Automatic measurement of lesions on medical images

• US 10,269,118 B2
• Filed: 11/09/2017
• Issued: 04/23/2019
• Est. Priority Date: 03/01/2013
• Status: Active Grant

• Current Assignee: Centre Hospitalier Universitaire d'Angers, Universite D'Angers
• Original Assignee: Centre Hospitalier Universitaire d'Angers, Universite D'Angers
• Inventors: Cales, Paul, Chaigneau, Julien, Hunault, Gilles, Rousselet, Marie-Christine, Aube, Christophe
• Primary Examiner(s): Azarian, Seyed H
• Application Number:
  US15/808,136
• Publication Number:
  US 20180075600A1
• Time in Patent Office: 530 Days
• Field of Search: 382100, 382103, 382106, 382128-134, 382154, 382162, 382168, 382173, 382181, 382190, 382193, 382199, 382203, 382209, 382219, 382 22, 382 4, 382232, 382254, 382274, 382276, 382291, 382305, 382312, 424464, 424 856, 707769, 435 611, 702 21
• CPC Class Codes: G06T 2207/10056 Microscopic image
• CPC Class Codes: G06T 2207/30056 Liver; Hepatic
• CPC Class Codes: G06T 2207/30096 Tumor; Lesion
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/90 Determination of colour characteristics

Abstract

A method for assessing the presence and/or the severity of a lesion in an organ or tissue of a subject through automated analysis of at least one image of the organ or tissue is provided. In particular, the organ or tissue is a liver organ or liver tissue. The method includes the calculation of a score that combines descriptors of the image, and includes the steps of: a) measuring on the image at least two descriptors of the image, b) mathematically combining the descriptors in a score; and c) assessing the presence and/or the severity of a lesion in the organ or tissue based on the value of the score calculated at step b).

Claims

1. A non-invasive method for assessing the presence of non-alcoholic fatty liver disease (NAFLD), non-alcoholic steatohepatitis (NASH) or severe fibrosis in a liver organ or liver tissue of a subject through automated analysis of at least one image of said liver organ or liver tissue, the method utilizing a microprocessor comprising a computer algorithm configured to analyze the at least one image and calculate a score combining descriptors of said image, said method comprising the steps of:

a. measuring on said at least one image at least two descriptors of said at least one image, the descriptors being selected from the group consisting of;

fractal dimension of the edges of the organ or fragment thereof, linearity percentage of the edges, nodularity of the curved and irregular edges of the organ, angularity, length of the organ or fragment thereof, length of a biopsy, length of numeric specimen, height of the organ, perimeter of the organ or fragment thereof, mean and total native perimeter of the organ, smoothed perimeter of the organ, ratio between the native and smoothed perimeters, also referred as anfractuosity, largest perimeter of the organ, indentation of the organ, area of the organ or fragment thereof, granularity percentage, fragmentation, mean intensity of the image on the red component, mean intensity of the image on the green component, mean intensity of the image on the blue component, area of fibrosis, fractal dimension of fibrosis, mean of percentage of fibrosis around areas, number of nodules, number of nodules with more than 30% of fibrosis around, area of steatosis, relative area of steatosis, fractal dimension of steatosis, area of porto-septal fibrosis, fractal dimension of porto-septal fibrosis, area of peri-sinusoidal fibrosis, fractal dimension of peri-sinusoidal fibrosis, area of lobular peri-sinusoidal fibrosis, ratio of peri-sinusoidal fibrosis among the whole fibrosis, luminosity of fibrosis staining in the red, green and/or blue components, luminosity of the parenchyma staining in the red, green and/or blue components, luminosity contrast between fibrosis and parenchyma, luminosity contrast between fibrosis and the organ or fragment thereof, area of stellar fibrosis among the total surface of the LB specimen, area of stellar fibrosis among the surface of porto-septal regions, area of stellar fibrosis among the surface of lobular regions, number of porto-septal regions;

mean area of stellar fibrosis, mean area of porto-septal regions, number of bridges, ratio of bridges among the porto-septal areas, area of fibrosis in the bridges, bridges thickness, bridges perimeter, bridges surface, portal distance, organ fat ratio, abdominal fat ratio, hypertrophy of liver segment I, surface of the segment I, width of the liver segment IV, ratio between segment I and segment IV dimensions, furrow thickness, surface of the furrow I, internal nodularity in the liver, diameter of the portal vein, heterogeneity of the density intensity, fractal organization of the organ, mean total density of the image, standard deviation of total density of the image, coefficient of variation of total density of the image, median total density of the image, interquartile range of total density of the image, ratio between interquartile range of total density and median of total density of the image, mean density of a region of interest (ROI) on the image, standard deviation of ROI density on the image, coefficient of variation of ROI density on the image, median ROI density on the image, interquartile range of ROI density on the image, ratio between interquartile range of ROI density and median of ROI density on the image, mean surface of the organ or fragment thereof, total surface of the organ or fragment thereof, total mean surface of the organ or fragment thereof, ratio between the organ perimeter and the organ surface, ratio between spleen surface and liver surface, ratio between spleen perimeter and liver perimeter, ratio between segment I surface and liver surface, Arantius furrow thickness, Arantius furrow surface, portal furrow thickness and number of fragments;

b. mathematically combining said at least two descriptors from step (a) in a score, wherein said mathematical combination is a binary logistic regression, a linear discriminant analysis, or a multiple linear regression; and

c. assessing the presence of non-alcoholic fatty liver disease (NAFLD), non-alcoholic steatohepatitis (NASH) or severe fibrosis in the liver organ or liver tissue, based on the value of the score calculated at step (b).

2. The method according to claim 1, further comprising measuring at least one non-invasive test marker selected from the group consisting of a biomarker, a clinical data, a physical data;

and/or a score.

3. The method according to claim 1, wherein the at least one image is a histological section image or an image of the entire organ or tissue.

4. The method according to claim 1, comprising recovering:

at least one histological section image; and

at least one image of the entire organ or tissue.

5. The method according to claim 1, wherein the at least one image is obtained by an optical technique selected from the group consisting of microscopic physical imaging, second harmonic generation (SHG), multiphoton imaging, coherent anti-Stokes Raman scattering (CARS), two-photon excitation fluorescence (TPEF), and diffuse optical imaging.

6. The method according to claim 1, wherein the at least one image is obtained by a non-optical technique selected from the group consisting of radiography, nuclear medicine, photoacoustic methods, and thermal methods.

7. The method according to claim 1, wherein said method is for assessing the presence of non-alcoholic fatty liver disease (NAFLD) or non-alcoholic steatohepatitis (NASH) and wherein the score obtained in step (c) is a diagnostic score indicative of the presence of non-alcoholic fatty liver disease (NAFLD) or non-alcoholic steatohepatitis (NASH).

8. The method according to claim 7, wherein said method comprises measuring in step (a) and mathematically combining in a score in step (b) the following descriptors:

fractal dimension of steatosis, fractal dimension of fibrosis and relative area of steatosis.

9. The method according to claim 1, wherein said method is for assessing the presence of definitive or borderline and definitive non-alcoholic steatohepatitis (NASH) and wherein the score obtained in step (c) is a diagnostic score indicative of the presence of definitive or borderline and definitive non-alcoholic steatohepatitis (NASH).

10. The method according to claim 9, wherein said method is for assessing the presence of definitive non-alcoholic steatohepatitis (NASH) and comprises measuring in step (a) and mathematically combining in a score in step (b) the following descriptors:

luminosity contrast between fibrosis and parenchyma, biopsy length and area of stellar fibrosis among the total surface of the LB specimen; and

wherein the score obtained in step (c) is a diagnostic score indicative of the presence of definitive non-alcoholic steatohepatitis (NASH).

11. The method according to claim 9, wherein said method is for assessing the presence of borderline and definitive non-alcoholic steatohepatitis (NASH) and comprises measuring in step (a) and mathematically combining in a score in step (b) the following descriptors:

luminosity contrast between fibrosis and parenchyma, area of stellar fibrosis among the surface of porto-septal regions and area of stellar fibrosis among the total surface of the LB specimen; and

wherein the score obtained in step (c) is a diagnostic score indicative of the presence of borderline and definitive non-alcoholic steatohepatitis (NASH).

12. The method according to claim 1, wherein said method is for assessing the presence of a severe fibrosis and wherein the score obtained in step (c) is a diagnostic score indicative of the presence of severe fibrosis.

13. The method according to claim 12, wherein said method comprises measuring in step (a) and mathematically combining in a score in step (b) the following descriptors:

ratio segment I surface to liver surface, second transversal segment I length (D4) and ratio native total liver perimeter to smoothed total liver perimeter.

14. The method according to claim 12, wherein said method comprises measuring in step (a) and mathematically combining in a score in step (b):

the following descriptors;

ratio segment I surface to liver surface, second transversal segment I length (D4) and ratio native total liver perimeter to smoothed total liver perimeter; and

liver stiffness.

15. A device for implementing the method of claim 1, wherein said device is a microprocessor comprising a computer algorithm to perform the mathematical combination of step (b).

16. A non-invasive method for predicting the METAVIR stage of a lesion in a liver organ or liver tissue of a subject through automated analysis of at least one image of said liver organ or liver tissue, the method utilizing a microprocessor comprising a computer algorithm configured to analyze the at least one image and calculate a score combining descriptors of said image, said method comprising the steps of:

a. measuring on said at least one image at least two descriptors of said at least one image, the descriptors being selected from the group consisting of;

fractal dimension of the edges of the organ or fragment thereof, linearity percentage of the edges, nodularity of the curved and irregular edges of the organ, angularity, length of the organ or fragment thereof, length of a biopsy, length of numeric specimen, height of the organ, perimeter of the organ or fragment thereof, mean and total native perimeter of the organ, smoothed perimeter of the organ, ratio between the native and smoothed perimeters, also referred as anfractuosity, largest perimeter of the organ, indentation of the organ, area of the organ or fragment thereof, granularity percentage, fragmentation, mean intensity of the image on the red component, mean intensity of the image on the green component, mean intensity of the image on the blue component, area of fibrosis, fractal dimension of fibrosis, mean of percentage of fibrosis around areas, number of nodules, number of nodules with more than 30% of fibrosis around, area of steatosis, relative area of steatosis, fractal dimension of steatosis, area of porto-septal fibrosis, fractal dimension of porto-septal fibrosis, area of peri-sinusoidal fibrosis, fractal dimension of peri-sinusoidal fibrosis, area of lobular peri-sinusoidal fibrosis, ratio of peri-sinusoidal fibrosis among the whole fibrosis, luminosity of fibrosis staining in the red, green and/or blue components, luminosity of the parenchyma staining in the red, green and/or blue components, luminosity contrast between fibrosis and parenchyma, luminosity contrast between fibrosis and the organ or fragment thereof, area of stellar fibrosis among the total surface of the LB specimen, area of stellar fibrosis among the surface of porto-septal regions, area of stellar fibrosis among the surface of lobular regions, number of porto-septal regions;

mean area of stellar fibrosis, mean area of porto-septal regions, number of bridges, ratio of bridges among the porto-septal areas, area of fibrosis in the bridges, bridges thickness, bridges perimeter, bridges surface, portal distance, organ fat ratio, abdominal fat ratio, hypertrophy of liver segment I, surface of the segment I, width of the liver segment IV, ratio between segment I and segment IV dimensions, furrow thickness, surface of the furrow I, internal nodularity in the liver, diameter of the portal vein, heterogeneity of the density intensity, fractal organization of the organ, mean total density of the image, standard deviation of total density of the image, coefficient of variation of total density of the image, median total density of the image, interquartile range of total density of the image, ratio between interquartile range of total density and median of total density of the image, mean density of a region of interest (ROI) on the image, standard deviation of ROI density on the image, coefficient of variation of ROI density on the image, median ROI density on the image, interquartile range of ROI density on the image, ratio between interquartile range of ROI density and median of ROI density on the image, mean surface of the organ or fragment thereof, total surface of the organ or fragment thereof, total mean surface of the organ or fragment thereof, ratio between the organ perimeter and the organ surface, ratio between spleen surface and liver surface, ratio between spleen perimeter and liver perimeter, ratio between segment I surface and liver surface, Arantius furrow thickness, Arantius furrow surface, portal furrow thickness and number of fragments;

b. mathematically combining said at least two descriptors from step (a) in a score, wherein said mathematical combination is a binary logistic regression, a linear discriminant analysis, or a multiple linear regression; and

c. predicting the METAVIR stage of a lesion in the liver organ or liver tissue, based on the value of the score calculated at step (b).

17. The method according to claim 16, wherein said method comprises measuring in step (a) and mathematically combining in a score in step (b) the following descriptors:

linearity percentage of the edges, granularity percentage, fragmentation, mean of percentage of fibrosis around areas, fractal dimension of porto-septal fibrosis, fractal dimension of peri-sinusoidal fibrosis, ratio of peri-sinusoidal fibrosis among the whole fibrosis and ratio of bridges among the porto-septal areas.

18. A device for implementing the method of claim 16, wherein said device is a microprocessor comprising a computer algorithm to perform the mathematical combination of step (b).