JP5348998B2 - Image search apparatus and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To retrieve image data corresponding to another examination object with high similarity on the basis of image measurement data and physiological examination data obtained from image data to be examined. <P>SOLUTION: The image retrieval apparatus calculates first time-series parameter ratios by a parameter ratio calculation unit 22 on the basis of image measurement data and physiological examination data which are obtained on the basis of various image data of a patient to be examined, evaluates similarity between the first parameter ratios and time-series parameter ratios of a plurality of patients different from the patient to be examined by using a similarity parameter comparison/extraction unit 24, retrieves various image data of the other patient representing the second parameter ratios having the highest similarity with the first parameter ratios of the patient to be examined, and displays the retrieved data on a display. <P>COPYRIGHT: (C)2010,JPO&amp;INPIT

Description

  The present invention relates to an image search apparatus and a method for searching for image data of another test object that exhibits a condition such as a medical condition similar to the condition of a test target such as a patient.

  When diagnosing the medical condition, etc. of the subject to be examined, the patient's medical condition, etc., is searched for similarities with other patient conditions, and from this search result, past diagnosis results, treatment methods, etc. for the patient are searched. A way to refer is needed. As such a method, for example, there is a representative technique such as ontology in Web technology. When utilizing information such as past patient diagnosis results, there is a technique for obtaining similarity between information such as past patient diagnosis results and current patient diagnosis result information. For example, Patent Documents 1 to 6.

In addition, there is a technique for facilitating comparison between information on the results of statistical processing of a large number of past patient information and information on a patient to be diagnosed and obtaining an objective understanding of graph display based on statistical processing information. For example, it is patent document 7. FIG. This patent document 7 also discloses that image data of other patients exhibiting similar symptoms are displayed.
As described above, there is a method of performing statistical processing from a large number of past patient information and searching for similarities between the statistical processing results and the information of the patient to be diagnosed. There is also a method of presenting the diagnostic effectiveness rate in order to improve the accuracy of diagnosis of the patient.

However, in the search technology using the Web technology represented by ontology, it is not possible to directly search for patient image data, and it is necessary to replace the wording used for the search with some information.
JP 2007-275216 A JP 2007-275408 A JP 2007-287018 A JP 2007-275440 A JP 2007-286945 A JP 2007-287027 A JP 2005-4780 A

  An object of the present invention is to provide an image search apparatus and method for searching image data corresponding to another test object having high similarity based on image measurement data or physiological test data acquired from the image data of the test object. There is.

According to a first aspect of the present invention, there is provided an image search apparatus according to claim 1, wherein at least a time-series image measurement data acquired based on image data to be inspected and time-series physiological examination data to be inspected. A parameter ratio calculation unit for calculating a parameter ratio of 1; time-series image measurement data acquired based on image data of a plurality of inspection objects different from the inspection object; and a time-series physiological examination of each inspection object A similarity evaluation unit that evaluates the similarity of changes over time between the second parameter ratio and the first parameter ratio of a plurality of time series calculated based on the data, and an evaluation by the similarity evaluation unit As a result, the second parameter ratio that is most similar to the change of the first parameter ratio with the passage of time among the plurality of changes of the second parameter ratio is determined, and the determined second parameter ratio is indicated. Another inspection Comprising an image retrieval unit for retrieving at least the image data corresponding to the elephants.

According to a sixth aspect of the present invention, there is provided a first time-series search method based on time-series image measurement data acquired based on image data to be examined and time-series physiological examination data to be examined. A plurality of parameters calculated based on time-series image measurement data acquired based on image data of a test object different from the test object and time-series physiological test data of another test object The time-series second parameter ratio and the first parameter ratio are evaluated by the similarity of changes over time . As a result of the evaluation, the first parameter ratio among the plurality of second parameter ratio changes is evaluated. The second parameter ratio having the highest similarity to the change with the passage of time is determined, and at least image data corresponding to another inspection object indicating the determined second parameter ratio is searched.

  ADVANTAGE OF THE INVENTION According to this invention, the image search apparatus which can search the image data corresponding to another test object with high similarity based on the image measurement data and physiological test data acquired from the image data of test object, and its method can be provided. .

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 shows a configuration diagram of an in-hospital image search system to which an image search apparatus is applied. This in-hospital image search system has a radiation department 1, a physiological examination department 2, and a medical department 3. Although there are other departments in the hospital, they are omitted here.
In the radiation department 1, for example, an ultrasonic apparatus 4, a CT apparatus 5, and various image diagnostic apparatuses such as an MRI apparatus are installed. Each of the ultrasonic device 4, the CT device 5, the MRI device, and the like acquires image data to be examined such as a patient, for example, ultrasonic image data, CT image data, MRI image data, and the like. The radiation department 1 includes an image information database (hereinafter referred to as an image information DB) 6 as a first image database, a parameter information database (hereinafter referred to as a parameter information DB) 7, and a personal computer (PC). A terminal 8 (hereinafter referred to as a PC terminal) and a server 9 are installed. The ultrasonic device 4, CT device 5, MRI device, etc., image information DB 6, parameter information DB 7, PC terminal 8, and server 9 are connected via a network 10.

  Among these, the image information DB 6 stores various image data such as ultrasonic image data, CT image data, and MRI image data for each patient, and analyzes them as image measurement data obtained when acquiring these image data. Data and image-accompanying data are stored. The analysis data is acquired by analyzing ultrasonic image data such as a three-dimensional echocardiogram, CT image data obtained by capturing a blood flow image, MRI image data, and the like. The image supplementary data is data attached to, for example, ultrasonic image data, CT image data, MRI image data, or the like. For example, when patient ID information is attached to a patient, various image data such as ultrasound image data, CT image data, and MRI image data of the patient, and analysis data and image incidental data as image measurement data are included in the patient image. A data group is stored linked to the patient ID information of the patient. Note that, for example, various image data such as ultrasonic image data, CT image data, and MRI image data are provided with date and time acquired by imaging, and can be arranged in time series.

  Specifically, the image measurement data includes, for example, image data itself such as ultrasound image data, CT image data, MRI image data, and the form and blood flow obtained from these image data, or obtained from the image data. Physical information such as CAG and PCI, and measurement information such as restenosis rate, EF (ejection amount), strain, and pulse Doppler.

  Among these, EF (ejection amount) is acquired from image data obtained by injecting contrast medium into the left ventricle during coronary angiography examination of patients. It can be seen from such image data that the left ventricle of the heart repeatedly contracts and expands. This is called a left ventricular contrast examination and is useful when examining cardiac function. The amount of blood that is pumped when the left ventricle contracts once is the amount obtained by subtracting the volume when contracted from the volume when the heart is expanded. This is the ejection amount. A value obtained by dividing the ejection amount by the capacity when the heart is expanded is called ejection fraction (EF), and is an index representing the function of the pump function of the heart. When the ejection rate is 50% or more and the function is lowered, the ejection rate is lowered.

  The parameter information DB 7 includes time-series parameter ratios calculated by the server 9, that is, image measurement data such as analysis data of various image data and image incidental data, and examination targets such as patients acquired from the physiological examination department 2. The time-series parameter ratio calculated based on the physiological test data is stored.

  The PC terminal 8 receives image data such as ultrasound image data, CT image data, and MRI image data acquired by various image diagnostic apparatuses such as the ultrasound apparatus 4, the CT apparatus 5, and the MRI apparatus in response to an operator's operation. Appears on the display. In this case, the PC terminal 8 receives various image data directly from various image diagnostic apparatuses such as the ultrasonic apparatus 4, CT apparatus 5, and MRI apparatus and displays them on the display, or reads various image data from the image information database 6. On the display.

In the physiological examination department 2, various examination apparatuses such as a physiological examination apparatus 11 and a DNA examination apparatus 12 are installed. The physiological examination department 2 includes a physiological examination information database (hereinafter referred to as physiological examination information DB) 13 as a second image database, a personal computer (PC) and its terminal (hereinafter referred to as PC terminal) 14. And are installed. The physiological examination apparatus 11, the DNA examination apparatus 12, the physiological examination information DB 13, and the PC terminal 14 are connected via the network 10.
The physiological examination apparatus 11 acquires physiological data to be examined such as a patient, for example, BNP, CTR, FBS, HbA1c, and the like. Among these, BNP (brain natriuretic peptide) is a hormone secreted from the heart and synthesized mainly in the ventricle. Blood BNP is thought to quickly reflect the state of heart failure. The reference value of BNP for healthy people is 20 pg / mL or less. There is an opinion that 50 pg / mL and 100 pg / mL are appropriate for screening for the presence of heart disease and early heart failure. There is also data that sensitivity is 97% and specificity is 84% in the diagnosis of heart failure if 22 pg / mL is cut off.

  CTR is the cardiothoracic ratio of the chest X-ray examination. In chest X-rays, the heart appears white in the center of the chest and protrudes slightly to the left. On X-rays, the width of the heart is normal less than half the width of the chest. When the width of the heart is greater than half the width of the chest, the heart expands. This ratio is called a cardiothoracic ratio, and a cardiothoracic ratio of less than 50% is regarded as normal. It is possible to predict whether or not the heart is burdened by looking at the thickness of the blood vessels visible in the lungs and how they appear.

The DNA testing device 12 tests DNA to be tested such as a patient.
In the physiological examination information DB 13, physiological data such as BNP, CTR, FBS, and HbA1c acquired by the examination of the physiological examination apparatus 11 and DNA data to be examined such as a patient obtained by the examination of the DNA examination apparatus 12 are stored. Stored. Similarly to the above, when patient ID information is attached to a patient, the patient's physiological data and DNA data are linked and stored as the patient's physiological data group by the patient ID information of the patient. Note that, for example, physiological data and DNA data can be arranged in time series because date and time acquired by the examination are attached.

  The PC terminal 14 displays on the display physiological data such as BNP, CTR, FBS, HbA1c and DNA data acquired by various testing devices such as the physiological testing device 11 and the DNA testing device 12 in response to the operation of the operator. To do. In this case, the PC terminal 14 receives various physiological data and DNA data directly from various inspection apparatuses such as the physiological inspection apparatus 11 and the DNA inspection apparatus 12 and displays them on the display, or various physiological data and DNA data from the physiological inspection information DB 13. Is read and displayed on the display.

In the medical department 3, a medical record information database (hereinafter referred to as a medical record information DB) 15, a personal computer (PC) and its terminal (hereinafter referred to as a PC terminal) 16 are installed. The chart information DB 15 and the PC terminal 16 are connected via the network 10.
The medical record information DB 15 stores medical record information such as medical records of examination subjects such as patients, records of examination results, and medication records.
In response to the operation of an operator such as a doctor, the PC terminal 16 associates data such as a medical record of a subject to be examined, a record of a test result, a record of medication, etc. with a subject to be examined such as a corresponding patient in the medical record information DB 15. And stored in the medical record information DB 15. Further, the PC terminal 16 displays data such as medical records of examination targets such as patients, examination results, and medication records stored in the medical chart information DB 15 in response to an operator's operation.

  The server 9 searches for image data of other examination objects that exhibit a condition such as a medical condition similar to the condition of the examination target such as a patient, for example, ultrasonic image data, CT image data, MRI image data, and the like. It has a function as a search device. FIG. 2 shows a functional block diagram of the server 9. The server 9 includes a target patient selection unit 20, a target patient data acquisition unit 21, a parameter ratio calculation unit 22, a database storage unit 23, a similar parameter comparison / extraction unit 24, a data extraction unit 25, and a display unit. 26.

  The target patient selection unit 20 sends a list of patients for selecting a patient to be examined to the PC terminal 8, for example, and displays the list of patients on the display of the PC terminal 8. When a patient is selected on the PC terminal 8, the target patient selection unit 20 sends an identification number for identifying the patient, for example, patient ID information, to the target patient data acquisition unit 21.

  When receiving the patient ID information from the target patient selection unit 20, the target patient data acquisition unit 21 receives, for example, ultrasound image data, CT image data, and MRI images of the patient having the patient ID information from the image information DB 6 according to the patient ID information. Various types of image data such as data, and analysis data and image incidental data are read out as image measurement data obtained when the image data is acquired. At the same time, when receiving the patient ID information from the target patient selection unit 20, the target patient data acquisition unit 21 reads out physiological data and DNA data of the patient Q from the physiological examination information DB 13 according to the patient ID information.

  The parameter ratio calculator 22 calculates at least a first parameter ratio in time series based on image measurement data and physiological examination data acquired based on various types of image data of the patient. The first parameter ratio includes a ratio of the time-series image measurement data and the time-series physiological examination data to a preset parameter standard value, an increase rate in the image measurement data and the physiological examination data, and the image measurement data. And the rate of time increase in physiological test data.

  FIG. 3 shows a specific example of the calculation of the first parameter ratio F. For example, periods D1 and D2 are attached to the time t. A, B, and C in the figure are parameters, and are, for example, EF (ejection amount) acquired from, for example, BNP, CTR, and image data of patient physiological data.

  The parameters A, B, and C are, for example, forms obtained from image data such as ultrasound image data, CT image data, and MRI image data, physical information such as blood flow, CAG, and PCI, restenosis rate and strain, It may be measurement information such as pulse Doppler, or physiological data such as BNP, CTR, FBS, HbA1c, or DNA data.

  The parameter A changes to a0, a1, and a2 over time t. Similarly, the parameter B changes to b0, b1, and b2 with the passage of time t. The parameter C changes to c0, c1, and c2 with the passage of time t.

However, the parameter ratio calculation unit 22 calculates the following equation to obtain the first parameter ratios F (A), F (B), F (C),..., F (N). Here, std indicates a parameter standard value.
F (A) = {a1 / a_std, a2 / a_std, ..., (a1-a0) / a0, (a2-a1) / a0, ...
(a1-a0) / D1, (a2-a1) / D2, ...}
F (B) = {b1 / b_std, b2 / b_std, ..., (b1-b0) / b0, (b2-b1) / b0, ...
(b1-b0) / D1, (b2-b1) / D2, ...}
F (C) = {c1 / c_std, c2 / c_std, ..., (c1-c0) / c0, (c2-c1) / c0, ...,
(c1-c0) / D1, (c2-c1) / D2, ...}
:
F (N) = {n1 / n_std, n2 / n_std, ..., (n1-n0) / n0, (n2-n1) / n0, ...
(n1-n0) / D1, (n2-n1) / D2, ...}
{A1 / a_std, a2 / a_std,...} In the first parameter ratio F (A) and {b1 / b_std, b2 in the first term in the first parameter ratio F (B). / b_std,...} represents a ratio (standard ratio) to the parameter standard value std.
{(A1-a0) / a0, (a2-a1) / a0,...} Of the second term in the first parameter ratio F (A) or the second term in the first parameter ratio F (B) {(B1-b0) / b0, (b2-b1) / b0,...,}, Etc. represent the rate of increase in the image measurement data and the physiological examination data.
{(A1-a0) / D1, (a2-a1) / D2, ...} of the third term in the first parameter ratio F (A) and the third term in the first parameter ratio F (B) {(B1-b0) / D1, (b2-b1) / D2,...} Represents the rate of time increase in the image measurement data and the physiological examination data.

  In addition, the parameter ratio calculation unit 22 calculates the above equation in the same manner as the first parameter ratios F (A), F (B), F (C),. Second parameter ratios F ′ (A), F ′ (B), F ′ (C),... Of other patients based on image measurement data and physiological examination data acquired based on various types of patient image data. , F ′ (N).

  The database storage unit 23 stores the first parameter ratios F (A) to F (N) of the patient calculated by the parameter ratio calculation unit 22 in the parameter information DB 7. In this case, the database storage unit 23 associates with the patient ID information of the patient, and maintains a link with various image data such as ultrasound image data, CT image data, MRI image data of the patient, and the first parameter. The ratios F (A) to F (N) are stored in the parameter information DB 7.

  The similar parameter comparison / extraction unit 24 receives the first parameter ratios F (A) to F (N) of the patient to be examined calculated by the parameter ratio calculation unit 22, and the patient to be examined from the database storage unit 23. Read out the second parameter ratios F ′ (A), F ′ (B), F ′ (C),..., F ′ (N) of the other non-patients, and these first parameter ratios F (A) -F (N) and a function as a similarity evaluation unit for evaluating the similarity between the second parameter ratios F '(A) to F' (N).

For example, the similar parameter comparison / extraction unit 24 performs the first parameter ratios F (A) to F (N) of the patient P1 to be examined as shown in FIGS. The similarity between the second parameter ratios F ′ (A) to F ′ (N) of patients P2 to P4 is evaluated. The first and second parameter ratios F (A) to F (N) and F ′ (A) to F ′ (N) are respectively composed of a standard ratio, an increase rate, and a time increase rate.
Therefore, the similar parameter comparison / extraction unit 24 uses the standard ratio, the increase rate, and the time increase rate of the first and second parameter ratios F (A) to F (N) and F ′ (A) to F ′ (N). A group of patients having a high degree of similarity with respect to the patient P1 to be examined from a plurality of patients stored in the parameter information DB7, for example, A plurality of other patients P2 to P4 are searched.
For example, when the similar parameter comparison / extraction unit 24 is in the state of the change portion H among the changes in the parameters of the patient P1 to be examined shown in FIG. 4A, the state of the change portion H is the other patients P2 to P4. To determine which patient is most similar.
The similar parameter comparison / extraction unit 24 calculates the time lapse between the standard ratio and the time increase rate of the first and second parameter ratios F (A) to F (N) and F ′ (A) to F ′ (N). If the degree of coincidence of the change accompanying the change is verified, it is determined that the change in the parameter of the patient P2 is most similar to the change in the parameter of the patient P1 to be examined.

As a result of the evaluation by the similar parameter comparison / extraction unit 24, the data extraction unit 25 determines the first parameter ratios F (A) to F (F) among the plurality of second parameter ratios F ′ (A) to F ′ (N). N) a function as an image search unit that searches for at least various image data corresponding to another patient having the second parameter ratio F ′ (A) to F ′ (N) having the highest similarity, for example, the patient P2. Have.
The data extraction unit 25 displays and outputs at least various image data corresponding to another patient P2 showing the second parameter ratio having the highest similarity to the first parameter ratio on the display through the display unit 26 as a result of the search. .

Next, the data flow from the calculation of the first parameter ratios F (A) to F (N) of the patient P1 to be examined to the search of the image data of the patient P2 having the highest similarity is shown in FIG. This will be described with reference to a flowchart.
The target patient selection unit 20 sends a list of patients for selecting a patient to be examined to the PC terminal 8, for example. A list of patients is displayed on the display of the PC terminal 8. When a patient is selected on the PC terminal 8 by the operation of an operator such as a doctor, the server 9 stores, for example, ultrasound image data of a patient having the patient ID information from the image information DB 6 according to the patient ID information of the patient to be examined. Various image data such as CT image data and MRI image data, and analysis data and image incidental data are read as image measurement data obtained when the image data is acquired. At the same time, the server 9 reads the physiological data and DNA data of the patient Q from the physiological examination information DB 13 in accordance with the patient ID information of the patient to be examined.

On the other hand, the PC terminal 8 receives an operation of an operator such as a doctor and issues an instruction to select another patient having the same disease as that of the patient P1 to be examined. The selection of the other patients is performed, for example, by disease, in a hospital to which an operator such as a doctor belongs, or in a group in which a plurality of patients are arbitrarily divided.
The server 9 performs various image data such as ultrasound image data, CT image data, MRI image data, and the like of the patient having the patient ID information from the image information DB 6 according to the patient ID information of other patients not to be examined, and the image data. Analytical data and image-accompanying data are read out as image measurement data obtained when acquiring. At the same time, when the server 9 receives the patient ID information of the other patient who is not the examination target, the server 9 reads the physiological data and DNA data of the patient Q from the physiological examination information DB 13 according to the patient ID information.

  Next, the server 9 uses at least first time-series parameter ratios F (A) to F (N) based on image measurement data and physiological examination data acquired based on various image data of the patient P1 to be examined. ) Is calculated. Further, the server 9 calculates the above formula, and based on the image measurement data and the physiological examination data acquired based on the various image data of the other patient, the second parameter ratio F ′ (A) of the other patient. -F '(N) is obtained.

Next, the server 9 includes first parameter ratios F (A) to F (N) of the patient to be examined and second parameter ratios F ′ (A) of a plurality of other patients who are not the patient P1 to be examined. Evaluate the similarity to ~ F '(N).
Next, the server 9 includes first parameter ratios F (A) to F (N) of the patient to be examined and second parameter ratios F ′ (A) of a plurality of other patients who are not the patient P1 to be examined. ˜F ′ (N) and their similarity are stored in the parameter information DB 7.

  Next, the server 9 has second parameter ratios F ′ (A) to F of other patients having the highest similarity to the first parameter ratios F (A) to F (N) of the patient to be examined. '(N) is searched, the various image data of the other patients having the highest similarity are searched, and the various image data of the patients are displayed on the display through the display unit 26.

Next, an image search operation in the server 9 will be described with reference to an image search flowchart shown in FIG.
First, in step S1, the target patient selection unit 20 sends a patient list in which a plurality of patients for selecting the patient P1 to be examined is described, for example, to the PC terminal 8, and the patient list is displayed on the PC terminal 8 display. Display a list. When the patient P1 to be examined is selected on the PC terminal 8, the target patient selection unit 20 sends an identification number for identifying the patient P1, for example, patient ID information, to the target patient data acquisition unit 21.

  Next, when the target patient data acquisition unit 21 receives patient ID information from the target patient selection unit 20 in step S2, for example, ultrasound image data of a patient having the patient ID information from the image information DB 6 according to the patient ID information. Analysis data and image-accompanying data are read out as various image data such as CT image data and MRI image data, and image measurement data obtained when the image data is acquired. At the same time, the target patient data acquisition unit 21 reads out physiological data and DNA data of the patient Q from the physiological examination information DB 13 according to the patient ID information.

The analysis data is acquired by analyzing ultrasonic image data such as a three-dimensional echocardiogram, CT image data obtained by photographing a blood flow image, MRI image data, and the like. The image accompanying data is attached to, for example, ultrasonic image data, CT image data, MRI image data, and the like. The image measurement data is, for example, image data itself such as ultrasound image data, CT image data, MRI image data, and the form, blood flow, and the like obtained from these image data. The image measurement data includes physical information such as CAG and PCI acquired from the image data, and measurement information such as restenosis rate, EF (ejection amount), strain, and pulse Doppler.
The physiological data is, for example, BNP, CTR, FBS, HbA1c, etc. acquired by the inspection of the physiological inspection apparatus 11.

  Next, in step S3, the parameter ratio calculation unit 22 is based on image measurement data and physiological examination data acquired based on image data of the patient P1, such as ultrasound image data, CT image data, and MRI image data. The above formula is calculated to calculate first time-series parameter ratios F (A) to F (N). The first parameter ratios F (A) to F (N) are respectively the ratio (standard ratio) to the parameter standard value std, the increase rate in the image measurement data and the physiological examination data, the image measurement data and the physiological examination data. And the time increase rate.

  Further, the parameter ratio calculation unit 22 calculates the above expression in the same manner as the first parameter ratios F (A) to F (N) in step S4, and based on various image data of other patients. The second parameter ratios F ′ (A) to F ′ (N) of other patients are obtained based on the image measurement data and physiological test data acquired in this manner. The second parameter ratios F ′ (A) to F ′ (N) are also the ratio (standard ratio) to the parameter standard value std, the increase rate in the image measurement data and the physiological examination data, the image measurement data and the physiological period, respectively. And the time increase rate in the inspection data.

  Next, in step S5, the database storage unit 23 calculates the first parameter ratios F (A) to F (N) of the patient calculated by the parameter ratio calculation unit 22 and the second parameter ratio F ′ (A). ˜F ′ (N) are stored in the parameter information DB 7. In this case, the database storage unit 23 associates with the patient ID information of the patient P1, and maintains a link with various image data of the patient P1, such as ultrasound image data, CT image data, MRI image data, and the like. Parameter ratios F (A) to F (N) are stored in the parameter information DB 7.

  Next, in step S6, the similar parameter comparison / extraction unit 24 receives the first parameter ratios F (A) to F (N) of the patient P1 to be examined calculated by the parameter ratio calculation unit 22, and the database Second parameter ratios F ′ (A), F ′ (B), F ′ (C),..., F ′ (N) of a plurality of other patients who are not the patient P1 to be examined are read from the storage unit 23. The similarity between the first parameter ratios F (A) to F (N) and the second parameter ratios F ′ (A) to F ′ (N) is evaluated.

For example, the similar parameter comparison / extraction unit 24 performs the first parameter ratios F (A) to F (N) of the patient P1 to be examined as shown in FIGS. The similarity between the second parameter ratios F ′ (A) to F ′ (N) of patients P2 to P4 is evaluated.
Here, in step S7, the similar parameter comparison / extraction unit 24 increases the standard ratio and increase of the first and second parameter ratios F (A) to F (N) and F ′ (A) to F ′ (N). The degree of coincidence of changes with the passage of time between the standard ratio and the time increase rate among the rate and the time increase rate is verified, and the similarity between the plurality of patients stored in the parameter information DB 7 and the patient P1 to be examined is determined. A high patient group, for example, a plurality of other patients P2 to P4, etc. are searched.

For example, when the similar parameter comparison / extraction unit 24 is in the state of the change portion H among the changes in the parameters of the patient P1 to be examined shown in FIG. 4A, the state of the change portion H is set to each of the other patients P2. It is determined which patient of P4 is most similar.
The similar parameter comparison / extraction unit 24 calculates the time lapse between the standard ratio and the time increase rate of the first and second parameter ratios F (A) to F (N) and F ′ (A) to F ′ (N). If the degree of coincidence of the change accompanying the change is verified, it is determined that the change in the parameter of the patient P1 is most similar to the change in the parameter of the patient P2 to be examined.
The similar parameter comparison / extraction unit 24 calculates the time between the standard ratio of the first and second parameter ratios F (A) to F (N) and F ′ (A) to F ′ (N) and the time increase rate. When verifying the degree of coincidence of changes with progress, the degree of coincidence may be set so that it cannot be determined to be the most similar if the degree of coincidence does not coincide, for example, 90% or more. This degree of coincidence can be variably set arbitrarily.

  As a result of the evaluation by the similar parameter comparison / extraction unit 24 in step S8, the data extraction unit 25 determines the first parameter ratio F (A) among the plurality of second parameter ratios F ′ (A) to F ′ (N). ) To F (N) are searched for various image data corresponding to another patient showing the second parameter ratios F ′ (A) to F ′ (N) having the highest similarity, for example, the patient P2. As a result of this search, the data extraction unit 25 displays and outputs at least various image data corresponding to another patient P2 showing the second parameter ratio having the highest similarity to the first parameter ratio on the display through the display unit 26. .

  As described above, according to the embodiment, the first parameter ratio F (A) in time series based on the image measurement data and the physiological examination data acquired based on the various image data of the patient P1 to be examined. ˜F (N) is calculated, and the first parameter ratios F ′ (A) in time series of the first parameter ratios F (A) to F (N) and a plurality of patients P2 to P4 different from the patient P1 are calculated. ) To F ′ (N), and the second parameter ratio F ′ (A) having the highest similarity to the first parameter ratios F (A) to F (N) of the patient P1 to be examined. ) To F ′ (N) are searched for various image data of another patient P2, and the various image data are displayed on the display.

  As a result, various image data of other patients P2 having a disease similar to the disease of the patient P1 to be examined, such as ultrasound image data, CT image data, MRI image data, etc. can be displayed on the display. The accuracy of diagnosis for the subject patient P1 can be increased, and a direction for determining a future treatment policy for the subject patient P1 can be determined.

  FIG. 7 shows a time-series relationship between changes in parameters A, B, and C of a patient and various image data of the patient. The patient parameters A, B, and C are, for example, BNP, CTR, and EF (ejection amount) acquired from image data, or ultrasonic image data, CT image data, Forms obtained from image data such as MRI image data, physical information such as blood flow, CAG, and PCI, measurement information such as restenosis rate, strain, and pulse Doppler, or physiological data such as BNP, CTR, FBS, and HbA1c DNA data.

  Further, as the various image data of the patient, for example, three-dimensional echocardiograms g1 and g2, a blood flow image h, and various measurement data i1 and i2 are acquired. Each of these image data is acquired by an examination when the patient goes to a hospital, for example, and the three-dimensional echocardiogram g1 and various measurement data i1 are obtained at the visit / examination f1, and the visit / examination f2 is obtained. A blood flow image h is acquired in FIG. 3, and a three-dimensional echocardiogram g2 and various measurement data i2 are acquired at the visit / inspection f3. Since these three-dimensional echocardiogram g1, various measurement data i1, blood flow image h, three-dimensional echocardiogram g2, and various measurement data i2 are each attached with a date and time acquired by photographing, they are arranged in time series. Is possible.

  Thus, since the changes in the parameters A, B, and C of each patient and the various image data of these patients have time-series information, the time-series first parameter ratio F of the patient P1 to be examined. If another patient showing the second parameter ratio F ′ (A) to F ′ (N) having the highest similarity to (A) to F (N), for example, the patient P2 can be searched, another patient P2 The three-dimensional echocardiogram g1, various measurement data i1, blood flow image h, three-dimensional echocardiogram g2, and various measurement data i2 can be retrieved and displayed on the display.

  Therefore, the accuracy of diagnosis for the patient P1 to be examined can be increased from the three-dimensional echocardiogram g1, various measurement data i1, blood flow image h, three-dimensional echocardiogram g2, and various measurement data i2 of the other patient P2. In addition, it is possible to make an orientation for determining a future treatment policy for the patient P1 to be examined.

  The similar parameter comparison / extraction unit 24 is a patient having a high degree of similarity with respect to the patient P1 to be examined from the plurality of patients P2 to P4 stored in the parameter information DB 7 as shown in FIGS. Since a group, for example, a plurality of other patients B, C, D, etc. are searched, not only the patient P1 having the highest similarity but also various image data of each of the other patients P2 to P4 are searched and the various images Data can be displayed on the display. Accordingly, it is possible to improve the accuracy of diagnosis for the patient P1 to be examined from various image data of these patients P2 to P4, and to give directions for determining future treatment policies for the patient P1 to be examined. Is possible.

In addition, since various image data of another patient having the highest similarity to the patient P1 to be examined can be searched, medical record information such as medical records of other patients, records of test results, records of medications, and the like can be retrieved. Can be searched and displayed on the display. However, by referring to the chart information of another patient, it can be used as a reference for determining the treatment method for the patient P1 to be examined.
In addition, diagnosis results, treatment methods, treatment results, and prognosis of another patient P2 can be presented from various image data and medical record information of the other patient P2. Further, the second parameter ratios F ′ (A) to F ′ () of another patient P2 having the highest similarity to the time-series first parameter ratios F (A) to F (N) of the patient P1 to be examined. N) can also be displayed.

Note that the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.
The first parameter ratio F (A) has a ratio (standard ratio) to the parameter standard value std. However, the parameter standard value std may have a range composed of an upper limit value and a lower limit value.

1 is a configuration diagram showing an in-hospital image search system to which a first embodiment of an image search device according to the present invention is applied. The functional block diagram of the server in the apparatus. The figure which shows the specific example of calculation of the 1st parameter ratio by the parameter ratio calculation part in the apparatus. The figure which shows the search effect | action of other patients with high similarity with respect to the patient of the test object by the similar parameter comparison / extraction part in the same apparatus. FIG. 6 is a data flow chart showing from calculation of a first parameter ratio of a patient to be examined in the apparatus to retrieval of image data of a patient having the highest similarity. FIG. The image search flowchart in the same apparatus. The figure which shows the time-sequential relationship between the change of the parameter of the patient acquired by the same apparatus, and the various image data of the said patient.

Explanation of symbols

  1: Radiation department, 2: Physiological examination department, 3: Medical department, 4: Ultrasound device, 5: CT device, 6: Image information database (image information DB), 7: Parameter information database (parameter information DB), 8: personal computer (PC) and its terminal (PC terminal), 9: server, 10: network, 11: physiological test apparatus, 12: DNA test apparatus, 13: physiological test information database (physiological test information DB), 14: Personal computer (PC) and its terminal (PC terminal), 15: Medical record information database (medical record information DB), 16: Personal computer (PC) and its terminal (PC terminal), 20: Target patient selection unit, 21: Target patient Data acquisition unit, 22: parameter ratio calculation unit, 23: database storage unit, 24: similar parameter comparison / extraction , 25: data extraction unit, 26: display unit.

Claims (7)

A parameter ratio calculator that calculates at least a first parameter ratio in time series based on time-series image measurement data acquired based on image data to be examined and time-series physiological examination data of the examination object;
When a plurality of which are calculated based on the physiological test data of the time series of the image measurement data with said another each inspection target time series is obtained based on the image data of another plurality of inspected and the inspection target A similarity evaluation unit that evaluates the similarity of changes over time between a second parameter ratio of a sequence and the first parameter ratio;
As a result of the evaluation by the similarity evaluation unit, the second parameter ratio having the highest similarity to the change with time of the first parameter ratio among the plurality of second parameter ratio changes is determined; An image search unit for searching at least the image data corresponding to the other inspection object indicating the determined second parameter ratio ;
An image search apparatus comprising: The parameter ratio calculator, the image and the ratio measurement data with respect to a preset parameter standard value in said physiological test data of the time series, the increase rate in the the sanitary inspection data and the image measurement data of the time series The image search device according to claim 1, wherein the first parameter ratio is calculated from a time increase rate in the image measurement data and the physiological examination data. As a result of the search, the image search unit obtains at least the image data corresponding to the other inspection object indicating the second parameter ratio having the highest similarity to the change with time of the first parameter ratio. 2. The image search apparatus according to claim 1, wherein the image search apparatus performs display output on a display. A first database for storing each image data in time series of each inspection object;
A second database for storing each physiological examination data in time series of each examination object;
The first and the first of the time series calculated by the parameter ratio calculation unit based on the image data stored in the first database and the physiological examination data stored in the second database. A parameter database for storing the parameter ratio of 2;
With
The similarity evaluation unit evaluates the similarity of the change over time of the second parameter ratio and the first parameter ratio stored in the parameter database;
The image search unit corresponds to the second inspection target indicating the second parameter ratio having the highest similarity to the change with the passage of time of the first parameter ratio as a result of the evaluation by the similarity evaluation unit. Search for at least the image data and display it on a display;
The image search apparatus according to claim 1, wherein: The parameter ratio calculation unit uses at least one of or both of analysis data obtained when the image data to be inspected is acquired and incidental data attached to the image data. The image search apparatus according to claim 1, wherein: Calculating a first parameter ratio of time series based on the physiological test data of the time series of the image measurement data and the inspection target time series is obtained based on the image data of the inspection object,
A plurality of time-series images calculated based on time-series image measurement data acquired based on image data of a test object different from the test object and time-series physiological test data of the other test object. The similarity of changes over time between the parameter ratio of 2 and the first parameter ratio,
As a result of the evaluation, the second parameter ratio that is most similar to the change with time of the first parameter ratio among the plurality of changes in the second parameter ratio is determined, and the determined second parameter ratio is determined. Searching for at least the image data corresponding to the other inspection object having a parameter ratio of 2 ;
An image search method characterized by that.   The first parameter ratio includes a ratio of the image measurement data and the physiological examination data to a preset parameter standard value, an increase rate in the image measurement data and the physiological examination data, the image measurement data, The image search method according to claim 6, comprising a rate of time increase in the physiological test data.

Images