Methods and systems for characterizing anatomical features in medical images

Patent metadata

Field	Value
Publication number	US-11227683-B2
Application number	US-202016750931-A
Country	US
Kind code	B2
Filing date	Jan 23, 2020
Priority date	Jan 23, 2020
Publication date	Jan 18, 2022
Grant date	Jan 18, 2022

Abstract

Various methods and systems are provided for determining and characterizing features of an anatomical structure from a medical image. In one example, a method comprises acquiring a plurality of medical images over time during an exam, registering the segmented anatomical structure between the plurality of medical images, segmenting an anatomical structure in a one of the plurality of medical images after registering the plurality of medical images, creating and characterizing a reference region of interest (ROI) in each of the plurality of medical images, determining characteristics of the anatomical structure by tracking pixel values of the segmented and registered anatomical structure over time, and outputting the determined characteristics on a display device.

First claim

The invention claimed is:

1. A method, comprising:
acquiring a plurality of medical images over time during an exam;
registering the plurality of medical images;
segmenting an anatomical structure in one of the plurality of medical images after registering the plurality of medical images;
identifying a boundary of a cancerous lesion of the anatomical structure;
creating and characterizing a reference region of interest (ROI) outside of the lesion and inside of the anatomical structure in each of the plurality of medical images;
determining characteristics of the anatomical structure by tracking pixel values of the segmented anatomical structure over time;
sub-segmenting regions of viable tissue within the lesion by generating a temporal profile of a change in brightness in each pixel of the lesion compared with the reference ROI across the plurality of medical images, the sub-segmented regions including a peripheral enhancement region; and
outputting the determined characteristics on a display device.

2. The method of claim 1 , wherein registering the plurality of medical images includes applying rigid and non-rigid registrations to a source image and a target image, the source image and the target image selected from the plurality of medical images.

3. The method of claim 1 , further comprising:
after registering the plurality of medical images and segmenting the anatomical structure, segmenting a lesion within the anatomical structure.

4. The method of claim 3 , wherein segmenting the anatomical structure includes using a convolutional neural network, and segmenting the lesion includes extracting and clustering textural features based on a received maximal axial diameter of the lesion, the textural features including at least one of a mean, a median, and a standard deviation of the pixel values within the segmented anatomical structure in each of the plurality of medical images.

5. The method of claim 3 , wherein the exam is a multi-phasic contrast-enhanced computed tomography (CT) exam, and segmenting the anatomical structure includes segmenting a liver.

6. The method of claim 5 , wherein the multi-phasic contrast-enhanced CT exam includes an unenhanced phase, an arterial phase after the unenhanced phase, a portal phase after the arterial phase, and a delayed phase after the portal phase, and the plurality of medical images include a first CT image of the liver obtained during the unenhanced phase, a second CT image of the liver obtained during the arterial phase, a third CT image of the liver obtained during the portal phase, and a fourth CT image of the liver obtained during the delayed phase.

7. The method of claim 5 , wherein the reference ROI comprises tissue outside of the segmented lesion and within the segmented liver, wherein the characteristics of the anatomical structure include a tissue classification of each portion of the segmented lesion, and wherein determining the characteristics of the anatomical structure by tracking the pixel values of the segmented anatomical structure over time includes determining the tissue classification of each portion of the segmented lesion based on a value of each pixel within the portion relative to the reference ROI in each of the plurality of medical images.

8. The method of claim 7 , wherein the tissue classification is one of necrotic, viable, chemoembolized, undefined, peripheral enhancement, and parenchymal.

9. The method of claim 8 , wherein the characteristics of the anatomical structure further include an effect of a treatment, and determining the characteristics of the anatomical structure by tracking the pixel values of the segmented anatomical structure over time further includes determining the effect of the treatment by comparing the tissue classification of each portion of the segmented lesion from a first exam and with the tissue classification of each portion of the segment lesion from a second exam, the first exam performed before the treatment and the second exam performed after the treatment.

10. The method of claim 7 , wherein the characteristics of the anatomical structure further include Liver Imaging Reporting and Data System (LI-RADS) scores, and determining the characteristics of the anatomical structure by tracking the pixel values of the segmented anatomical structure over time further includes:
determining arterial phase hyperenhancement (APHE), wash-out, and capsule probabilities of a contrast agent used in the multi-phasic contrast-enhanced CT exam based on the pixel values of the segmented lesion relative to the reference ROI in each of the plurality of medical images.

11. A method, comprising:
receiving a series of liver images from a multi-phasic exam;
aligning the series of liver images;
segmenting the liver and a lesion within the series of aligned liver images;
sub-segmenting the lesion based on changes in pixel values of the segmented lesion across the series of aligned liver images; and
outputting an analysis of the sub-segmented lesion,
wherein sub-segmenting the lesion based on the changes in pixel values of the segmented lesion across the series of aligned liver images includes:
comparing the pixel values of the segmented lesion to pixel values in a reference region, the reference region within the boundary of the liver and outside of the boundary of the lesion, in each image of the series of aligned liver images; and
determining a classification of each sub-segment of the lesion as one of viable tissue, necrotic tissue, undefined tissue, and chemoembolized tissue based on a change in the pixel values within the sub-segment relative to the pixel values in the reference region across the series of aligned liver images.

12. The method of claim 11 , wherein the series of liver images include one liver image from each phase of the multi-phasic exam, and segmenting the liver includes:
defining a boundary of the liver in one image of the series of aligned liver images using a convolutional neural network; and
applying the boundary of the liver to each additional image in the series of aligned liver images.

13. The method of claim 12 , wherein segmenting the lesion includes:
defining a boundary of the lesion within the boundary of the liver in the one image of the series of aligned liver images based on textural features of the one image, the textural features including at least one of a mean, a median, and a standard deviation of the pixel values within the boundary of the liver in the one image; and
applying the boundary of the lesion to each additional image in the series of aligned liver images.

14. The method of claim 11 , wherein outputting the analysis of the sub-segmented lesion includes outputting the classification of each sub-segment of the lesion as an annotated image of the sub-segmented lesion.

15. The method of claim 14 , further comprising, determining a diagnostic score of the lesion based on an intensity profile of each pixel within the viable tissue sub-segment over the series of aligned liver images, and wherein outputting the analysis of the sub-segmented lesion includes outputting the diagnostic score.

16. A system, comprising:
a computed tomography (CT) system;
a memory storing instructions; and
a processor communicably coupled to the memory and, when executing the instructions, configured to:
receive a plurality of images acquired by the CT system during a multi-phasic exam enhanced with a contrast agent;
identify a boundary of an anatomical feature in at least one of the plurality of images via a deep learning model;
determine a diagnostic score for the anatom

Assignees

• Ge Prec Healthcare Llc

Inventors

• Morard Vincent Jerome
• Quelever Ronan
• Dumont Laura
• Gogin Nicolas
• Coustaud Nicolas
• Nicodeme Luc

Classifications

• G06T7/0012 Primary

  Biomedical image inspection · CPC title

• G16H30/40 Primary

  for processing medical images, e.g. editing · CPC title

• G06N3/045

  Combinations of networks · CPC title

• G06F18/241

  relating to the classification model, e.g. parametric or non-parametric approaches · CPC title

• G06T7/0014

  using an image reference approach · CPC title