Method for generating a biomarker, system

Patent metadata

Field	Value
Publication number	US-12045980-B2
Application number	US-202017631258-A
Country	US
Kind code	B2
Filing date	Jun 3, 2020
Priority date	Jul 31, 2019
Publication date	Jul 23, 2024
Grant date	Jul 23, 2024

Abstract

A Method for generating a biomarker includes acquiring an image using an MRI system; processing the MRI image to generate a three-dimensional image of the lung; generating a first function corresponding to the distribution of the different signal intensity values; automatically calculating a filtering threshold of the first function from a second signal intensity value distribution function; segmenting a lung volume comprising: a main volume; a filtered volume of a volume of voxels quantified by the first function and filtered by at least the calculated filtering threshold, normalizing the values of the three-dimensional image of the lung volume; generating a biomarker indicating a normalized segmented volume ratio.

First claim

The invention claimed is:

1. A method for generating a biomarker comprising:
acquiring an MRI image using an MRI system;
processing said MRI image to generate a three-dimensional image of the lung;
generating a first function corresponding to a distribution of different signal intensity values of each voxel of a portion of the acquired three-dimensional image;
automatically calculating at least one filtering threshold of said first function from at least one second different signal intensity value distribution function of each voxel of a portion of the acquired three-dimensional image;
segmenting a volume comprising:
a main volume, corresponding to the lung volume;
a filtered volume of a volume of voxels quantified by the first function and filtered by at least the calculated filtering threshold,

normalizing the values of the three-dimensional image of the lung volume from absolute values of the signal intensity values of the voxels of the image and at least the calculated filtering threshold;
generating a biomarker indicating a normalized segmented volume ratio.

2. The method for generating a biomarker according to claim 1 , wherein:
the acquisition of a three-dimensional image by an MRI system is configured by:
a T 2 weighting;
an echo time greater than a predefined threshold;

the automatic calculation of at least one filtering threshold comprises:
acquiring a reference volume;
generating a reference function corresponding to a distribution of the different signal intensity values of each voxel of the reference volume, said reference function being the second distribution function;
calculating the standard deviation of the reference function;
determining a filtering threshold, designated reference threshold, from the calculated standard deviation of the reference function.

3. The method for generating a biomarker according to claim 2 , wherein the acquisition is parameterized and comprises:
an acquisition of an image recomposed of a plurality of images acquired over a number of cycle of at least 4 echo times, the echo times being configured according to increasing durations;
a spin echo sequence;
a parameterization aiming to emit a signal to pre-saturate or to saturate the acquired signal.

4. The method for generating a biomarker according to claim 2 , wherein a second acquisition of the image is made with a first configuration to outline the lung volume, the first configuration defining a parameterization of a T 1 or proton density weighted acquisition, a step of image processing being carried out to combine the image acquired with T 2 weighting with the image acquired by the second acquisition.

5. The method for generating a biomarker according to claim 4 , wherein a merger operation between at least one image acquired with T 1 weighting with an ultra-short echo time UTE and at least one image acquired with T 2 weighting is made to generate an image of which the data coming from each of the acquired images have been combined to segment the lung volume.

6. The method for generating a biomarker according to claim 2 , wherein the reference threshold is established from a combination between a reference distribution value of the reference function and a value comprised between 10 and 20 times the value of the standard deviation of the reference function.

7. The method for generating a biomarker according to claim 6 , wherein the reference distribution value of the reference function is the main mode of the distribution of the signal intensity values of the image acquired with T 2 weighting within the lung volume.

8. The method for generating a biomarker according to claim 2 , further comprising a step of normalization, the normalization comprising the calculation of a volume intensity product of the signal from the absolute values of the signal of the filtered volume, the volume resulting from the filtered volume and the lung volume.

9. The method for generating a biomarker according to claim 1 , wherein:
the acquisition of a three-dimensional image using an MRI system, is configured by:
a proton density or T 1 weighting, an echo time
an echo time less than a predefined threshold;

the automatic calculation of at least one filtering threshold comprises:
modeling at least two gaussian functions by adjustment of the first function;
determining the filtering threshold, called first threshold, by a calculation of the intersection of the first gaussian function and the second gaussian function;
determining a second threshold corresponding to a minimum value of the first gaussian function and a minimum value of a signal intensity value of a voxel;

the filtered volume is a first volume corresponding to the voxels quantified by the first gaussian function comprised between the first threshold and the second threshold, said voxels corresponding to an air medium,
normalizing the values of the three-dimensional image of the lung volume from the first threshold and the second threshold calculated;
generating a first biomarker indicating a ratio of a characteristic volume of the normalized segmented volume, said ratio being calculated between the characteristic volume and the lung volume.

10. The method for generating a biomarker according to claim 9 , wherein the echo time is less than 1 ms.

11. The method for generating a biomarker according to claim 9 , characterized in that wherein the segmentation comprises the definition of a second volume corresponding to the voxels quantified by the second gaussian function greater than the first threshold, said voxels corresponding to a fatty or intermediate medium.

12. The method for generating a biomarker according to claim 9 , wherein the segmentation comprises a step of extracting a characteristic volume comprising voxels of the filtered volume of which the signal intensity value is less than a third predefined threshold, said third predefined threshold being determined on a normalized scale of [0; 1].

13. The method for generating a biomarker according to claim 11 , wherein the segmentation comprises a step of exclusion/deletion of voxels disconnected from their neighborhood of voxels of a same quantification.

14. The method for generating a biomarker according to claim 9 , wherein the modeling of the gaussian functions comprises:
a gaussian smoothing applied to the acquired image with reduction of the one-off encoding time and with radial acquisition for denoising;
outlining the contours by application of a local filter;
using the curve adjustment method representing the frequency of distribution of voxels.

15. The method for generating a biomarker according to claim 1 , further comprising an acquisition of the signal intensity values of each voxel quantified by the first function, the intensity values corresponding to an image contrast datum.

16. The method for generating a biomarker according to claim 1 , characterized in that wherein the acquisition is carried out in a synchronized manner with a respirator, the respirator being a navigator or a respiratory belt.

17. The method for generating a biomarker according to claim 1 , wherein a step of extracting a volume image is carried out from the images acquired by MRI, said extracted image being realized at a determined instant of the sequence.

18. The method for generating a biomarker according to claim 1 , wherein the acquisition of the three-dimensional image is made by a stack of acquired 2D images, a thickness of the section being equal at least to the width of a voxel.

19. The method for gener

Assignees

• Univ Bordeaux
• Centre Hospitalier Univ Bordeaux
• Institut National De La Sante Et De La Rech Medicale—Inserm

Inventors

• Dournes Gael
• Benlala Ilyes
• Laurent François
• Berger Patrick

Classifications

• G06T2207/30061

  Lung · CPC title

• G06T2207/20024

  Filtering details · CPC title

• G06T2207/10088

  Magnetic resonance imaging [MRI] · CPC title

• G06T7/11

  Region-based segmentation · CPC title

• G06T7/136

  involving thresholding · CPC title