JP2023022460A - Information processing device, tonometry system and tonometry method - Google Patents

Abstract

To measure intraocular pressure of a subject at high accuracy by a learned model constructed based on a data set of pressing data from a pressure-sensitive sensor and intraocular pressure data.SOLUTION: An information processing device according to the present invention includes a detection part for detecting pressing force when an eyeball of a subject is pressed, an estimation part for estimating intraocular pressure of the subject by inputting data from the detection part into a learned model that has preliminarily machine-learned a relation between the output data of the detection part and a measurement value relating to the intraocular pressure, and an output part for outputting data relating to the intraocular pressure.SELECTED DRAWING: Figure 1

Description

The present invention relates to, for example, a technique for measuring intraocular pressure of a subject using a trained model.

Attempts have been made in the medical field to automate diagnosis by skilled doctors by utilizing artificial intelligence.

Here, for example, in Patent Document 1, a patient information acquisition unit that acquires information about a new patient's symptom, and the acquired information about the new patient's symptom are combined with information about the past patient's symptom and a past by a Kampo specialist. Using the learned model created based on the diagnostic result information for the patient, the artificial intelligence analysis unit that analyzes by artificial intelligence and derives the analysis results, and the analysis results that are derived from the analysis, the new patient's An automatic diagnosis support method is disclosed that includes a diagnosis output unit that outputs at least one of probable disease, optimal treatment, and drug information to be prescribed.

Japanese Patent Application Laid-Open No. 2021-47504

However, in Patent Document 1, by inputting data from a detection device equipped with a pressure-sensitive sensor or the like to a trained model, the eye pressure can be estimated as if a skilled ophthalmologist were estimating the intraocular pressure with the palm of the finger. Measuring intraocular pressure with high precision is not disclosed.

The present invention has been made in view of such problems, and the object thereof is to provide a subject with a trained model constructed based on a data set of pressure data and intraocular pressure data from a pressure sensor. To measure the intraocular pressure of a person with high precision.

In order to solve the above problems, an information processing apparatus according to an aspect of the present invention includes a detection unit that detects a pressing force output from a pressure-sensitive sensor when an eyeball of a subject is pressed; An estimating unit for estimating the intraocular pressure of the subject by inputting the output data from the detecting unit to a trained model that has undergone machine learning in advance on the relationship between the output data and the measured value related to the intraocular pressure and an output unit for outputting the estimation result.

An information processing apparatus according to another aspect of the present invention includes a detection unit that detects a pressing force output from a pressure-sensitive sensor when an eyeball of a subject is pressed; an estimating unit for estimating the intraocular pressure of the subject corresponding to the output data from the detecting unit by referring to a table in which measured values are associated with each other; and an output unit for outputting the result of the estimation. rice field.

An intraocular pressure measurement system according to another aspect of the present invention includes a detection device having a pressure sensor, and an information processing device that performs information processing based on the output of the detection device, the information processing device comprising:
estimating the intraocular pressure of the subject by inputting the output data from the detection device to a trained model in which the relationship between the output data of the detection device and the measured value related to the intraocular pressure is machine-learned in advance; and an output unit for outputting the result of the estimation.

According to still another aspect of the present invention, there is provided an intraocular pressure measuring method in which a detecting unit detects a pressing force when an eyeball of a subject is pressed, and an estimating unit detects the output data of the detecting unit and the intraocular pressure. The intraocular pressure of the subject is estimated by inputting the output data from the detection unit to a trained model that has undergone machine learning in advance on the relationship with the measured value, and the output unit outputs the result of the estimation. Output.

According to the present invention, it is possible to provide a technique for measuring the intraocular pressure of a subject with high accuracy using a trained model constructed based on a data set of pressure data and intraocular pressure data from a pressure sensor.

1 is a configuration diagram of an intraocular pressure measurement system according to an embodiment of the present invention; FIG. It is a flowchart which shows the flow of the process by the same system.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 shows the configuration of an intraocular pressure measurement system according to an embodiment of the present invention.

As shown in the figure, the intraocular pressure measurement system comprises an information processing device 1 such as a personal computer, and a detection device 2 having a pressure sensor 2a.

The information processing apparatus 1 has a control section 11 , a communication section 12 , a storage section 13 , an input/output I/F 17 , an operation input section 18 and a display section 19 . Each unit is communicably connected via a bus line. The communication unit 12 is realized by, for example, a network interface card (NIC) or the like, and is connected to a communication network such as the Internet by wire or wirelessly, and is a communication interface for communicating with external devices. is.

The storage unit 13 is implemented by, for example, a semiconductor memory device such as a RAM (Random Access Memory) or a flash memory, a hard disk drive (HDD), an optical disk device, or the like. The storage unit 13 stores control programs executed by the control unit 11 . Furthermore, the storage unit 13 has a user information storage unit 14 , a learned model storage unit 15 , and an estimation result storage unit 16 .

The user information storage unit 14 stores attribute information, anamnesis, biometric information, etc. of a patient who is a user in association with a user ID. The learned model storage unit 15 stores learned models used for estimation by the estimation unit 11b. The learned model storage unit 15 may also store a table that associates the output data of the detection device 2 with the intraocular pressure evaluated from the output signal. The estimation result storage unit 16 stores each estimation result obtained by the estimation unit 11b.

The operation input unit 18 is implemented by a mouse, a keyboard, or the like, and receives various operation inputs from the user. The display unit 19 is a display device such as a liquid crystal display. When the information processing device 1 is connected wirelessly or by wire to a smart phone or a tablet terminal, the touch panel can be used as an input I/F of the device.

The control unit 11 is realized by a CPU (Central Processing Unit), an MPU (Micro Processing Unit), or the like, and by executing a program stored in the storage unit 13, for example, an input/output control unit 11a, an estimation unit 11b, It functions as a learning/updating unit 11c, a display control unit 11d, and a transmission/reception control unit 11e. The control unit 11 may be configured by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field-Programmable Gate Array).

In such a configuration, the input/output control section 11a receives and processes the output signal of the pressure sensor 2a of the detection device 2 via the input/output I/F 17 .

The estimation unit 11b estimates the intraocular pressure by inputting data related to the pressing force received by the input/output control unit 11a into the learned model. For example, the intraocular pressure is estimated by inputting data related to the pressing force when the subject's eyeball is pressed against the trained model. A trained model means a machine learning model that has been trained based on a machine learning algorithm such as a support vector machine. Here, the learning data is composed of pairs of input data and output data (correct answers). Although it is used as learning data, it is needless to say that it is not limited to this. In addition, by referring to a table that associates the output data of the detection device 2 and the intraocular pressure evaluated from the output signal, which is stored in advance in the learned model storage unit 15, the output data from the detection device 2 is referenced. You may estimate the intraocular pressure of the subject who carries out.

The learning/updating unit 11c periodically performs machine learning based on the learning data set input via the input/output I/F 17 or the communication unit 12, calculates a learned model, and stores the data in the learned model storage unit 15. Update the trained model to the latest one as appropriate. The display control unit 11d controls the display unit 19 to display the estimation result by the estimation unit 11b. The transmission/reception control unit 11e controls a remote external device to transmit the result of estimation by the estimation unit 11b as display data.

Here, the detection device 2 has a sheet-like pressure-sensitive sensor 2a with a size of 1 mm in a square area of 1 to 2 cm. ing. Then, the distribution of data related to the pressing force of the pressure sensor 2a at, for example, 100 to 200 points when the eye is touched with a constant pressure is collected (measured and recorded as a pressure unit such as mmHg, and acquired in a format such as CSV Possible). Then, as training data, the intraocular pressure of the same patient is measured with a tonometer. For example, using the distribution of data related to pressing force on a scale of several hundred people, an algorithm is developed by machine learning using the training data using the tonometer. to create As a result, the learned model described above is generated by the learning/updating unit 11c.

Therefore, if distribution data of pressing force is collected using this system for a subject whose intraocular pressure is unknown, it is possible to estimate the intraocular pressure under estimation by the estimation unit 11b and display it on the display unit 19. Become. In the measurement of the subject's eyeball by the detection device 2, the distribution data of the pressing force is collected not once but several times, and the intraocular pressure is measured a plurality of times, and the intraocular pressure is specified by the average value or the like. You may make it display. It goes without saying that the functions of the detection device may be implemented in the information processing device 1 and may be part of the information processing device 1 .

Hereinafter, the processing procedure of eye examination by the intraocular pressure measurement system according to the embodiment of the present invention will be described in detail with reference to the flowchart of FIG.

The input/output control unit 11a acquires and processes data related to the pressing force from the detection device 2 via the input/output I/F 17 (S1), and the estimation unit 11b estimates the intraocular pressure (S2). That is, the estimation unit 11b estimates the intraocular pressure by inputting the data related to the pressing force received by the input/output control unit 11a into the learned model. For example, the intraocular pressure is estimated by inputting data related to the pressing force when the subject's eyeball is pressed against the trained model. Subsequently, the display control unit 11d generates display data related to the estimation result by the estimation unit 11b (S3), displays the estimation result on the display unit 19 (S4), and ends the process.

As described above, according to the embodiment of the present invention, the intraocular pressure of the subject is measured with high accuracy using a trained model constructed based on a data set of pressure data and intraocular pressure data from the pressure sensor. becomes possible.

Although the embodiments of the present invention have been described above, it goes without saying that the present invention is not limited to these and that various improvements and modifications are possible without departing from the scope of the invention.

For example, the estimation result by the estimation unit may be transmitted to an external terminal such as a doctor to perform remote diagnosis and treatment.

Due to the lack of ophthalmologists and their uneven distribution in urban areas, it is often difficult to access ophthalmologists. There are currently a certain number of situations. According to the present invention, this technology, which enables highly accurate measurement of intraocular pressure using a pressure-sensitive sensor, will enable a wide range of patients to have the opportunity to have their intraocular pressure measured more frequently and at a lower cost than at present. be able to.

Reference Signs List 1 information processing device 2 detection device 2a pressure sensor 11 control unit 11a input/output control unit 11b estimation unit 11c learning/updating unit 11d display control unit 11e transmission/reception Control unit 12 Communication unit 13 Storage unit 14 User information storage unit 15 Learned model storage unit 16 Estimation result storage unit 17 Input/output I/F 18 Operation input unit 19 … display.

Claims (7)

a detection unit that detects the pressing force output from the pressure sensor when the subject's eyeball is pressed;
The intraocular pressure of the subject is determined by inputting the output data from the detection unit to a trained model in which the relationship between the output data from the detection unit and the measured value related to the intraocular pressure is machine-learned in advance. an estimating unit for estimating;
an output unit that outputs the result of the estimation;
Information processing device with a detection unit that detects the pressing force output from the pressure sensor when the subject's eyeball is pressed;
an estimating unit for estimating the intraocular pressure of the subject corresponding to the output data from the detecting unit by referring to a table that associates the output data of the detecting unit with measured values related to intraocular pressure;
an output unit that outputs the result of the estimation;
Information processing device with The output unit is a display unit,
3. The information processing apparatus according to claim 1, further comprising a display control unit that causes the display unit to display data on the intraocular pressure. The information processing apparatus according to any one of claims 1 to 3, wherein the estimating unit refers to biological information of the subject when estimating. a detection device comprising a pressure sensor;
An information processing device that performs information processing based on the output of the detection device,
The information processing device is
estimating the intraocular pressure of the subject by inputting the output data from the detection device to a trained model in which the relationship between the output data of the detection device and the measured value related to the intraocular pressure is machine-learned in advance; an estimator that
an output unit that outputs the result of the estimation, and an intraocular pressure measurement system. The output unit is a display unit,
The intraocular pressure measurement system according to claim 5, further comprising a display control section that causes the display section to display data related to the intraocular pressure. The detection unit detects the pressing force when pressing the subject's eyeball,
The estimating unit inputs the output data from the detecting unit to a learned model in which the relationship between the output data of the detecting unit and the measured value related to intraocular pressure is machine-learned in advance, so that the subject's Estimate intraocular pressure,
The intraocular pressure measurement method, wherein the output unit outputs the result of the estimation.

Images