a wake-up signal that receives the acute symptoms of a sleep disorder by waking the patient 5 completed.

Claims (16)

Method for preventive health monitoring of a patient ( 5 ), in which by means of at least one on or in the vicinity of the patient ( 5 ) arranged sensor ( 4a - 4d ) a health-relevant parameter (G _i ; i = 1 - 4) and an evaluation unit ( 3 ), in which the or each parameter (G _i ) in the evaluation unit ( 3 ) is checked for the presence of a critical state in accordance with stored decision criteria (R), and in which, when a critical state is detected by the evaluation unit ( 3 ) a recommendation (W) to the patient ( 5 ) is output. A method according to claim 1, characterized in that as a parameter (G _i ) a body position, a body movement and / or an environmental influence is detected. A method according to claim 1 or 2, characterized in that the time course of at least one parameter (G _i ) as a characteristic function (F _i ; i = 1 - 4) of the time (t) by means of the evaluation unit ( 3 ) is recorded. A method according to claim 3, characterized in that at least one characteristic function (F _i ) by means of a data mining module ( 14 ) is examined for trends (T), whereby upon detection of a trend (T) classified as critical according to the decision criteria (R) the decision criteria (R) are adjusted to avert the trend (T), and / or a recommendation (W) to the patient ( 5 ) is output. Method according to one of claims 1 to 4, characterized in that at least two characteristic functions (F _i ) by means of a data mining module ( 14 ) are examined for correlations (C), wherein upon recognition of a decision criterion (R) as a critically classified correlation (C) the decision criteria (R) are adjusted to avert the correlation (C), and / or a recommendation (W ) to the patient ( 5 ) is output. Method according to one of claims 1 to 5, characterized in that from the or each characteristic function (F _i ) metadata (M) are derived, which have a relation to the or each underlying characteristic function (F _i ) increased information density, access to at least one Characteristic function (F _i ) facilitate and / or facilitate the representation of the characteristic variable function (F _i ) associated characteristic (G _i ). A method according to claim 6, characterized in that the metadata (M) from parameters (G _i ) of several patients ( 5 ) be derived. Method according to one of claims 1 to 7, characterized in that upon recognition of a critical condition in accordance with the decision criteria (R) a recommendation (W) to a medical expert ( 13 ) is delivered. Contraption ( 1 ) for the preventive health monitoring of a patient ( 5 ), - with at least one sensor ( 4a - 4d ) for the acquisition of a health-relevant parameter (G _i ), - with a decision module ( 10 ) evaluation unit ( 3 ), where the decision module ( 10 ) is designed to examine the or each parameter (G _i ) in accordance with stored decision criteria (R) for the presence of a critical condition and to generate a recommendation (W) when a critical condition is detected, and - with data transmission means ( 6 . 7 . 8th . 9 ) for transmitting the characteristic (G _i ) from the or each sensor ( 4a - 4d ) to the evaluation unit ( 3 ) and for the retransmission and issuing of the recommendation (W) to the patient ( 5 ). Contraption ( 1 ) according to claim 9, characterized in that the or at least one sensor ( 4a - 4d ) is designed to detect a body position, a body movement or an environmental influence as a parameter (G _i ). Contraption ( 1 ) according to claim 9 or 10, characterized in that the or at least one sensor ( 4a - 4d ) in the body of the patient ( 5 ) or on the body of the patient ( 5 ) is portable. Contraption ( 1 ) according to claim 9 or 10, characterized in that the or at least one sensor ( 4a - 4d ) is integrated in a working device, sports equipment or furnishing Contraption ( 1 ) according to one of claims 9 to 12, characterized in that the evaluation unit ( 3 ) a data mining module ( 14 ), which is designed to recognize trends (T) and correlations (C) from one or more characteristic quantities (F _i ) representing the time profile of a respective parameter (G _i ) and in accordance with the decision criteria (R) with regard to to classify their health relevance as critical or uncritical. Contraption ( 1 ) according to one of claims 9 to 13, characterized in that the evaluation unit ( 3 ) a data warehousing module ( 15 ), which is designed to derive metadata (M) from the or each characteristic function (F _i ) which have an increased information density compared with the or each characteristic parameter function (F _i ), access to at least one characteristic function (F _i ) facilitate and / or facilitate the representation of a characteristic variable function (F _i ) associated characteristic (G _i ). Contraption ( 1 ) according to one of claims 9 to 14, characterized in that the evaluation unit ( 3 ) remote from the patient and centrally implemented and for the patient-specific processing of parameters (G _i ) of several patients ( 5 ) is trained. Contraption ( 1 ) according to one of claims 9 to 15, characterized in that the transmission means ( 6 . 7 . 8th . 9 ) comprise a mobile telephone connection or internet connection.