Field of the Invention

This invention relates to method and apparatus for analyzing information gathered from symmetric areas of a living organism and, more particularly, to an innovative analysis technique which provides for early and reliable detection of abnormal bodily conditions.

Background of the Invention

The early detection of abnormal bodily conditions, in¬ dicative of possible disease, has long been a goal of the medical community. Indeed, it is well known that early detection of certain diseases, such as breast cancer, substantially increases the possibility of survival, while late detection of the disease often results in the untimely death of the patient.

There are, of course, numerous methods, techniques and procedures that have been developed in an effort to detect abnormal bodily conditions that may lead to disease. One such method is to gather data generated by the body, such as temperature, electrical activity and the like, and to analyze that information in an effort to detect abnor¬ malities. For example, it has been long known that a detectable electrical potential is generated by the action of the heart and it is common medical practice to monitor such potential levels. Electrodes are fastened to the skin of the patient and the voltages generated are recorded on a continuous roll of graph paper by a galvanometer stylus, which deflects in proportion to the voltage signal. The resultant electrocardiograph must then be analyzed by a trained observer to detect abnormalities, if any, in the action of the heart.

Although electrocardiograms are extremely useful in monitoring heart activity, they generally provide only a "snapshot" of such activity, which snapshot must be analyzed

and compared with previous electrocardiograms in order to give an indication of the condition of the heart. Electro¬ cardiograms, however, have the disadvantage of not providing data over an extended period of time, or under ambulatory conditions.

In order to overcome this disadvantage, the Holter twenty-four hour ECG monitor was developed. This device consists of a plurality of electrodes which are fastened to the skin, and a portable data collection system, which is connected to the electrodes and worn by the patient on a belt, or attached to the patient by other means. The patient wears the device for twenty-four or forty-eight hours, and the data collection system records all heart activity during that interval under ambulatory conditions. The data collection system is subsequently returned to a doctor's office, where the data is used to generate an electrocardiogram for analysis by the doctor.

Other systems exist in which electrical potential, or resistance generated by the human body is measured over pre¬ determined intervals, either under ambulatory, or static, conditions, with subsequent data analysis in an attempt to detect body abnormalities. Such systems are described, for example, in U.S. Patent Nos. 3,920,002, 3,971,366 and 4,328,809. In such known systems, however, in which electrical phenomena are monitored, monitoring of informa¬ tion does not occur over an extended period of time, or under ambulatory conditions. Also, such known systems require a trained observer to interpret the data obtained in order to render a diagnosis, which diagnosis may take hours, or days, depending on availability of the trained observer.

Another body parameter that is often monitored in an attempt to detect abnormal bodily conditions is body temperature, as it has been known since antiquity that an increase in body temperature often forecasts, or accom¬ panies, disease conditions. One particular area in which body temperature has been used in an attempt to detect disease is the earl -"detection of breast cancer. Clinical

studies have conclusively established that the great majority of malignant mammary tumors act as localized heat sources. The temperature of a breast effected by a malignant tumor remains elevated, while a normal breast fluctuates through a twenty-four hour temperature cycle. The normal twenty-four hour temperature cycle comprises a relatively lower temperature during the day and a higher temperature at night. The temperature difference between a normal breast and that of a breast containing a malignant tumor provides the basis for the diagnosis of breast cancer by known thermographic techniques.

Various systems exist in which the difference in tempera¬ ture between a normal breast and a breast containing a malignant tumor is used in an attempt to detect breast cancer at an early stage of the disease. For example, in U.S. Patent No. 3,960,138, there is provided a method and apparatus for detecting the presence of breast cancer in which the breast receiving cups of a brassiere are each provided with thermally conductive material next to the skin, with a thermistor attached to the thermally conductive material in each cup. The thermistors are connected to the adjacent arms of a Wheatstone bridge. In the absence of a tumor, both breasts remain substantially the same tempera¬ ture, and the Wheatstone bridge remains balanced. If a tumor is present in one breast, the higher temperature in the diseased breast unbalances the Wheatstone bridge and provides an indication of an existing abnormality.

Another such system is described in U.S. Patent No. 4,055,166. In this system a brassiere is provided, which includes a number of skin temperature sensors. The sensors are connected to battery operated integrated circuits, including storage registers, with all such circuits, including the battery, being integral with the brassiere. The breast temperature at predetermined intervals is recorded and the recorded temperature is subsequently printed for examination by an operator. The brassiere is designed to be worn normally while skin temperature monitor-

ing is performed. This system is specifically designed to collect temperature data over an extended period of time under ambulatory conditions.

Another approach at monitoring body temperature is described in U.S. Patent No. 4,310,003. In that system, a scanning array unit is provided consisting of a matrix of sensitive sensing elements such as thermopile devices, which are mounted in close but spaced relation to a body to produce an output signal proportional to the temperature of the body. In addition, a reference temperature sensor is provided, whose output is compared with the output of the .temperature sensing array. The sensor array unit is aligned with a desired first portion of an individual patient, and the multiplicity of the sensors are read and the data stored. The same sensor array unit is then accurately aligned with - a second body portion and similarly read. The stored signals are then processed in a pattern recognition program to directly provide an automated diagnosis based on the temperature data.

An additional temperature monitoring system is described in U.S. Patent No. 4,190,058. In this patent, there is disclosed a device for the early detection of breast cancer, the device comprising a flexible heat conductive web, preferably in the form of a disc shaped patch, having an - adhesive layer on one side thereof, and a peelable layer removably secured thereto by said adhesive layer. On the other side the device comprises an array of spaced apart indicators each of which comprise a dye or a pigment in a temperature sensitive substance, which melts at a relatively precise temperature. As many indicators are used as are necessary to cover the desired temperature range. The device is- incorporated into the breast receiving cups of a brassiere, mirror image quadrants of the two breasts are scanned, and the device is visually examined to determine the number of indicators which have displayed a change in color, thus ^apprising the doctor of the existence of potential abnormality in the mammary tissue. From the

foregoing, it is apparent that many systems and devices exist in the prior art, all of which attempt to utilize data gathered from the body for the early detection of disease. All known systems, however, suffer from one or more inherent deficiencies. Such deficiencies include, for example, the inability to monitor bodily conditions over extended periods of time and under ambulatory conditions, the lack of a rapid and automated analysis of the collected data, the need to provide trained personnel to interpret the data gathered and prepare a report for the patient, and the inability to accurately and consistently detect abnormal conditions which may lead to serious disease. It is apparent that if a system existed for the analysis of information gathered from the human body, which system did not suffer from the foregoing deficiencies, a marked advance in the early detection of certain diseases would be possible.

It is, therefore, an object of the instant invention to provide method and apparatus to analyze information gathered from a living organism in order to detect body abnor¬ malities.

It is a further object of the instant invention to provide for the automatic analysis of such information.

It is a still further object of the instant invention to perform a fully automated analysis of information gathered from a living organism, without requiring the intervention of a trained observer.

It is another object of the instant invention to gather information for analysis over an extended period of time and preferably under ambulatory conditions.

It is an additional, preferred and general object of the instant invention to provide an analysis of information gathered from a living organism, which analysis will accurately and reliably detect abnormal bodily conditions which may lead to disease.

Summary of the Invention

In accordance with various embodiments of the instant invention, apparatus and methods have been developed for de¬ termining abnormalities in a living organism, which abnor¬ malities may be indicative of disease.

A data collection device is provided which gathers and records data from symmetric areas of a living organism over a predetermined interval. The recorded data is thereafter transmitted from a first location, to a data analysis center at a second location.

It is a feature of the invention that the data is analyzed at the analysis center by performing ipsilateral and contralateral comparisons, and a chronobiologic analysis on selected portions of the recorded data.

It is a further feature of the invention that the analysis center automatically generates a patient report, indicative of the absence, or presence, of disease in the living organism, and transmits the report from the analysis center to the first location.

In accordance with a first embodiment of the instant invention, quasi-continuous recording of skin temperature under ambulatory, or in-bed conditions, is conducted at N symmetrical locations of the human body.

It is a feature of the invention that reading time can be up to several days with a predetermined measuring frequency.

It is another feature of the invention that the tempera¬ ture data gathered is analyzed by performing N/2 con¬ tralateral comparisons (opposite and symmetrical body locations), and 2L ipsilateral comparisons (same body location) , where L is the number of significant comparisons per body site, with the total number of comparisons equal to N/2 + 2L.

^" It is a still further feature of the invention that interdependence coefficients and attenuation coefficients are calculated from the temperature data, and a chrono¬ biologic analysis is performed based on the calculated

coefficients, which analysis provides information indicative of the absence, or presence, of potential bodily disease.

In accordance with a second embodiment of the instant invention, the inventive apparatus and method is utilized for the early detection of breast cancer. Temperature sensors are placed on predetermined areas of the breast, one temperature sensor being located on each breast quadrant and one additional temperature sensor being placed in the nipple/areola area of each breast. Temperature data from the temperature sensors is recorded every five minutes over a twenty-four hour interval under ambulatory conditions and this temperature data is stored in a portable data collec¬ tion device. The data collection device is lightweight, unobtrusive, and designed to be worn around the waist of a patient.

It is a further feature of the invention that the data collection device is removed from the patient at a doctor's office or similar location, and the recorded temperature data is transmitted over a communications link to a remote data Analysis Center.

It is another feature of the invention that the tempera¬ ture data is analyzed at the Analysis Center by performing ipsilateral and contralateral comparisons of the temperature data, along with a chronobiologic analysis of the data.

Ipsilateral comparisons consist of comparing the tempera¬ ture data from predetermined sensors on the same breast, while contralateral comparisons compare the temperatures from predetermined sensors on opposite breasts. It is a feature of the invention that 16 of 20 possible ipsilateral comparisons, and 5 of 25 contralateral comparisons have been found to be significant and non-redundant for analysis purposes.

It is a still further feature of the invention that a plurality of regression coefficients are calculated, based on the ipsilateral and contralateral temperature com¬ parisons, which regression coefficients are characterized by deviations in temperature values when compared with the

straight regression line that would represent equal tempera¬ tures at any test location at any time. The higher the value of the regression coefficient, the greater the thermal levels of the two comparison breast areas could be expected to differ.

It is another feature of the invention that a chrono¬ biologic analysis is performed on the recorded temperature data, which chronobiologic analysis concerns the time patterns of temperature of each breast area, with the time patterns of temperature of each of all other breast areas explored. Relative attenuation coefficients are calculated based on the chronobiologic analysis, which coefficients are characterized by the mean degree in amplitude of the temperature fluctuations between predetermined breast areas. A plurality of relative attenuation coefficients are calculated as part of the analysis process.

It is still another feature of the invention that a score and a chronothermodynamic class is assigned to each specific breast depending upon the calculated values of the regres¬ sion coefficients and the relative attenuation coefficients. The chronothermodynamic class represents, in an index of I- IV, an increasing degree of abnormality.

It is a still further feature of the invention that the entire analysis performed at the analysis center is fully automatic, including the printing of an examination report, which is then transmitted to the attending physician.

It is an advantage of the invention that the entire data analysis and evaluation process takes less than one-half hour, and does not require intervention by an operator to examine the data, or the prepared report, prior to transmis¬ sion to the physician.

It is a still further feature and advantage of the invention that the data analysis performed results in accurate and reliable predictions of the possibility of breast disease.

In accordance with another embodiment of the instant invention, the inventive method and apparatus can be

utilized for the chronothermodynamic examination of the upper extremities for the detection of certain vascular diseases.

It is another feature of the invention that in accordance with a chronothermodynamic examination of the upper ex¬ tremities for the determination of certain vascular disease, quasi-continuous recording of skin temperature in ambulatory or in-bed conditions is made at symmetrical locations of the human body.

It is a still further feature of the invention that N thermal sensors are utilized and temperatures are recorded for a time interval up to several days, the temperature response being ' ^• measured in accordance with various stress tests applied to the area of the body being monitored.

It is another feature of the invention that temperature records are maintained prior to a stress test, during a stress test, and after a stress test, wherein the stress applied to the human body can include temperature extremes, chemical agents, radioactivity, and/or other phenomena.

It is a still further feature of the invention that characteristic parameters of each temperature analysis prior to, during and after a stress test are recorded, and chronobiologic and/or ipsilateral comparisons are conducted on the data extracted from symmetrical areas of the body.

It is a still further feature of the invention that thereafter a chronobiologic analysis is performed, which analysis permits the detection of the possibility of vascular disease in the area being monitored.

It is a still further feature of the invention that the inventive analysis technique is not limited to the evalua¬ tion of temperature, but is advantageously capable of analyzing other parameters recorded on the body such as electrical potential, electrical resistance and the like.

The foregoing and other objects, features and advantages of this invention will be fully understood from the follow¬ ing description or an illustrative embodiment thereof, taken in conjunction with the accompanying drawings.

Brief Description of the Drawings

In the Drawings:

FIG. 1 illustrates utilization of the inventive method and apparatus for gathering and recording temperature data from the human breast for the detection of breast disease;

FIG. ^{" *} 2 illustrates use of the inventive method and apparatus to gather and record temperature data from fingertips for the detection of vascular disease;

FIG. 3 illustrates, in block diagram form, the apparatus of the instant invention;

FIG. 4 illustrates, in block diagram form, the components of a data collection device useful with the instant inven¬ tion;

FIG. 5 illustrates a generalized flow chart of one embodiment of the invention.

FIG. 6 illustrates preferred locations of temperature sensors on the left and right breasts;

FIG. 7 illustrates contralateral comparisons between the left and right breasts;

FIG. έ illustrates ipsilateral comparisons from the left breast;

FIG. 9 illustrates the contralateral and ipsilateral comparisons found to be significant and non-redundant in accordance with the analysis method of the instant inven¬ tion;

FIG. 10 is a flow chart of the analysis method for use with the instant invention when applied to the detection of breast disease;

FIG. ΪIA - HE illustrates temperature data in graph form taken from an actual patient examination;

FIG. 12 illustrates calculated values of the regression coefficient, attenuation coefficient and chronothermodynamic score for a typical breast examination;

FIG. 13 represents the chronothermodynamic class for a typical breast examination;

FIG. 14 illustrates typical temperature data taken from a

-•¥ patient over a 24 hour period;

FIGS. 15A-B - 18A-B illustrate the manner in which regression coefficients are determined and applied;

FIGS. 19-20A-B - 21A-B, illustrate the manner in which attenuation coefficients are determined and applied;

FIGS. 22A-D illustrates clinical results obtained through use of the instant invention for healthy breasts, breasts with benign disease, breasts with borderline disease and breasts having cancer; and

FIG. 23 is a generalized flow chart for the analysis method of the instant invention, when generally applied to vascular studies; and

FIG. 24 is a more specific description of a vascular embodiment of the instant invention.

Detailed Description of the Invention

The instant invention is designed to detect body abnor¬ malities which may lead to disease by recording prede¬ termined body parameters, such as skin temperature, electrical potential, electrical resistance and the like, under non-environmentally controlled/ambulatory and/or in- bed conditions, over variable periods of time ranging from less than one hour to several days, using a portable, multi¬ channel microprocessor based, data acquisition system. After the body parameters are recorded and stored, the data is transferred from the portable multi-channel microproces¬ sor based data acquisition device, to a main frame data processing system, at which a computerized analysis is performed based on the data received. Specific algorithms analyze the data and permit automatic and rapid data analysis, including the delivery of an examination report. Exemplary applications of the invention may include the chronothermodynamic examination of the breast for the detection of breast cancer, and the chronothermodynamic examination of the upper extremities for the detection of various vascular diseases such as Raynaud's syndrome of various etiologies, acrosyndromes resulting from the

professional use of vibrating tools, carpal tunnel syn¬ dromes, algodystrophia and rheumatoid arthritis.

It is also possible, through use of the instant inven¬ tion, to detect vascular diseases of the lower extremities, to assist in repair hand surgery after finger, reimplantation or toe transfer, and to possibly detect and/or predict migraines or other headaches based on temperature variations in the body.

In addition to the body temperature analysis described above, the basic method and apparatus described in conjunc¬ tion with the instant invention can also be utilized to collect and analyze electrical potential generated by various portions of the body, and/or varying electrical resistance of various body portions. In short, whether the incoming data collected is temperature data, voltage potential, electrical resistance, or other parameters gathered from a living organism, the innovative and novel analysis method described and claimed hereafter is the same, which method is able to reliably and accurately detect abnormalities which may be indicative of disease.

Referring now to FIG. 1, there is illustrated use of the instant invention in an application for monitoring breast temperature for the early detection of breast cancer. As indicated, temperature sensors are placed at predetermined areas of the breast (described below), and temperature data is collected in a data collection device over an extended period of time. The data collection device is designed to be portable and light weight such that the patient can easily and comfortably wear the device on a belt during data collection. The temperature sensors are also designed to be light weight and are affixed to the breast with special self-adhesive, non-allergenic pads to avoid irritation to the breast tissue.

FIG. 2 illustrates a similar arrangement for monitoring the temperature at the finger tips of each hand over a predetermined interval. The amount of blood flow in the fingertip area is often indicative of vascular disease of

the upper extremities and the amount of blood flow is directly related to temperature at the fingertips. Accord¬ ingly, the arrangement illustrated in FIG. 2 monitors fingertip temperature data over predetermined intervals, which temperature data is stored in a data collection device for subsequent processing and analysis.

It is to be understood that the uses illustrated in FIGS. 1 and 2 are exemplary only and are not meant to limit the scope or spirit of the invention.

Referring now to FIG. 3, there is illustrated a preferred embodiment for use of the instant invention. More par¬ ticularly a plurality of sensors 1 are attached to a living organism and transfer data to a portable data collection device 2 , which data collection device, as described above, may be worn by the patient under ambulatory conditions for an extended period of time. After collection of the data over the predetermined interval, the data which has been stored in data collection device 2 may be transferred to a personal computer 3, which includes keyboard 4 and a modem 5. It is anticipated that personal computer 3 will be located at a doctor's office, where the patient will return after the data has been stored in the data collection device. Personal computer 3 can be a standard personal computer, or alternatively an equivalent device merely capable of storing the data collected by data collection device 2 and, thereafter, transferring that data to an Analysis Center.

After storage of the collected data in personal computer 3, the operator of the system calls the Analysis Center 6, which can be situated at any location which permits a communications link between personal computer 3 and Analysis Center 6. Subsequent to contacting the Analysis Center, the operator of this system will identify the particular patient by name, age, or any other parameters necessary for complete identification of a particular patient and, of course, identify the particular location at which the personal computer is located. After sufficient identification

information has been entered into the system, personal computer 3 will transmit to the Analysis Center all of the data that has been collected by the data collection device.

The Analysis Center is designed to rapidly analyze the incoming data, and generate a patient report for transmis¬ sion to the doctor's office. It is anticipated that the time necessary for the doctor to remove the data collection device from a patient, transmit the data to the Analysis Center and receive a written report will possibly be no longer than fifteen to twenty minutes. Alternatively, data communications systems exist which will store the report until voluntary retrieval by the doctor. It is also to be understood that the Analysis Center is fully automated and that no trained personnel are required at the Analysis Center to analyze or examine incoming data, or outgoing reports.

Referring now to FIG. 4, there is illustrated further details of data collection device 2 and sensors 1. The data collection device consists of multiplexer 7, which accepts the incoming data from the sensors, multiplexes that data, and applies the data to amplifier 8. The amplified data is thereafter applied to analog to digital (A/D) converter 9, which converts the analog incoming data into digital data for manipulation and storage. Microprocessor 11 is con¬ trolled by a stored program, resident in EPROM 12. External clock 10 drives the microprocessor, and also provides timing signals to other portions of the circuit where necessary. The incoming data, after being processed by microprocessor 11, is stored in random access memory 13. It is, of course, understood that at the time data is transferred from data collection device 2 to the processing system 3, microproces¬ sor 11 will control the extraction of data from random access memory 13 and transmission of that data to personal computer 3 pursuant to instructions stored in EPROM 12, and via communications port 14. The block diagram configuration for data collection device 2 can be readily understood by one skilled in this art and, thus, no further description

will be provided for the functions performed by the data collection device.

In FIG. 5, there is illustrated a generalized flow chart describing the concept of utilizing a computerized chrono¬ thermodynamic analysis for ambulatory data acquisition and computerized data processing of temperature data from a human body, which analysis would be performed at the Analysis Center discussed above. Chronothermodynamic examinations consist essentially of recording skin tempera¬ ture under non-controlled (ambulatory) or controlled conditions, over variable periods of time ranging from less than one hour to several days, utilizing the portable multi¬ channel microprocessor based data acquisition system shown in FIG. 3.

Chronothermodynamic examinations rely on the fact that most diseases involve metabolic and vascular disorders and consequently give rise to significant changes in the chronobiologic behavior of skin temperature in reaction to physiological (internal) and environmental (external) stimuli. The advantages of the chronothermodynamic examina¬ tion are that it explores function and not structure and is strictly non-invasive, which permits fully automatic and computerized data acquisition and processing, and it can be used under a large variety of conditions such as in-bed, ambulatory or with short or long recording periods. The flow chart illustrated in FIG. 5 represents a generalized description of a clinical chronothermodynamic analysis procedure.

In accordance with the procedure set forth in FIG. 6, quasi-continuous recording of skin temperature at sym¬ metrical locations of the body are performed in either an ambulatory or in-bed condition. A total of N thermal sensors are used, one-half of which are located at one symmetrical location and the remaining one-half being located at the opposite symmetrical location of the body. As described in Block 15, the recording time for the temperature is variable, up to several days, and the

measuring frequency is dependent upon specific information and protocols.

The temperature data gathered from symmetrical areas of the body is analyzed by contralateral and ipsilateral com¬ parisons, where ipsilateral comparisons refer to temperature measures taken from the same symmetrical body area and con¬ tralateral comparisons refer to measurements taken from mirror image areas of the body.

As described in Block 16, the inventive technique an¬ ticipates the performance of N/2 contralateral comparisons, and 2L ipsilateral comparisons, where L is defined as the number of significant comparisons per body side. Thus, the total number of comparisons performed, M, is equal to N/2 plus 2L. For each comparison the analysis described in FIG. 6 is performed as set forth below.

Characterization of the data sample occurs at Block 17, which is defined as the temperature series recorded at two locations, which characterization includes either ip¬ silateral or contralateral comparisons. An interdependence relationship is assessed at Block 19, which evaluates the relationship between the temperatures at the two locations ^' being compared. Thereafter, a system of M equations with N unknown parameters are solved to achieve the calculation of an interdependence coefficient (Blocks 21 and 23). The interdependence coefficient, also referred to hereinafter as a regression coefficient, characterizes deviations in temperature values when compared with the straight regres¬ sion line that would represent equal temperatures at any test location and at any time.

The description and solution of such equations will be described hereinafter.

In addition to the characterization of the data sample as previously described, a spectral analysis (described below) is performed on the data after choosing significant fre¬ quency bands, which analysis is performed for each band of frequency chosen (Block 18). Based on the spectral anal¬ ysis, an attenuation coefficient ratio is calculated, again

based on ipsilateral and contralateral comparisons (Block 20).

A second system of M equations with N unknown parameters is solved (Block 22) to calculate an attenuation coefficient (Block 24). The attenuation coefficient characterizes the mean decrease in amplitude of temperature fluctuations.

The manner in which the attenuation coefficients and the interdependence (regression) coefficients are calculated for a specific embodiment of the invention is set forth below.

The attenuation and interdependence (regression coeffi¬ cients) are stored in a statistical data base (Block 25). Thereafter, a preliminary calculation of a chronother¬ modynamic score is performed for each sensor location (Block 26), which score characterizes the possibility of abnormal bodily conditions potentially indicative of disease. Thereafter, (Block 27), the calculated score is compared with previously stored clinical data to assess correlations between newly calculated coefficients and previously stored coefficients to determine whether the calculated CT score is within statistical bounds based on prior stored data (Block 27). Subsequently, an assessment of the chronothermodynamic class CT (Block 28) is performed, which CT class is indica¬ tive of the absence or presence of bodily diseases as described below.

A specific embodiment of the instant invention will now be described wherein there is demonstrated the analysis of temperature data gathered from human female breasts for the purpose of early detection of breast cancer.

In the case of clinical analysis of the breast, tempera¬ ture readings are recorded for twenty-four hours contin¬ uously under ambulatory conditions. Temperature readings are taken every five minutes, from five specific locations on each breast, i.e. four breast quadrants and the nipple/areola area. The thermal sensors are kept in contact with the skin surface by means of self-adhesive non-aller- genic pads. Temperature readings being obtained in this manner demonstrate fluctuations of breast skin temperature

in relation to the metabolic and vascular reactions to multiple stimuli, (physiological, behavioral and environ¬ mental) which effect the patient during the recording process _*

Referring to FIG. 6, there is illustrated the location of five sensors on both the left and right breasts. The nipple/areola area is shown in the center of each breast il¬ lustration, with one sensor 5 being located in the nipple area, and the remaining four sensors 1-4 being located in the four quadrants of the breast as illustrated.

As described above, the temperature from each sensor is recorded at five minute intervals for a twenty-four hour • period and stored in data collection device 2. Subsequent thereto, the data is transmitted to Analysis Center 6, at which time specific algorithms are implemented to process the data and generate a report indicative of the potential , of breast cancer. The thermobiologic analysis performed includes recording the temperature levels of each designated breast area, and comparing that temperature data with the temperatures of each of the other areas examined on the same breast (ipsilateral comparisons), as well as comparing the temperature data with the temperature levels on designated areas of the opposite breast (contralateral comparisons).

FIG. 7 illustrates the concept of contralateral com¬ parisons where the temperature at quadrant one on the left breast is compared with temperature readings taken at each of the other quadrants and the nipple area of the con¬ tralateral breast. Similarly, FIG. 8 illustrates the concept of ipsilateral comparisons where the temperature of each quadrant of the breast is compared with the temperature at each other quadrant of the same breast. The ipsilateral comparisons shown in FIG. 8 for the left breast are also performed on the right breast, although not shown in FIG. 8.

Based on experimental clinical studies, preliminary evaluations have indicated that 16 out of a possible 20 ipsilateral comparisons, and 5 out of a possible 25 con¬ tralateral comparisons, are significant and non-redundant in

the early detection of breast cancer. The significance and non-redundancy of such data was determined by comparing a large number of known "normal" breasts with breasts in various stages of disease. This analysis of clinical data resulted in the determination that 16 out of the 20 ip¬ silateral, and 5 out of the 25 contralateral comparisons were important in terms of providing an automatic, accurate and reliable analysis of the data received.

FIG. 9 illustrates the preferred 5 of 25 contralateral comparisons, and the 16 of 20 ipsilateral comparisons which were found to be significant. As will be described below, the comparisons illustrated in FIG. 9 permit a regression coefficient to be calculated, in accordance with the instant invention, which characterizes the deviation in temperature values when compared with the straight regression line that would represent equal temperatures at any test location, at any particular period in time. The larger the regression coefficient, the more the thermal levels of the two com¬ parison breast areas could be expected to differ. In fact, in diseased conditions, the regression coefficient in the areas overlying the pathological process have been found to be significantly higher when compared with those of the other ipsilateral and contralateral areas. A total of 21 regression coefficients are developed, four for each breast quadrant and five for the nipple/areola area. The manner in which the regression coefficients are calculated will be de¬ scribed below.

In addition to calculating a regression coefficient de¬ pendent upon the ipsilateral and contralateral temperature comparisons, the instant invention also performs a chrono¬ biologic analysis which concerns the time patterns of temperature of each breast area with the time patterns of temperature of all the other breast areas explored. Such comparisons generally reveal a desynchronization of a variable degree, i.e., over time the temperature fluctua¬ tions at one particular area of the breast will exhibit a somewhat different frequency variation then the temperature

0/13092

- 20 - fluctuations of an adjacent or contralateral area of the breast. Accordingly, in order to analyze this chrono¬ biologic data, a cross-spectral type of analysis of the time series of temperature is used to determine the charac¬ teristics of the distribution of the amplitude of the thermal fluctuations by frequency. As it has been deter¬ mined that the cross-amplitude spectra are rather smooth, only six different ranges of periods were considered: 15 minutes; 15-30 minutes; 30-60 minutes; 60-80 minutes; 180- 360 minutes; and more than 360 minutes. Twenty-one such comparisons have been found to be significant and non- redundant in preliminary evaluations based on clinical studies. For the twenty-one comparisons a gain function is calculated in each band of frequency resulting in 126 attenuation coefficients, i.e., 6 frequencies multiplied by 21 comparisons. The attenuation coefficients characterize the mean decrease in amplitude of the temperature fluctua¬ tions as a function of time. It has been found that breast diseases, in particular carcinomas, generally give rise to marked damping of thermal fluctuations. Therefore, the attenuation coefficient in an area overlying a lesion is high, compared with those of other ipsilateral and con¬ tralateral areas. A total of 60 relative attenuation coefficients (10 breast areas multiplied by 6 frequency ranges) are calculated as described below.

Based on the calculation of regression coefficients and attenuation coefficients, a chronothermodynamic score is assigned to each specific breast area depending upon the value of the 21 regression coefficients, and the 60 relative attenuation coefficients. Thereafter, a chronothermodynamic class (CT class) is determined where this classification is derived from an expert system type of analysis based on correlations between the regression and attenuation coef¬ ficients, and the relative findings for physical, mono¬ graphic, echographic, and pathological studies. The CT class represents, from an index from of I-V, an increasing degree of abnormality. After calculation of the CT class, a

report is prepared for the physician, which report is transmitted to the physician's office in the manner described above. The report, a sample of which is discussed below, indicates the score and CT class for each of the five breast areas explored on each breast.

Referring now to FIG. 10, there is shown a data process¬ ing flow chart which describes in block diagram form the chronothermodynamic assessment of breast health that occurs at the Analysis Center with respect to the analysis of temperature data for the early detection of breast cancer. The analysis described in FIG. 10 is performed each time data is received for a particular patient. More par¬ ticularly, referring to FIG. 10, temperature records over a twenty-four hour period, with temperature sampling every five minutes, are received at the Analysis Center, il¬ lustrated by the Data Acquisition portion of the flow chart (Block 29). That data is then analyzed, as described above, wherein 5 contralateral comparisons, and 16 ipsilateral comparisons are performed. (Block 30). Proceeding down the left side of the flow chart for calculation of the regres¬ sion coefficient, the first process that is performed is characterization of the data sample (Block 31), wherein the temperature series recorded at two designated breast locations is compared. Thereafter, a calculation of the regression curve (Block 32) is made in a manner to be described below. The next step in the flow chart is solving a system of 21 equations, with 10 unknown parameters, (Block 33), the result of which is a calculation of a regression coefficient for each breast site (Block 34). The precise manner in which this is done is described below. Each regression coefficient is then stored in a statistical data base, (Block 35), which data base is designed to store all previous data that has been collected for the numerous clinical studies performed in accordance with the instant invention.

Similarly, proceeding to the right side of the flow chart, after the contralateral and ipsilateral comparisons

90/13092

- 22 - are made, the first step in the calculation of the attenua¬ tion coefficient is a spectral analysis for six distinct bands of frequency (Block 36). Thereafter, a calculation of attenuation coefficient ratios occur (Block 37), followed by the solution of a second system of 21 equations, with 10 unknown parameters to calculate the attenuation coefficient for each breast site (Blocks 38 and 39). The attenuation coefficients are also stored in the statistical data base (Block 35).

Next, a preliminary calculation of the CT score for each breast site is made (Block 40), and that calculation is then compared with statistical selection criteria to determine whether the CT score calculated is within statistical bounds based on prior patient data (Block 41). Thereafter, the choice of the chronothermodynamic CT class is made for each breast site (Block 42), and this data becomes part of the report which is forwarded to the physician.

A typical report for a patient may consist of three separate pages. The first page, as shown in FIGS. 11A-11E, consists of the chronogramms for each of the five breast areas for the left and right breasts. More particularly, shown in FIGS. 11A-11E, is temperature in degrees cen¬ tigrade, versus time (over 24 hours) for the areas 1-5 described above. As shown, the temperature is taken every five minutes for a twenty-four hour period to obtain the graphs shown in FIG. 11.

After accumulation of the data as described above an analysis is performed, in accordance with the instant invention, in the manner described in FIG. 10. That analysis consists of determination of a regression co¬ efficient, an attenuation coefficient and a CT score for each breast area. In FIG. 12 is illustrated an example of a patient report, wherein the regression coefficient for the left and right breast, the attenuation coefficient for the left and right breast, and the CT score for the left and right breast is shown.

After calculation of the CT score, a CT class is assigned for each patient. Based on an analysis of clinical data, as will be described below, it has been determined that a CT score of 0-7 is a class I classification; a CT score of 7-13 is a class II classification; 13-20 is class III; 20-34 is class IV; and 34 and above is class V. Classes I and II are considered to be normal conditions for normal breasts; class III is considered to be possibly abnormal and suggestive of disease; and classes IV and V are considered to be a definite disease condition for the breast.

The last page of an exemplary report provided to a physician would consist of what is illustrated in FIG. 13, along with a written recommendation from the Analysis Center to the physician. As shown in FIG. 13, the upper left quadrant of the right breast has a CT class III classifica¬ tion. This is considered abnormal and, therefore, the recommendation from the analysis center would most likely be that this particular patient should have strict and regular surveillance in order to determine whether a disease condition exists, and/or the breast should be examined with mammographic or other techniques to check precisely for the presence of disease.

The precise analysis technique performed in accordance with the instant invention will now be described in detail.

Referring now to FIG. 14, there is shown a typical chronothermogram derived from a twenty-four hour monitoring of breast temperature. As illustrated on the x-axis, temperature monitoring commenced at 1400 hours on day one, and continued until 1400 plus hours on day two. The temperature at this particular breast location varied from a low of 31°C to a high of nearly 35°C. Note particularly the rapid increase in breast temperature at the commencement of darkness, where the darkness interval is illustrated by the dark line between 2200 and 0600 hours. During a typical breast examination, a chronothermogram, as illustrated in FIG. 14, will be generated for all ten areas of the breast being monitored.

Calculation of the regression and attenuation coeffi¬ cients will now be described. Refer first to FIG. 15A wherein, for the sake of explanation, two different breast locations, X and Y, are shown as having constant and equal temperatures of 34^C over a twenty-four hour period. It is understood that location X and Y can be on the same breast, or on different breasts. Similarly, in FIG. 16A, locations X and Y are shown as having identical temperature readings over a twenty-four hour period, which temperature readings, however, vary between temperatures Tl and T2.

Recall from the description set forth above that the regression coefficient characterized deviations in tempera¬ ture values when compared with the straight regression line that would represent equal temperatures at any test location at any point in time. A straight line regression curve is illustrated in FIGS. 15B and 16B, wherein the temperatures at locations X and Y are plotted versus each other. As the temperatures in FIG. 16A are equal throughout the twenty- four hour period, the plot in FIG. 15B is simply a point, i.e., the temperatures were equal and unvarying through the

^{* l"~} 24 hour period. However, as the temperatures at locations X and Y varied over time in FIG. 16A, the plot in FIG. 16B is a straight line indicating equal temperatures at each location over the twenty-four hour period, which line extends between temperatures Tl and T2.

Referring now to FIGS. 17 and 18, FIG. 17A represents temperature variations over time at locations X and Y, wherein tlie temperatures at each location are not equal to each other at any particular point in time. FIG. 17B is a plot of the temperature at point X, versus the temperature at point Y, and represents the regression curve for the temperature variations illustrated in FIG. 17A. As the temperatures at locations X and Y in FIG. 17A are not equal, the regression curve in FIG. 17B deviates from the dotted 45° straight line plot by a certain angle of deviation.

The examples described with respect to FIGS. 15-17 do not represent actual temperature plots taken from a patient, but

are presented for purposes of explanation only. In con¬ trast, FIG. 18A represents actual temperature readings taken from a patient breast area over a twenty-four hour period. As illustrated, the temperature readings at locations X and Y vary greatly with respect to each other over the twenty- four hour period.

Due to the substantial and somewhat random variations between X and Y in FIG. 18A, when X versus Y is plotted in FIG. 18B, the regression curve consists of a cloud of data points, each of which represents a difference in temperature between X and Y at a particular point in the twenty-four hour cycle.

During an actual breast examination, as described above, 16 of the possible 20 ipsilateral comparisons, and 5 of the 25 possible contralateral comparisons, were found to be significant and non-redundant. A plot such as shown in FIG. 18A and the calculation of the regression curve shown in FIG. 18B, is done for each of these 21 comparisons. Once this data is generated, the regression coefficient for each of the possible 10 breast examination sites is calculated, as set forth below.

More particularly, assume that the five sites on the left breast are identified as C1-C5, and the five sites on the right breast are identified as C6-C10. If α is defined as the regression coefficient, then the following system of 21 equations apply: α (Cl, C2) = -180 ARCTAN (b** ) π α (Cl, C3) = -180 ARCTAN (b ₂ ) π α (Cl, C4) = -180 ARCTAN (b _? ) π α (Cl, C5) = -180 ARCTAN (bi _j ϊ

Tt

a ( Cl , Cl ) = -180 ARCTAN (b-?) π

α (Cm, Cn) = -180 ARCTAN (b _n ) it where b-^ b _n is equal to the coefficient of the regres¬ sion curve.

The coefficient of the regression curve b2 bn is calcu¬ lated by inspecting the data sample with respect to a system of coordinates defined by the center of a scattergram and a 45° deviation line.

Data is grouped by: (i) defining classes (subsamples of data), using the text of the variance homogeneity; and (ii) by calculating the median for each class. Thereafter, a regression curve is derived, along with relative regression coefficients b2-bn based upon the median values.

The system of 21 equations is then solved to ^" find the 10 unknown parameters which are the specific regression coefficients for each site, i.e., α (Cl) α (CIO).

The regression coefficients are calculated as α-^ (cm,cn) wherein cm,cn refers to the temperature sensors being compared, and ^" i refers to the comparison made (1-21). The value α characterizes each of the ten breast sites (sensor locations) compared with all other breast sites and is calculated by means of the Huber's robuts estimator.

Each of the 10 calculated regression coefficients are then used iji the manner described below.

The next step in determining the CT classification for a particular patient is calculation of the attenuation coefficient for each breast site. More particularly, as previously described, the attenuation coefficient is derived from the chronobiologic comparison of the time patterns of temperature of each breast area, with the time patterns of all other areas explored. It is known that the time patterns of temperature for particular breast and other bodily areas exhibit a desynchronization with respect to

other bodily areas in terms of responsiveness to external stimuli. For example, referring to FIG. 19, there is represented, by rectangles 1 and 2, the breast tissue surrounding temperature sensors placed at breast locations X and Y. When these breast tissue areas are exposed to external stimuli, such as the temperature variations shown as waveform E in FIG. 19, they exhibit markedly different response characteristics. More particularly, the tempera¬ ture at location X will vary, as shown by waveform X in FIG. 19, while the temperature at location Y will exhibit the response shown by waveform Y in FIG. 19. The differences in response are a result of the natural filtering charac¬ teristics of breast (or other bodily) tissue and, thus, the temperature at each location X and Y will vary at different frequencies in response to the same external stimuli.

Referring now to FIG. 20A, there is shown the temperature response over a twenty-four hour period at two breast locations XI and Yl. As the temperature at each location varies at the same frequency f^, the attenuation coef¬ ficient, which as recalled is the mean decrease in amplitude of the temperature fluctuations, will be equal to 1 as the amplitude and frequency of both waveforms is the same.

In FIG. 2OB, X2 and Y2 are varying at the same frequency f2, but are different in amplitude and the instantaneous amplitude of Y2 is approximately one-third the instantaneous amplitude of X2. Accordingly, in this instance, the attenuation coefficient will be equal to one-third.

FIGS. 21A and 21B represent further examples of tempera¬ ture variation at different frequencies at locations X and Y. In these instances, the relationship between X-*_ + X ₂ and ^γ l ^{+ γ} 2' ^anα - ***** ^an< *** ^γ i- ³ sufficiently complex that the attenuation coefficient cannot be readily ascertained by mere observance, but must be calculated as described below. FIG. 21A illustrates that the frequency f- _j _ and the two chronothermograms (X^ + X ₂ ) and (Y^ + Y ₂ ) have equal attenuation coefficients. In contrast, at frequency f ₂ , the

attenuation coefficient is higher for the chronothermogram (Y ^* L + Y2) compared with the chronothermogram (X*j_ + ^-2 ' _'

FIG. 21B illustrates that at the frequency f 1 ' the chronothermograms X and Y have equal attenuation coeffi¬ cients. By contrast, at any other frequency different from ±1 , the chronothermogram X is attenuated, i.e. all frequencies but f*-_ are filtered.

Assume, as described previously, that there are 10 examination locations (C1-C10), for which it is desired to calculate an attenuation coefficient. Also, as described above, data to calculate the attenuation coefficients is gathered at six different frequencies and 21 separate comparisons were found to be significant and non-redundant. Each of the 60 attenuation coefficients (10 examination sites times 6 different frequencies) is governed by the following relationship, where A is the attenuation coeffi¬ cient: r _β j _e -13 2 π fl/fe j=0

A (Cl, C2) _fl = r _{α e} -ij 2 π fl/fe

1+ j=l where fe is the sampling frequency c-j βj are the coefficients of the A.R.M.A. model (r,r)

A description of the Auto Regressive Moving Average (ARMA) method of spectral analysis is found in Topics in Applied Physics, Volume 34, entitled "Non Linear Methods of Spectral Analysis, edited by S. Haykin, Published by Springer-Verlag, 1983.

There are 21 such equations for each of frequencies fl- f6. These equations are then solved to determine the 60 attenuation coefficients, i.e, 10 examination sites, multiplied by each of the six different frequencies.

The manner of calculating α and β utilizing the A.R.M.A. model (p,q) is as follows:

Given S (n) the data recorded by the sensor C^ S ₂ (n) the data recorded by the sensor C ₂ e (n) a signal representing the internal and external stimuli as a whole a recursive equation is derived from e (n) in order to express S- _j _ (n) and S ₂ (n) as follows:

^a li ^s l ^{{ n} - ^{± ] +} b _lje ⁽ n-j ⁾ i=l j=0

^a 2i ^s l ⁽ⁿ -i ^{) + g} b _2j e ⁽ n-j ⁾ i=l j=0

Taking z transforms in (1) and (2) q

∑ *>lj ² -3 j=o

S ₂ (Z) = E (Z)

P

1 + ,-1 Σ ^L li i=l

q

_. ∑ ^b 2j z ^"3 j=0

S ₂ (z) = E (z)

P 1 + 2 ^a 2i ^Z ~ i=l with z = e -i2πf/fe . f _e . sampling frequency

Given Hi (z) the transfer function characterizing the breast tissue located underneath the sensor C*-_

Si. (z)

Hj; (z) =

E (z)

Given H ₂ (z) the transfer function characterizing the breast tissue located underneath the sensor C ₂

S ₂ (z)

H ₂ (z) =

E (z) S^ (z) and S ₂ (z) are correlated as follows:

p+g

Σ βj z-3 j=0

R _? 1 ( z )

3=1

-α^ and βj are determined using a least square approxima¬ tion of S ₂ (n) by a function S (n) defined as follows: r 2r+l

S (n) = ι Φi s ₂ (n-i) + ^ Φi ^s l (n-i) i=l i=l φ-j_ : estimation of α- _j . and β^ r = p + q

-The order (p + q) is determined based upon an optimum procedure defined by BOX and JENKINS.

Referring again to FIG. 10, it is seen that the foregoing description has reached the point on the flow chart where the regression coefficients, and the attenuation coef¬ ficients, have been calculated for each breast site. Each

coefficient is then stored in a statistical data base and compared with a range of stored coefficients to determine the degree of correlation and whether anomalies exist. Next is the calculation of a CT score and CT class for each breast site, which are determined as follows.

First, a weighting factor is applied to the attenuation coefficients in accordance with the following relationship:

A = pl ^A fl + p2 ^A f2 — + p6 ^A f6

(fl,f2,-~f6) pl+p2+p3 + p6

where pl-p6 are the weighting factor which is dependent upon the degree of correlation of the attenuation coefficient with what is expected from an analysis of the statistical data base. The weighting factor would be large with good correlation, and small with poor correlation. The CT score is then determined as follows:

CT score = g-ια(Cl) + g ₂ A (Cl) (Cl) fl-f6 gi + g2 where gl and g2 are also a weighting factor determined in the same manner as is the weighting factor p.

Once a CT score is established for each examination site, that score is compared (see FIG. 10) with statistical selection criteria gathered from a number of clinical trials (see below). The selection criteria dictates that if a CT score is between 0-7 that site is designated as Class I; 7- 13 as Class II; 13-20 as Class III; 20-34 as Class IV; and 34 and above as Class V. The chronothermodynamic Class CT represents, from an index of I to V, an increasing degree of abnormality with classes I and II, generally representing none, or very small, chronothermobiologic anomaly, and with Classes III, IV and V representing markedly increasing chronothermobiologic anomalies respectfully.

Experimental clinical studies, in accordance with the invention described herein, have consisted of an examination of a total of 339 patients. Prior to utilization of the

invention, all patients were thermographically, mammo- graphically and physically examined. In addition, roentgen- ographic, echographic and cytologic tests were ordered for selected patients depending upon breast structure and density', questionable thermovascular patterns and/or hyper- thermia, and the presence of mammographic opacities or palpable anomalies.

Each patient consented to examination in accordance with the instant invention, in which thermal sensors were taped to appropriate regions of the breast and connected to the portable recorder worn on a belt. Patients were advised to perform their usual domestic and professional activities, but to avoid taking a bath or shower. Temperature measure¬ ments were recorded every two-five minutes over an ap¬ proximate twenty-four hour period. Test procedures were well accepted, with an initial refusal or early removal rate of less than one (1%) percent.

Technical problems were rare, consisting of less than two (2%) of the patients. After removal of the sensors, temperature data was interpreted and analyzed as described above.

Refer now to FIGS. 22 A-D, wherein the depicted his¬ tograms describe correlations between major diagnostic classifications and chronothermobiologic findings. More particularly, patients in the clinical studies were grouped into four major categories: (a) physically and mammo- graphically healthy breasts; (b) benign conditions, such as dysplasia, tumor or mastopathy; (c) borderline conditions, representing a diseased state with a significant possibility of future cancer; and (d) cancers of variable histology and stage.

FIG. 22A illustrates the 117 patients examined, with physical and mammographically healthy breasts. Upon utilization of the instant invention no chronothermodynamic anomalies were found in eighty-six (86%) percent of the patients (CT Class I or II). Accordingly, the rate of false

- 33 - positive findings, i.e., patients with apparently healthy breasts, but classified as CT class III, IV, or V, was 14%.

FIG. 22B demonstrates the distribution of 144 patients with benign disease. Thirty (30%) percent of these patient had no chronothermobiologic anomalies, while the majority of seventy (70%) percent had anomalies of variable degrees (Class CT II-V) . The 101 patients with positive chrono¬ thermodynamic findings had more or less severe glandular or fibrotic dysplasia or solitary and multiple cysts or fibrocystic mastopathy. The 12 patients with marked chronothermodynamic abnormalities (Class CT IV or V), had severe fibrocystic mastopathy with dense and very irregular structure.

The distribution of 18 patients with borderline lesion, i.e. a diseased state with a significant possibility of malignancy is shown in FIG. 22C. This distribution is intermediary with regard to the actual rates of the dif¬ ferent CT classes.

FIG. 22D shows the distribution of 59 patients with cancer. Ninety (90%) percent of these patients were classified by the instant invention as having marked chronothermobiologic anomalies (Classes CT III, IV or V).

It is also significant that during the clinical studies wherein utilization of the instant invention was followed, a timely diagnosis of breast malignancy was concluded in three patients when spot mammograms were ordered on the basis of isolated chronothermodynamic and thermographic anomalies in the presence of a normal standard mammogram which had shown no anomalies. Significantly, in two additional patients who had clinically and mammographically healthy breasts, but positive chronothermodynamic findings (Class CT IV) on their first examination, were later diagnosed as having micro- invasive carcinomas after several months of follow-up monitoring by the attending physician.

Accordingly, the instant invention, when used in the application of breast cancer detection, has at least the following advantages: (1) detecting breast cancer earlier

0/13092

- 34 - and, in particular, in young women with dense breasts, and women with fibrocystic mastopathy; (2) identifying women at high risk of developing breast cancer; and (3) accessing the pre-therapeutic diagnosis of Stage I and II breast cancers based upon the relationship between tumor growth rate and thermovascular changes.

The foregoing embodiment has focused on use of the instant invention in the early detection of breast cancer. However, the invention can also be utilized to analyze any data extracted from a living organism, when that data is extracted from symmetrical portions of the body, e.g., the hands, elbows, knees, feet, contralateral areas of the face, and so forth. Further, the invention is not limited to analyzing temperature data, but can generally analyze any analog signals extracted from symmetric areas of the body, including by way of example, but not way of limitation, electric voltage, current and resistance generated by the body.

FIG. 23 illustrates a generalized data processing flow chart application to clinical chronothermodynamic studies, which studies include a dynamic stress test, utilizing either a ^s physical or pharmacological agent. It is to be noted that previous descriptions of the instant invention did not include the use of a stress test.

With the procedure illustrated in FIG. 23, quasi-con¬ tinuous recording of skin temperature is performed under ambulatory, or in-bed conditions. Temperature is recorded at N symmetrical locations from right/left mirror areas of the human body with recording time being up to several days. The measuring frequency and type of stress test is deter¬ mined in accordance with specific protocols (Block 43).

The temperature records are analyzed during three distinct phases; prior to, during and after a stress test (Blocks 44-47). The stress imposed can include a variety of phenomena including noise, microwave radiation, shock, heat, cold and so forth.

Prior to the stress test, a spectral analysis is per¬ formed in the same manner as was described previously with respect to FIG. 5. If the chronothermograms present contain significant frequencies (Block 48) the data processing continues in accordance with the flow chart set forth in FIG. 5 (Block 51).

As in FIG. 5, the parameters are stored in a statistical data base (Block 53), calculation of single scores for each sensor location is accomplished (Block 54), an assessment of the correlations between parameter/scores and clinical data is performed (Block 55), and a chronothermodynamic class is calculated (Block 56). Alternatively, should the text at Block 50 indicate an absence of significant frequencies prior to the stress test, then an evaluation of charac¬ teristic parameters is performed for each phase of each chronothermogram at Block 49. This same evaluation is conducted during the stress test and after the stress test.

An assessment of the scattering and/or synchronization is performed at Block 52 where a comparative analysis based upon contralateral and/or ipsilateral comparisons is conducted. Thereafter, the information is evaluated, as previously described above, in accordance with the remaining steps of the flow chart in FIG. 23.

FIG. 24 is a more specific chronothermodynamic assessment of neurovascular conditions of the upper extremities, utilizing a cold stress test. It is to be understood that the flow chart set forth in FIG. 24 is a more specific description of the generalized chronothermodynamic study just described with respect to FIG. 23.

Referring to FIG. 24, there is described a chronothermo¬ dynamic assessment of vascular conditions of the hands and fingers, wherein quasi-continuous recording of skin tempera¬ ture is performed for each fingertip under ambulatory or in- bed conditions. Ten thermal sensors are utilized, five sensors being placed at symmetrical locations on each hand. A recording time of 45-60 minutes is utilized, with a sampling period of six seconds. In this embodiment of the

invention, a stress test which includes emersion of both hands in cold water at 15°C is utilized (Block 57).

As set forth with respect to FIG. 23, the chronother¬ mograms are then analyzed in three distinct phases; prior to, during and after the stress test (Block 58). Prior to the stress test, temperatures are recorded for 10-15 minutes (Block 59), during the stress test cooling time is limited to three minutes (Block 60), and after the stress test, recording time is set at 30-45 minutes (Block 61). The data is then evaluated to determine the characteristics para¬ meters for each category taking into account initial teBJperature, rewarming delay, rewarming rate, interval recovering rate, and final recovering rate (Block 62). Thereafter, a scattering assessment is performed utilizing a comparative analysis based upon ipsilateral and contra¬ lateral comparisons (Block 63). Subsequent thereto, the data is then evaluated in Blocks 64-67 in the same manner as has beerC described above.

The instant invention is directed to the collection of data from symmetrical areas of a living organism over predetermined intervals, preferably through use of a portable and lightweight data collection device. The collected data is then stored and transmitted to an Analysis Center, where ipsilateral and contralateral comparisons of data gathered from the symmetrical areas is performed. In addition,, a chronobiologic analysis of time patterns in the data is accomplished for each test location. The combina¬ tion of the ipsilateral and contralateral comparisons, and the chronobiologic analysis advantageously permits accurate and reliable detection of bodily abnormalities that could indicate bodily disease.

Although specific embodiments of this invention have been shown and described, it will be understood that various modifications may be made without departing from the scope and spirit of the invention.