JPH05266002A - Disease condition estimating system applying inter-viscous tissue network - Google Patents

Abstract

PURPOSE:To provide a disease condition estimating system which backs up a doctor to exactly estimate the progress of the disease condition of a patient without relying on the experiences of the doctor. CONSTITUTION:Plural viscera which may possibly be morbid metastasis are connected to each other via the transfer paths and at the same time each viscus is divided into internal tissues in an inter-viscus tissue network model. Based on this model, a disease estimating system carries out the calculation to estimate the change of the disease condition. Then a processor of this system estimates the capacity rate of each internal tissue to be morbid at an optional future time point based on a data base 22 related to the self-growing speed, the inter- tissue metastasis frequency statistics, and the inter-viscus metastasis frequency statistics and also based on the capacity rate of the tissue to be morbid set against the total capacity of each internal tissue of the relevant patient inputted as the initial conditions 18.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a support system for doctors for cancer and the like, surgery, and early determination of the time of admission and discharge. It also relates to a system that can be used for predicting and presenting the future state of lesion distribution in patients suffering from alcoholism, chronic intoxication, etc. Furthermore, it is also possible to use it as a learning simulator system that trains a doctor's expertise in medication by inputting virtual lesion distribution and virtual medication data as initial conditions.

[0002]

2. Description of the Related Art At present, most of the decisions on drug administration to patients, surgery on patients, and timing of admission and discharge depend on the experience of doctors. Several studies have been published on the optimal medication support system for treating gastric cancer, but all of them are limited to the support system within each organ. As a recent example, the National Cancer Center has probabilistically predicted future disease progression by statistically organizing medical records for the past 10 years and expanding the case database. In addition, examples of application of mathematical models to predicting epidemics of infectious diseases and treatment plans for leukemia have been reported. further,
There is a report from Kawasaki Medical School that constructed a medical condition diagnosis support system from clinical laboratory information and evaluated the prognosis (exacerbation or death) for risk diseases as a statistical discriminant value.

[0003]

[Problems to be Solved by the Invention] At present, medication to patients,
Determining the timing of surgery and hospitalization for patients depends largely on the experience of the doctor. Moreover, now that the lack of beds and the lack of nurses are being emphasized, it is an important issue to optimize the number of patients and the timing of surgery or medication. Conventional support system for medication judgment and prediction system of disease progression do not provide predictive information of disease progression or medication effect for lesions of multiple organs related to each other, and also in terms of supporting comprehensive judgment of doctors. It was also not sufficient as a support system for doctors' learning.

An object of the present invention is to provide a medical condition predicting system which assists a doctor to accurately predict the future medical condition progress of a plurality of mutually related organs or the therapeutic effect of medication regardless of experience. To do.

Another object of the present invention is to provide a medical condition predicting system capable of virtually setting a lesion state of a patient in the initial stage and giving an advice on an optimum timing of medication.

Another object of the present invention is to provide a medical condition predicting system suitable for predicting a lesion state and learning a drug effect regardless of actual experience.

[0007]

The system of the present invention comprises an interactive input device for inputting initial conditions for lesion prediction calculation and medication data during lesion progression, database accumulating means, and calculation processing. It includes a processing device and a display device for presenting a prediction calculation result to a doctor, and connects a plurality of organs having a possibility of metastasis of a lesion by body fluid circulation by the metastatic route, and connects each organ to a plurality of internal tissues. The medical condition prediction process is performed based on the divided network model between organ tissues. As the initial condition, the volume ratio of the diseased tissue to the total volume of each internal tissue of each organ of the target patient is input, and as a database, at least the self-proliferation rate of lesion tissue, the metastasis frequency statistics between tissues, and A database of metastasis frequency statistics is stored in advance. The processing device performs a predictive calculation for obtaining a lesion tissue volume ratio of each tissue at an arbitrary future time point using the volume ratio of the lesion tissue for each internal tissue under the initial condition and the database.

Another feature of the present invention is that the database accumulating means further accumulates a case database created by using the volumetric ratio of lesion tissue and the manifestation symptom as a search key item, and the processing device is configured to perform predictive calculation. The volume ratio of the lesion tissues of each organ and each tissue obtained as a result is sorted in descending order, and the case database is searched in order from the lesion tissues having the largest volume ratio using the corresponding tissue number as a key item.

Still another feature of the present invention is that the database storage means further stores a medication database indicating a healing tissue number corresponding to a drug name and its healing rate,
The processing device is for predicting and calculating the volumetric ratio of the lesioned tissue for each internal tissue at any future time point in which the medication effect is taken into consideration by using the medication data during the progress of the lesion and the medication database.

The software of the processor in a representative embodiment of the invention comprises the following subroutines:

(1) Output a volumetric distribution map of lesioned tissue of each organ or tissue at a certain time step, and (2) calculate a volumetric rate at the next time step under the set interaction of organs and tissues. (3) Sort the capacity rates in descending order of value, search the case database using the tissue numbers corresponding to them as search key items, and output the symptoms that appear in the patient at that time. (4) Premedication At the time step specified in the database, the target tissue number and healing rate are extracted using the drug name as a key item, and the volume ratio data is converted.

[0012]

According to the present invention, it is possible to realize, on software, the expertness in the prediction of disease progression and the determination of medication, surgery, and the time of admission and discharge, which has been conventionally determined by a doctor. As a result, it is possible to support the doctor so that the patient's subsequent condition does not greatly differ due to the discretionary difference depending on the experience of the doctor.

[0013]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a diagram showing the outline of a prediction system according to an embodiment of the present invention. In the figure, reference numerals 20 and 26 denote network models between organs and tissues which are targets of prediction calculation in the processing device. This organ-tissue network model is a system in which a plurality of organs 10 having a possibility of metastasis of a lesion due to circulation of body fluid are connected by a metastatic path 12 and each organ is divided into a plurality of internal tissues. For example, brain 1, lung 2, heart 3,
Number each organ such as 4, stomach, 5, liver, and so on. Each organ is divided into multiple internal tissues and numbered in order. When the organ number and the tissue number are collectively written as (5, 2), it means the second tissue of the liver. Organization (i, j)
The ratio of the lesion tissue volume to the total volume is defined as N (i, j). Here, N generally has a variable range of 0 or more and 1 or less, and N = 0 means a normal tissue.

In the prediction system, the present time, that is, t = 0
The N (i, j) examination result of the target patient in is input as the initial condition 18 of the prediction calculation. As a method of evaluating N (i, j), for example, a numerical value obtained by dividing the area of a lesion visually or automatically discriminated from CT or radiograph by the front area of the target tissue is used. Specifically, the initial condition 18 is patient name, disease name, organ number i, tissue number j, and N of those.
It consists of (i, j) and is keyed in from an interactive input device.

Since the processing device performs the display shown in 20 of FIG. 1, the data is output to the display device in accordance with the input of the initial condition. A rectangle 14 in the drawing corresponds to each tissue, and a hatched area 15 represents a tissue in which N is equal to or larger than a certain threshold value. By displaying and displaying the current lesion distribution of the target patient on such an organ-tissue network model, it is possible to grasp the patient's medical condition or the tissue requiring surgery or medication with a visual image. Another expression method of the lesion distribution is a method of displaying the color on the color output device by associating the value of N with the color of each internal tissue of each organ. For example, a warmer color is designated as the value of N increases, and normal tissues are expressed in cold colors.

The processing device mathematically solves the growth and metastasis of lesions as a time development problem. For this reason, a non-linear differential equation concerning the time change rate of N is introduced. If the rate of change of N (i, j) with time is M (i, j), this variable dependence is N in the tissue (i, k) adjacent to (i, j) in addition to self-reproduction.
It depends not only on (i, k) but also on N (m, n) of other organ tissues (m, n) (m is not equal to i) through the network with other organs. Therefore, M (i,
j) is described in the next three sections. 1. Self-propagation: This is proportional to the square of N (i, j).
The proportionality coefficient is A (i, j), or A for short. 2. Metastases from other tissues (i, k) within the same organ: This is proportional to the product of N (i, j) and N (i, k). The proportionality coefficient is B (i, j, i, k), or B for short. 3. Metastases from other tissues (m, n) of different organs: This is proportional to the product of N (i, j) and N (m, n). The proportional coefficient is C (i, j, m, n), or C for short.

From the above, M (i, j) is M (i, j) = A (i, j) * N (i, j) * N (i, j) + B (i, j, i, k) ) * N (i, j) * N (i, k) + C (i, j, m, n) * N (i, j) * N (m, n) (Equation 1). Since the left side of (Equation 1) is the rate of change over time, N (1) after one hour step of M (i, j) is evaluated according to the evaluation of the difference method.
The difference between (i, j) and the current N (i, j) is evaluated by (Equation 2). M (i, j) = (N '(i, j) -N (i, j)) / DT (Equation 2) Here, DT is one time step width.

If N (i, j) of a certain time step is given from (Equation 1) and (Equation 2), N '(i, j) after one time step can be calculated by (3). N '(i, j) = N (i, j) + DT {A (i, j) * N (i, j) * N (i, j) + B (i, j, i, k) * N ( i, j) * N (i, k) + C (i, j, m, n) * N (i, j) * N (m, n)} (Equation 3) (Equation 3) For all i and j, N (i, j) = N ′ (i, j) (Equation 4) is set and the calculation of (1) and (2) is repeated again.

The coefficient A of (Equation 3) is determined by interpolating the growth curve of the number of cells when the focal tissue cells are self-cultured into the curve of A * N * N. Coefficients B and C generate inter-tissue metastasis data of disease from clinical data on corresponding disease as follows. From other tissues (i, k) in the same organ (i, k
j) and other tissues of different organs (m,
Regarding the transfer from (n) to (i, j), a transfer number database is constructed with the form items as shown in FIGS. 4 and 5, respectively. In FIG. 4, i, j, k, the number of transfer cases, and in FIG. 5, i, j, k
The key items are j, m, n, and the number of transfers. That is, the database accumulating means 22 of FIG. 1 stores a database of the coefficient A obtained as a result of the self-culture of the focal tissue of each internal tissue of each organ as the data of the interaction between each organ and the internal organs shown in FIG. The database of the coefficient B calculated from the database of the number of cases of transfer between organizations and the database of the coefficient C calculated from the database of the number of cases of transfer between organs shown in FIG. 5 are included. The processor uses the input initial conditions and these databases, and repeats the calculation of (Equation 3) to obtain N ′ (i, i, i) at any future time point t (for example, one month later).
j) Find the value for each (i, j). The result is output to the display device in the same manner as the initial lesion distribution, for example, in FIG.
Display as shown in FIG.

According to the above-described system for predicting and displaying lesion distribution, the time of hospitalization, the time of medication,
It is possible to support the doctor's judgment such as the timing of surgery. It is also effective as a doctor's learning support system.

Database storage means 22 of the embodiment shown in FIG.
Further, using the drug name as a search key, the organ number i, the tissue number j, and the healing rate that are healed by the drug are V
A medication database recording (i, j) is stored. The above-mentioned calculation of the future medical condition prediction can be performed by assuming that the drug has been administered at a certain time and taking into consideration the healing effect thereof.

That is, an array variable F (t) indicating whether or not to administer medication is set up at each time step i. F
(T) = 0 means no medication at this time step, F (i) = 1 means dosing. The array variable F (t) is created by the processing device by inputting the data 24 of the drug name and the administration time into the interactive input device. In the calculation of (Equation 3) and (Equation 4), this flag is checked at each time step, and if F (t) = 1, the following N change calculation by medication is executed for t. In this change calculation, first, the drug database is searched using the drug name as the search key, and the organ number i, which is cured by the drug,
The tissue number j and the healing rate V (i, j) are output from the medication database. Next, N ′ at the next time step when F (t) = 1 is calculated as N ′ (i, j) = N ′ (i, j) * exp (−V (i, j) * DT) (number By replacing with 5), it becomes possible to perform predictive calculation of lesion distribution in consideration of the drug effect.

FIG. 2 shows a processing flow in the processing apparatus when the healing effect of this medication is taken into consideration. The calculation is repeated for the specified number of steps, and future values of N are obtained for all i and j, and as shown in FIG.
26 is displayed. By such a prediction process, it is possible to predict the transition of the medical condition by variously setting the medication timing, which is useful for determining the appropriate medication timing.

The embodiment shown in FIG. 1 illustrates a symptom predicted to appear in a patient at that time from the lesion distribution at a certain point in the future thus obtained. For this reason, the medical condition prediction system includes a case database 28 listing cases corresponding to the volume ratio of the diseased tissue of each tissue of each organ, and the processing device performs the processing of the flow shown in FIG. First, the predicted value N of the volumetric ratio of the diseased tissue for each internal tissue of each organ obtained by calculation
Sort (i, j) in descending order of value. Next, the case database is searched in descending order of capacity ratio using the corresponding organization number as a search key item. These case items are output to the display device. For example, it is preferable that the output character display color is divided according to the value of N, and the case of high urgency is displayed in warm color, and the case of low urgency is displayed in cold color. Further, for a case that requires urgent treatment, the doctor can be alerted by blinking the character display.

[0025]

INDUSTRIAL APPLICABILITY As described above, according to the present invention, it is possible to assist a doctor to accurately predict the future disease state progression of a plurality of mutually related organs or the therapeutic effect of medication regardless of experience. Therefore, it becomes possible for the doctor to appropriately determine the timing of medication, the time of hospitalization, or the time of surgery. Moreover, it can be used as a warning to patients who are suffering from alcoholism or chronic poisoning.

[Brief description of drawings]

FIG. 1 is a conceptual diagram showing a schematic configuration of a prediction system of the present invention.

FIG. 2 is a flowchart of time evolution calculation according to the embodiment.

FIG. 3 is a flowchart for outputting a case from a capacity ratio in the final step of the time evolution calculation of the embodiment.

FIG. 4 is a diagram showing a database structure of the number of metastases between different tissues in the same organ according to the example.

FIG. 5 is a diagram showing a database of the number of cases of metastasis between tissues of different organs in the example.

Front page continuation (72) Inventor Akihide Hashizume 1-280 Higashi Koikekubo, Kokubunji, Tokyo Metropolitan Research Center, Hitachi Ltd.

Claims (5)

[Claims] 1. A lesion prediction calculation based on an organ-tissue network model in which a plurality of organs having a possibility of metastasis of a lesion due to body fluid circulation are connected by their metastasis path and each organ is divided into a plurality of internal tissues. This is a system for predicting pathological conditions, which is an interactive input device for inputting initial conditions for lesion prediction calculation and medication data during the course of lesion, self-propagation rate of lesion tissues, frequency statistics between tissues, and metastasis between organs. By accessing the database storage means from the database storage means that stores the database relating to frequency statistics and the volume ratio of the diseased tissue to the total volume of each internal tissue of the target patient input as the initial condition, the database storage means can be accessed at any future time point. It is characterized by including a processing device for predictively calculating the volumetric ratio of the diseased tissue for each internal tissue, and a display unit for displaying the result of the prediction calculation. Jo prediction system. 2. The predictive calculation of the processing device is such that the term proportional to the square of the volume ratio of its own tissue number corresponding to self-proliferation of pathogenic cells and the capacity of different tissue number corresponding to metastasis from another tissue. The medical condition predicting system according to claim 1, wherein the medical condition predicting system is executed by an equation including a term of a product of rates. 3. The database storing means further stores a healing tissue number corresponding to a drug name and a medication database showing a healing rate thereof, and the processing device stores the medication data during the progress of the lesion and the medication database. The disease state prediction system according to claim 1, wherein the volume ratio of the diseased tissue for each of the internal tissues at any future point in time that takes into account the medication effect is calculated by using. 4. The medical condition predicting system according to claim 1, wherein the database accumulating means further accumulates a case database created by using a volumetric ratio and a manifestation symptom of the diseased tissue as search key items. 5. The processing device sorts the volume ratios of the lesion tissues of each organ and each tissue obtained as a result of the prediction calculation in descending order, and selects from the lesion tissues having a large volume ratio using the corresponding tissue number as a key item. The medical condition prediction system according to claim 4, wherein the case database is searched in order.