Ophthalmic image processing device, ophthalmic image processing program, and ophthalmic image processing system
A control section of an ophthalmologic image processing apparatus acquires an ophthalmologic image captured by an ophthalmologic image capturing apparatus. The control section acquires, for the acquired ophthalmologic image, evaluation information indicating an appropriateness for acquiring medical data including at least any of an analysis result relating to a disease of a subject eye, an analysis result relating to the structure of the subject eye, and an image converted from the acquired ophthalmologic image. The control section acquires the medical data based on the acquired ophthalmologic image. The control section changes a medical data acquisition method according to whether or not the evaluation information on the ophthalmologic image satisfies a criterion.
1. An ophthalmologic image processing apparatus for processing an ophthalmologic image as an image of a tissue of a subject eye, the ophthalmologic image processing apparatus comprising a control section configured to execute:
an image acquisition step of acquiring an ophthalmologic image captured by an ophthalmologic image capturing apparatus,
an evaluation information acquisition step of acquiring, for the ophthalmologic image acquired at the image acquisition step, evaluation information indicating an appropriateness for acquiring medical data including at least any of an analysis result relating to a disease of the subject eye, an analysis result relating to a structure of the subject eye, and an image converted from the acquired ophthalmologic image, and
a medical data acquisition step of acquiring the medical data based on the ophthalmologic image acquired at the image acquisition step, wherein
at the medical data acquisition step, the control section changes a medical data acquisition method according to whether or not the evaluation information acquired at the evaluation information acquisition step satisfies a criterion,
at the medical data acquisition step, the control section acquires the medical data in such a manner that the ophthalmologic image is input to a mathematical model trained according to a machine learning algorithm,
the mathematical model is trained using multiple training data sets of the ophthalmologic image captured by the ophthalmologic image capturing apparatus as input training data and the medical data corresponding to the input training data as output training data, and
the criterion is set based on multiple pieces of the input training data used for training of the mathematical model.
2. The ophthalmologic image processing apparatus according to claim 1 , wherein
at the medical data acquisition step, the control section acquires the medical data in such a manner that information relating to the evaluation information on the ophthalmologic image is input to the mathematical model together with the ophthalmologic image.
3. The ophthalmologic image processing apparatus according to claim 1 , wherein
at the medical data acquisition step, the control section
executes, for the ophthalmologic image, evaluation information improvement processing of causing the evaluation information on the ophthalmologic image to approach the criterion in a case where the evaluation information on the ophthalmologic image does not satisfy the criterion, and
acquires the medical data based on the ophthalmologic image subjected to the evaluation information improvement processing.
4. The ophthalmologic image processing apparatus according to claim 1 , wherein
at the medical data acquisition step, the control section
outputs an instruction for additionally capturing an image of a location identical to that of the ophthalmologic image in a case where the evaluation information on the ophthalmologic image does not satisfy the criterion, and
acquires the medical data based on the additionally-captured ophthalmologic image.
5. The ophthalmologic image processing apparatus according to claim 1 , wherein
at the medical data acquisition step, the control section
outputs an instruction for additionally capturing an image of a location identical to that of the ophthalmologic image in a case where the evaluation information on the ophthalmologic image does not satisfy the criterion, and
acquires the medical data based on an ophthalmologic image obtained by addition averaging processing of the additionally-captured ophthalmologic image.
6. The ophthalmologic image processing apparatus according to claim 1 , wherein
the control section
acquires multiple ophthalmologic images of the identical tissue of the identical subject eye at the image acquisition step, and
acquires the medical data based on at least one, of which evaluation information satisfies the criterion, of the multiple ophthalmologic images.
7. The ophthalmologic image processing apparatus according to claim 1 , wherein at the medical data acquisition step, the control section outputs warning information for a user in a case where the evaluation information on the ophthalmologic image does not satisfy the criterion.
8. The ophthalmologic image processing apparatus according to claim 1 , wherein
at the medical data acquisition step, the control section:
determines whether or not the evaluation information satisfies the criterion for each of multiple partial regions of the ophthalmologic image, and
changes the medical data acquisition method for each partial region according to whether or not the evaluation information for that partial region satisfies the criterion.
9. A storage device storing an ophthalmologic image processing program executed by an ophthalmologic image processing apparatus for processing an ophthalmologic image as an image of a tissue of a subject eye, an execution of the ophthalmologic image processing program by a control section of the ophthalmologic image processing apparatus causing the ophthalmologic image processing apparatus to execute:
an image acquisition step of acquiring an ophthalmologic image captured by an ophthalmologic image capturing apparatus,
an evaluation information acquisition step of acquiring, for the ophthalmologic image acquired at the image acquisition step, evaluation information indicating an appropriateness for acquiring medical data including at least any of an analysis result relating to a disease of the subject eye, an analysis result relating to a structure of the subject eye, and an image converted from the acquired ophthalmologic image, and
a medical data acquisition step of acquiring the medical data based on the ophthalmologic image acquired at the image acquisition step, wherein
at the medical data acquisition step, a medical data acquisition method is changed according to whether or not the evaluation information acquired at the evaluation information acquisition step satisfies a criterion,
at the medical data acquisition step, the control section acquires the medical data in such a manner that the ophthalmologic image is input to a mathematical model trained according to a machine learning algorithm,
the mathematical model is trained using multiple training data sets of the ophthalmologic image captured by the ophthalmologic image capturing apparatus as input training data and the medical data corresponding to the input training data as output training data, and
the criterion is set based on multiple pieces of the input training data used for training of the mathematical model.
10. An ophthalmologic image processing system for processing an ophthalmologic image as an image of a tissue of a subject eye, comprising:
an OCT apparatus configured to receive interfering light of reference light and reflected light of measurement light emitted to the tissue of the subject eye to capture an ophthalmologic image as a laminagram image of the tissue, a control section of the ophthalmologic image processing system being configured to execute:
an image acquisition step of acquiring an ophthalmologic image captured by the OCT apparatus,
an evaluation information acquisition step of acquiring, for the ophthalmologic image acquired at the image acquisition step, evaluation information indicating an appropriateness for acquiring medical data including at least any of an analysis result relating to a disease of the subject eye, an analysis result relating to a structure of the subject eye, and an image converted from the acquired ophthalmologic image,
an addition step of executing, in a case where the evaluation information acquired at the evaluation information acquisition step does not satisfy a criterion, processing of additionally capturing an image of a location identical to that of the ophthalmologic image and addition averaging processing of the image captured in addition to the ophthalmologic image, and
a medical data acquisition step of acquiring the medical data in such a manner that the ophthalmologic image subjected to the addition averaging processing is input to a mathematical model trained according to a machine learning algorithm, wherein
at the medical data acquisition step, the control section acquires the medical data in such a manner that the ophthalmologic image is input to a mathematical model trained according to a machine learning algorithm,
the mathematical model is trained using multiple training data sets of the ophthalmologic image captured by the ophthalmologic image capturing apparatus as input training data and the medical data corresponding to the input training data as output training data, and
the criterion is set based on multiple pieces of the input training data used for training of the mathematical model.