Method for identifying biomarkers using a probability map

10347,015

15/925082

March 19, 2018

A method of forming a probability map is disclosed. According to one embodiment, a method may include: (1) obtaining multiple measures of multiple imaging parameters for every stop of a moving window on an image, wherein two neighboring ones of the stops of the moving window are partially overlapped with each other; (2) obtaining first probabilities of an event for the stops of the moving window by matching the measures of the imaging parameters to a classifier; and (3) obtaining second probabilities of the event for multiple voxels of a probability map based on information associated with the first probabilities.

Schieke, Moira F. (Milwaukee, WI, US); Lin, Erica (Cambridge, MA, US)

1. A method for detecting a probability of an event in a tissue sample based on generating a probability map of an image of the tissue sample, the method comprising: defining, by the imaging system, boundaries of a plurality of computation voxels over at least a portion of the image of the tissue sample, wherein each the plurality of computation voxels is a unit of the probability map; applying, by the imaging system, a moving window to the plurality of computation voxels in incremental steps; obtaining, by the imaging system, multiple measures of multiple imaging parameters for every step of the moving window across the image; and determining, by the imaging system, values for each of the plurality of computation voxels, wherein the values for each of the plurality of computation voxels are based on the multiple measures of the multiple imaging parameters; and generating the probability map using the values for each of the plurality of computation voxels, wherein the probability map provides an indication of a likelihood of an event at a plurality of locations within the tissue sample.

2. The method of claim 1 , wherein the imaging parameters comprise at least four types of magnetic resonance imaging (MRI) parameters.

3. The method of claim 1 , wherein two neighboring ones of the steps of the moving window are shifted from each other by a distance substantially equal to a side length of one of the computation voxels.

4. The method of claim 1 , wherein the event is that a cancer occurs.

5. The method of claim 1 , wherein determining the probability map of the event comprises: calculating multiple assumed probabilities for the respective computation voxels of the probability map based on the first probabilities of the event for the steps of the moving window covering the respective computation voxels; calculating multiple probability guesses for the respective steps of the moving window based on the assumed probabilities for the computation voxels of the probability map within the respective steps of the moving window; calculating multiple differences each between one of the probability guesses and one of the first probabilities for one of the steps of the moving window; and updating the assumed probabilities for the respective computation voxels of the probability map based on the differences for the steps of the moving window covering the respective computation voxels of the probability map.

6. A method of forming a probability map, comprising: applying, by an imaging system, a moving window over a first portion of an image of a tissue at a first stop; obtaining, by the imaging system, multiple first measures of multiple imaging parameters for the first stop of the moving window from the image; moving, by the imaging system, the moving window relative to the image in a direction at a fixed interval to a second stop; obtaining, by the imaging system, multiple second measures of the multiple imaging parameters for the second stop, wherein the first stop of the moving window is partially overlapped with the second stop of the moving window; obtaining, by the imaging system, a first probability of a first event for the first stop of the moving window, wherein the first probability of the first event is based on the first measures; obtaining, by the imaging system, a second probability of the first event for the second stop of the moving window, wherein the second probability of the first event is based on the second measures; and determining, by the imaging system, the probability of a biomarker at a first location in the tissue based, at least in part, on the first probability and the second probability.

7. The method of claim 6 , wherein the image of the tissue comprises a magnetic resonance imaging (MRI) image.

8. The method of claim 6 , wherein the imaging parameters comprise at least four types of magnetic resonance imaging (MRI) parameters.

9. The method of claim 6 , wherein the first event is that a cancer occurs.

10. The method of claim 6 , wherein the first and second stops of the moving window are shifted from each other by a distance substantially equal to a side length of the computation voxel of the probability map.

11. The method of claim 6 , wherein the dimension of the moving window has a size defined based on volumes of multiple biopsy tissues in a subset of tissue samples.

12. The method of claim 6 , wherein the dimension of the moving window has a volume defined based on volumes of multiple biopsy tissues in a subset of tissue samples.

13. The method of claim 6 , wherein the moving window has a circular shape with a radius defined based on information associated with volumes of multiple biopsy tissues in the subset of tissue samples.

14. The method of claim 6 , wherein the first classifier comprises a Bayesian classifier.

15. The method of claim 6 , wherein the first classifier is created based on information associated with multiple third measures of the imaging parameters for multiple biopsy tissues in a subset of tissue samples and multiple diagnoses for the multiple biopsy tissues in the subset of tissue samples.

16. The method of claim 6 further comprising calculating a third probability of the first event for the at least one of the computation voxels of the probability map based on the first and second probabilities of the first event, wherein said first and second stops of the moving window overlap the at least one of the computation voxels of the probability map.

17. The method of claim 6 further comprising calculating a third probability of the first event for one of multiple computation voxels of the probability map, wherein calculating the third probability of the first event comprises: assuming the third probability of the first event by averaging the first and second probabilities; calculating a first probability guess for the first stop of the moving window by averaging a first group of probabilities of the first event for a first group of the plurality of computation voxels inside the first stop of the moving window, wherein the first group of probabilities of the first event comprise the third probability of the first event; calculating a second probability guess for the second stop of the moving window by averaging a second group of probabilities of the first event for a second group of the plurality of computation voxels inside the second stop of the moving window, wherein the second group of probabilities of the first event comprise the third probability of the first event; and determining whether a first absolute value of a first difference between the first probability guess and the first probability of the first event and a second absolute value of a second difference between the second probability guess and the second probability of the first event are less than or equal to a preset threshold value.

18. The method of claim 17 , wherein calculating the third probability of the first event further comprises updating the third probability of the first event based on information associated with the first and second differences.

19. The method of claim 6 further comprising: obtaining a third probability of a second event for the first stop of the moving window by matching the first measures to a second classifier; obtaining a fourth probability of the second event for the second stop of the moving window by matching the second measures to the second classifier; calculating a fifth probability of the first event based on the first and second probabilities of the first event; calculating a sixth probability of the second event based on the third and fourth probabilities of the second event; and creating a composite probability map based on information associated with the fifth probability of the first event and the sixth probability of the second event.

20. The method of claim 19 , wherein the first event is that a cancer occurs, and the second event is associated with a Gleason score.

6123647 September 2000 Mitchell
6567684 May 2003 Chenevert et al.
6956373 October 2005 Brown et al.
8509570 August 2013 Degani et al.
8605980 December 2013 Li et al.
8768431 July 2014 Ross et al.
8781214 July 2014 Davis et al.
8805619 August 2014 Sorensen et al.
8818484 August 2014 Liew et al.
8873836 October 2014 Dietrich et al.
9092691 July 2015 Beaumont et al.
9165362 October 2015 Siewerdsen et al.
9615028 April 2017 Mizutani
2002/0186818 December 2002 Arnaud et al.
2002/0193677 December 2002 Thornton
2003/0072479 April 2003 Sofia Totterman et al.
2006/0269476 November 2006 Kuo
2009/0161928 June 2009 Khamene
2010/0158332 June 2010 Rico et al.
2010/0284927 November 2010 Lu et al.
2011/0243417 October 2011 Madabhushi
2013/0197349 August 2013 Blumhagen et al.
2013/0329973 December 2013 Cao et al.
2014/0003697 January 2014 Qian et al.
2014/0010429 January 2014 Highnam et al.
2014/0010430 January 2014 Chandelier et al.
2014/0037172 February 2014 Madabhushi et al.
2014/0064580 March 2014 Madabhushi et al.
2014/0079302 March 2014 Sato et al.
2014/0101080 April 2014 Lee et al.
2014/0126794 May 2014 Ahn et al.
2014/0153795 June 2014 Lenox
2014/0185888 July 2014 Kelm et al.
2014/0185900 July 2014 Lee et al.
2014/0195472 July 2014 Kawagishi
2014/0205163 July 2014 Stark et al.
2014/0219535 August 2014 Chen et al.
2014/0228667 August 2014 Dankerl et al.
2014/0233826 August 2014 Agaian
2014/0241606 August 2014 Park et al.
2014/0309511 October 2014 Stal
2015/0003706 January 2015 Eftestol et al.
2016/0019693 January 2016 Silbersweig
2016/0038095 February 2016 Schieke
2016/0086326 March 2016 Raschke et al.
2016/0117816 April 2016 Taylor
2016/0217576 July 2016 Kabus et al.
2016/0292194 October 2016 Farkash
2016/0350933 December 2016 Schieke
2016/0350946 December 2016 Schieke
2017/0263023 September 2017 Zhou
2018/0114312 April 2018 Palma
3 117 771 January 2017
WO-2015/175746 November 2015
WO-2016/206942 December 2016

“Determining Single Voxel Value from Larger Region of Interest (ROI),” Dec. 18, 2014, 21 pages. cited by applicant
Antipolis, “MEDIAN Technologies strengthens IP portfolio with US patent,” MEDIAN Technologies (ALMDT), Sep. 10, 2015, 4 pages. cited by applicant
Ashraf, et al., “Identification of Intrinsic Imaging Phenotypes for Breast Cancer Tumors: Preliminary Associations with Gene Expression Profiles,” Radiology, Aug. 2014, pp. 374-384, vol. 272, No. 2. cited by applicant
Baselga, et al., “Everolimus in Postmenopausal Hormone-Receptor 2013 Positive Advanced Breast Cancer,” New England Journal of Medicine, Feb. 9, 2012, pp. 520-529, vol. 366, No. 6. cited by applicant
Boes, et al., “Image Registration for Quantitative Parametric Response Mapping of Cancer Treatment Response,” Translational Oncology, Feb. 2014, pp. 101-110, vol. 7, No. 1. cited by applicant
Bornn, et al., “Herded Gibbs Sampling,” Mar. 16, 2013, 21 pages. cited by applicant
Buckley,“Uncertainty in the Analysis of Tracer Kinetics Using Dynamic Contrast-Enhanced T1-Weighted MRI,” Magnetic Resonance in Medicine, Feb. 20, 2002, pp. 601-606, vol. 47. cited by applicant
Chan, et al., “Detection of Prostate Cancer by Integration of Line-Scan Diffusion, T2-Mapping and T2-Weighted Magnetic Resonance Imagine; a Multichannel Statistical Classifier,” Medical Physics, Sep. 2003, pp. 2390-2398, vol. 30, No. 9. cited by applicant
Colen, et al., “NCI Workshop Report: Clinical and Computational Requirements for Correlating Imaging Phenotypes with Genomics Signatures,” Translational Oncology, Oct. 2014, pp. 565-569, vol. 7, No. 5. cited by applicant
Ellingson, et al., “Volumetric Analysis of Functional Diffusion Maps is a Predictive Imaging Biomarker for Cytotoxic and Anti-Angiogenic Treatments in Malignant Gliomas,” Journal of Neuro-Oncology, Mar. 2011, pp. 95-103, vol. 102, Issue 1. cited by applicant
Ellingson, et al.,“Graded Functional Diffusion Map 2013 Defined Characteristics of Apparent Diffusion Coefficients Predict Overall Survival in Recurrent Glioblastoma Treated with Bevacizumab,” Neuro-Oncology, Oct. 2011, pp. 1151-1161, vol. 13, No. 10. cited by applicant
Galavis, et al.,“Variability of Textural Features in FDG PET Images Due to Different Acquisition Modes and Reconstruction Parameters,” Acta Oncologica, Sep. 2010, pp. 1012-1016, vol. 49, Issue 7. cited by applicant
Galban, et al., “The Parametric Response Map: An Imaging Biomarker for Early Cancer Treatment Outcome,” Nature Medicine, 2009, pp. 572-576, vol. 15, No. 5. cited by applicant
Galbraith, et al., Reproducibility of Dynamic Contrast-Enhanced MRI in Human Muscle and Tumours: Comparison of Quantitative and Semi-Quantitative Analysis, NMR in Biomedicine, Apr. 2002, pp. 132-142, vol. 15, Issue 2. cited by applicant
Gillies, et al., “MRI of the Tumor Microenvironment,” Journal of Magnetic Resonance Imaging, Oct. 2002, pp. 430-450, vol. 16, Issue 4. cited by applicant
Haq, et al., “A Data-Driven Approach to Prostate Cancer Detection From Dynamic Contrast Enhanced MRI,” Computerized Medical Imaging and Graphics, Apr. 2015, pp. 37-45, vol. 41. cited by applicant
International Search Report and Written Opinion in International Application No. PCT/US2017/040456 dated Oct. 19, 2017 (14 pages). cited by applicant
Irani, et al., “Motion Analysis for Image Enhancement: Resolution, Occlusion, and Transparency,” Journal of Visual Communication and Image Representation, Dec. 1993, pp. 325-335, vol. 4, No. 4. cited by applicant
Kwak, et al., “Correlation of Magnetic Resonance Imaging With Digital Histopathology in Prostate,” International Journal of Computer Assisted Radiology and Surgery, Apr. 2016, pp. 657-666, vol. 11, Issue 4. cited by applicant
Kwak, et al., “Automated Prostate Cancer Detection Using T2-Weighted and High-B-Value Diffusion-Weighted Magnetic Resonance Imaging,” The International Journal of Medical Physics Research and Practice, May 2015, pp. 2368-2378, vol. 42, Issue 5. cited by applicant
Kwak, et al., “Prostate Cancer: A Correlative Study of Multiparametric MR Imaging and Digital Histopathology,” Radiology, Oct. 2017, pp. 1-10, vol. 285, Issue 1. cited by applicant
Langer, et al., “Prostate Cancer Detection With Multi-parametric MRI: Logistic Regression Analysis of Quantitative T2, Diffusion-Weighted Imaging, and Dynamic Contrast-Enhanced MRI,” Journal of Magnetic Resonance Imaging, Aug. 2009, pp. 327-334, vol. 30, Issue 2. cited by applicant
Lee, et al, “Maximum Entropy and Bayesian Methods,” Kluwer Academic Publishers, 1993, 414 pages. cited by applicant
Li, et al., “Cell Membrane Water Exchange Effects in Prostate DCE-MRI,” Journal of Magnetic Resonance, May 2012, pp. 77-85, vol. 218. cited by applicant
Maani, et al., “Voxel-Based Texture Analysis of the Brain,” PLOS ONE, Mar. 10, 2015, pp. 1-19. cited by applicant
Maenpaa, et al., “Texture Analysis With Local Binary Patterns,” WSPC, May 13, 2004, pp. 1-20, vol. 8, Issue 19. cited by applicant
Method for Determining in Vivo Tissue Biomarker Characteristics Using Multiparameter MRI Matrix Creation and Big Data Analytics (Draft application), 22 pages. cited by applicant
Moffat, et al., “Functional Diffusion Map: A Noninvasive MRI Biomarker for Early Stratification of Clinical Brain Tumor Response,” Proceedings of the National Academy of Sciences, Apr. 2005, pp. 5524-5529, vol. 102, Issue 15. cited by applicant
Moradi, et al., “Multiparametric MRI Maps for Detection and Grading of Dominant Prostate Tumors,” Journal of Magnetic Resonance Imaging, Jun. 2012, pp. 1403-1413 vol. 35, Issue 6. cited by applicant
Nasrollahi, et al., “Super-Resolution: A Comprehensive Survey,” Machine Vision & Applications, Aug. 2014, pp. 1423-1468, vol. 25, Issue 6. cited by applicant
Niaf, et al., “Computer-Aided Diagnosis of Prostate Cancer in the Peripheral Zone Using Multiparametric MRI,” Physics in Medicine Biology, May 2012, pp. 3833-3851, vol. 57, No. 12. cited by applicant
Non-Final Rejection Office Action in U.S. Appl. No. 14/821,703 dated Mar. 22, 2017 (56 pages). cited by applicant
Non-Final Rejection Office Action on U.S. Appl. No. 15/165,644 dated Nov. 30, 2017 (40 pages). cited by applicant
Notice of Allowance in U.S. Appl. No. 14/821,703 dated Jan. 23, 2018 (11 pages). cited by applicant
Notice of Allowance in U.S. Appl. No. 14/821,703 dated Oct. 5, 2017 (11 pages). cited by applicant
Oto, et al., “Diffusion-Weighted and Dynamic Contrast-Enhanced MRI of Prostate Cancer: Correlation of Quantitative MR Parameters With Gleason Score and Tumor Angiogenesis,” American Journal of Roentgenology, Dec. 2011, pp. 1382-1390, vol. 197, No. 6. cited by applicant
Padhani, et al., “Reproducibility of Quantitative Dynamic MRI of Normal Human Tissues,” NMR in Biomedicine, Apr. 2002, pp. 143-153, vol. 15, Issue 2. cited by applicant
Peng, et al., “Quantitative Analysis of Multiparametric Prostate MR Images: Differentiation Between Prostate Cancer and Normal Tissue and Correlation with Gleason Score—A Computer-aided Diagnosis Developmental Study,” Radiology, Jun. 2013, pp. 787-796, vol. 267, No. 3. cited by applicant
Purysko, et al., “LI-RADS: A Case-based Review of the New Categorization of Liver Findings in Patients With End-Stage Liver Disease,” RadioGraphics, Nov.-Dec. 2012, pp. 1977-2012, vol. 32, Issue 7. cited by applicant
Rijpkema, et al., “Method for Quantitative Mapping of Dynamic MRI Contrast Agent Uptake in Human Tumors,” Journal of Magnetic Resonance Imaging, Oct. 2001, pp. 457-463, vol. 14, Issue 4. cited by applicant
Roberts, et al., “The Effect of Blood Inflow and B1-Field Inhomogeneity on Measurement of the Arterial Input Function in Axial3D Spoiled Gradient Echo Dynamic Contrast-Enhanced MRI,” Magnetic Resonance in Medicine, Jan. 2010, pp. 108-119, vol. 65, Issue 1. cited by applicant
Senseney, et al., “Tumor Treatment Response Identification Using Combination Post-Treatment Mapping to Quantify Voxel-Wise Multiparameter MRI Biomarker Changes: A Simulation Study,” International Symposium on Biomedical Imaging, May 2012, Barcelona, Spain. cited by applicant
Shah, et al., “Decision Support System for Localizing Prostate Cancer based on Multiparametric Magnetic Resonance Imaging,” Medical Physics, Jul. 2012, pp. 4093-4103, vol. 39, No. 7. cited by applicant
Wang, et al., “Computer Aided-Diagnosis of Prostate Cancer on Multiparametric MRI: A Technical Review of Current Research,” BioMed Research International, Aug. 2014, pp. 1-12, vol. 2014. cited by applicant
Yang, et al., “Comparison of Quantitative Parameters in Cervix Cancer Measured by Dynamic Contrast 2013 Enhanced MRI and CT,” Magnetic Resonance in Medicine, Jun. 2010, pp. 1601-1609, vol. 63, Issue 6. cited by applicant
Yang, et al., “Reproducibility Assessment of a Multiple Reference Tissue Method for Quantitative Dynamic Contrast Enhanced 2013 MRI Analysis,” Magnetic Resonance in Medicine, Apr. 2009, pp. 851-859, vol. 61, Issue 4. cited by applicant
Non-Final Office Action on U.S. Appl. No. 15/640,107 dated Jan. 23, 2019. cited by applicant
U.S. Office Action on U.S. Appl. No. 14/821,700 dated May 31, 2018. cited by applicant
Final Office Action on U.S. Appl. No. 14/821,700 dated Mar. 18, 2019. cited by applicant
International Search Report and Written Opinion in PCT/US2018/028679 dated Jan. 29, 2019 (14 pages). cited by applicant

edspgr.10347015