Making measurements in images

Patent metadata

Field	Value
Publication number	US-12482123-B2
Application number	US-202017785071-A
Country	US
Kind code	B2
Filing date	Dec 16, 2020
Priority date	Dec 19, 2019
Publication date	Nov 25, 2025
Grant date	Nov 25, 2025

Abstract

A computer implemented method of making a measurement associated with a feature of interest in an image. The method comprises using ( 302 ) a model trained using a machine learning process to take the image as input and predict a pair of points between which to make the measurement of the feature of interest in the image. The method then comprises determining ( 304 ) the measurement, based on the predicted pair of points.

First claim

The invention claimed is:

1 . A computer implemented method of making a measurement associated with a feature of interest in a image, the method comprising:
using a model trained using a machine learning process to take the image as input and predict a pair of points between which to make the measurement of the feature of interest in the image, wherein the model provides as output one or more confidence maps indicating confidence values associated with different placements of the pair of points; and determining the measurement, based on the predicted pair of points.

2 . The method of claim 1 , the model having been trained using training data comprising:
i) example images; and ii) for each example image, a ground truth pair of points indicating appropriate locations between which to make the measurement of the feature of interest in the example image.

3 . The method of claim 1 , wherein the image comprises a medical image and the pair of points are positioned so as to satisfy a clinical requirement associated with the measurement.

4 . The method of claim 3 , wherein the model further takes as input information relating to a clinical context of the measurement and/or the medical image.

5 . The method of claim 4 , wherein the information relating to the clinical context of the measurement and/or the medical image is obtained from meta-data stored in the image.

6 . The method of claim 1 , wherein the pair of points comprise:
end points of an electronic caliper; a first point representing the centre of a circle or sphere, and a second point representing a boundary of the circle or sphere; or a first point representing one corner of a rectangular box, and a second point representing a second corner of the rectangular box.

7 . The method of claim 1 , wherein the method further comprises:
performing a segmentation of the image; and providing a segment from the segmentation as a further input to the model.

8 . The method of claim 1 , wherein the pair of points are determined using keypoint detection.

9 . The method of claim 1 , wherein the image comprises a medical image, further wherein the pair of points are positioned so as to satisfy a clinical requirement associated with the measurement, and wherein the method further comprises:
performing a segmentation of the medical image; and providing a segment from the segmentation as a further input to the model.

10 . A method of training a model for use in making a measurement associated with a feature of interest in a image, the method comprising:
providing training data to the model, the training data comprising: i) example images; and ii) for each example image, a ground truth pair of points indicating appropriate locations between which to make the measurement of the feature of interest in the example image; and training the model to predict the associated ground truth pair of points for each example image and to provide as output one or more confidence maps indicating confidence values associated with different placements of the pair of points.

11 . The method of claim 10 , wherein the ground truth pair of points for each example image are obtained from meta data of the associated example image.

12 . The method of claim 10 , wherein the ground truth pair of points for each example image are obtained by:
displaying the example image to a user; and receiving the ground truth pair of points from an input provided by the user.

13 . The method of claim 12 , wherein the training data further comprises:
for each example image, a display setting associated with displaying the example image to the user; and wherein the display setting is further provided as input to the model during the training.

14 . A system for making a measurement associated with a feature of interest in a image, the system comprising:
a memory comprising instruction data representing a set of instructions; and a processor configured to communicate with the memory and to execute the set of instructions, wherein the set of instructions, when executed by the processor, cause the processor to:
use a model trained using a machine learning process to take the image as input and predict a pair of points between which to make the measurement of the feature of interest in the image, and wherein the model provides as output one or more confidence maps indicating confidence values associated with different placements of the pair of points; and
determine the measurement, based on the predicted pair of points.

15 . The system of claim 14 , wherein the image comprises a medical image, further wherein the pair of points are positioned so as to satisfy a clinical requirement associated with the measurement, and wherein the set of instructions, when executed by the processor, further cause the processor to:
perform a segmentation of the medical image; and provide a segment from the segmentation as a further input to the model.

16 . The system of claim 14 , the model having been trained using training data comprising:
i) example images; and ii) for each example image, a ground truth pair of points indicating appropriate locations between which to make the measurement of the feature of interest in the example image.

17 . The system of claim 14 , wherein the model further takes as input information relating to a clinical context of the measurement and/or the image.

18 . The system of claim 17 , wherein the information relating to the clinical context of the measurement and/or the image is obtained from meta-data stored in the image.

19 . A non-transitory computer readable medium storing instructions that, when executed by one or more processors, cause the one or more processors to:
use a model trained using a machine learning process to take an image as input and predict a pair of points between which to make a measurement of a feature of interest in the image; determine the measurement, based on the predicted pair of points; wherein the model is trained by:
providing training data to the model, the training data comprising: i) example images; and ii) for each example image, a ground truth pair of points indicating appropriate locations between which to make the measurement of the feature of interest in the example image; and
training the model to predict the associated ground truth pair of points for each example image and to provide as output one or more confidence maps indicating confidence values associated with different placements of the pair of points.

20 . The non-transitory computer readable medium of claim 19 , wherein the ground truth pair of points for each example image are obtained from meta data of the associated example image.

21 . The non-transitory computer readable medium of claim 19 , wherein the ground truth pair of points for each example image are obtained by:
displaying the example image to a user; and receiving the ground truth pair of points from an input provided by the user.

Assignees

• Koninklijke Philips Nv

Inventors

• Wiemker Rafael
• Brosch Tom
• Deshpande Hrishikesh Narayanrao
• Goossen André
• Harder Tim Philipp
• Saalbach Axel

Classifications

• G06T2207/20081

  Training; Learning · CPC title

• G06T7/0012

  Biomedical image inspection · CPC title

• G06T7/11

  Region-based segmentation · CPC title

• G16H50/70

  for mining of medical data, e.g. analysing previous cases of other patients · CPC title

• G16H50/20

  for computer-aided diagnosis, e.g. based on medical expert systems · CPC title