Recording medium, case data generation support system, and case data generation support method

Patent metadata

Field	Value
Publication number	US-10713539-B2
Application number	US-201815987023-A
Country	US
Kind code	B2
Filing date	May 23, 2018
Priority date	Nov 27, 2015
Publication date	Jul 14, 2020
Grant date	Jul 14, 2020

Abstract

A non-transitory recording medium stores a program causing a computer to execute a process including determining a reference value on a decision surface classifying a group of image data represented by a plurality of types of feature quantities into two classes in a feature quantity space, based on a feature quantity of given image data that is closest to the decision surface and is classified into one of the classes of the group of image data in the feature quantity space, plotting and displaying pieces of image data of the group of image data along at least one axis indicating a degree of separation of each of the pieces of image data relative to the reference value, modifying, upon receiving a single piece of image data identified from the plotted and displayed pieces of image data, the decision surface based on the identified single piece of image data.

First claim

What is claimed is:

1. A non-transitory recording medium having stored therein a program for causing a computer to execute a case data generation support process comprising:
determining a reference value on a decision surface used for classifying a group of image data represented by a plurality of types of feature quantities into first and second classes in a feature quantity space, based on a feature quantity of a reference image data piece, the reference image data piece being closest to the decision surface in the feature quantity space and determined as belonging to the first class;
plotting and displaying in a two-dimensional plane pieces of image data extracted from the group of image data, each of the plotted and displayed pieces of image data being distributed along each of two orthogonal axes of the two-dimensional plane in accordance with a degree of separation calculated for each of the pieces of image data relative to the reference value; and
modifying, upon receiving a single piece of image data identified from among the plotted and displayed pieces of image data in the two-dimensional plane, the decision surface based on the identified single piece of image data;
wherein the plotted and displayed pieces of image data in the two-dimensional plane include:
pieces of image data to be determined as belonging to the first class plotted in a first region on an opposite side of the decision surface when viewed from a position of the reference image data piece; and
pieces of image data to be determined as belonging to the second class plotted in a second region on the opposite side of the decision surface, the second region being visually distinct from the first region.

2. The non-transitory recording medium according to claim 1 , wherein the degree of separation calculated for each of the pieces of image data indicates a distance of a position of the piece of image data from the position of the reference image data piece in the two-dimensional plane.

3. The non-transitory recording medium according to claim 1 , wherein, in response to a possibility of one of the pieces of image data being overlapped with another of the pieces of image data when plotted and displayed along one axis indicating the degree of separation of each of the pieces of image data relative to the reference value, the one of the pieces of image data is plotted and displayed by being shifted in a direction orthogonal to the one axis.

4. The non-transitory recording medium according to claim 3 , wherein the degree of separation of each of the pieces of image data relative to the reference value is a degree of separation in a direction approximately orthogonal to the decision surface.

5. The non-transitory recording medium according to claim 3 , wherein the degree of separation of each of the pieces of image data relative to the reference value is a degree of separation in a direction along the decision surface.

6. A case data generation support system comprising a computer configured to execute a process including:
determining a reference value on a decision surface used for classifying a group of image data represented by a plurality of types of feature quantities into first and second classes in a feature quantity space, based on a feature quantity of a reference image data piece, the reference image data piece being closest to the decision surface in the feature quantity space and determined as belonging to the first class;
plotting and displaying in a two-dimensional plane pieces of image data extracted from the group of image data, each of the plotted and displayed pieces of image data being distributed along each of two orthogonal axes of the two-dimensional plane in accordance with a degree of separation calculated for each of the pieces of image data relative to the reference value; and
modifying, upon receiving a single piece of image data identified from among the plotted and displayed pieces of image data in the two-dimensional plane, the decision surface based on the identified single piece of image data;
wherein the plotted and displayed pieces of image data in the two-dimensional plane include:
pieces of image data to be determined as belonging to the first class plotted in a first region on an opposite side of the decision surface when viewed from a position of the reference image data piece; and
pieces of image data to be determined as belonging to the second class plotted in a second region on the opposite side of the decision surface, the second region being visually distinct from the first region.

7. A case data generation support method executed by a computer, the method comprising:
determining a reference value on a decision surface used for classifying a group of image data represented by a plurality of types of feature quantities into first and second classes in a feature quantity space, based on a feature quantity of a reference image data piece, the reference image data piece being closest to the decision surface in the feature quantity space and determined as belonging to the first class;
plotting and displaying in a two-dimensional plane pieces of image data extracted from the group of image data, each of the plotted and displayed pieces of image data being distributed along each of two orthogonal axes of the two-dimensional plane in accordance with a degree of separation calculated for each of the pieces of image data relative to the reference value; and
modifying, upon receiving a single piece of image data identified from among the plotted and displayed pieces of image data in the two-dimensional plane, the decision surface based on the identified single piece of image data;
wherein the plotted and displayed pieces of image data in the two-dimensional plane include:
pieces of image data to be determined as belonging to the first class plotted in a first region on an opposite side of the decision surface when viewed from a position of the reference image data piece; and
pieces of image data to be determined as belonging to the second class plotted in a second region on the opposite side of the decision surface, the second region being visually distinct from the first region.

Assignees

• Fujitsu Ltd

Inventors

• Takebe Hiroaki
• Ishihara Masaki
• Sugimura Masahiko
• Endo Susumu
• Baba Takayuki
• Uehara Yusuke

Classifications

• A61B6/5217 Primary

  extracting a diagnostic or physiological parameter from medical diagnostic data · CPC title

• G06V10/7788

  the supervisor being a human, e.g. interactive learning with a human teacher · CPC title

• G06V10/7784

  based on feedback from supervisors · CPC title

• G06V10/764

  using classification, e.g. of video objects · CPC title

• G06F18/2411

  based on the proximity to a decision surface, e.g. support vector machines · CPC title