Multi-resolution super learner for voxel-wise classification of prostate cancer using multi-parametric MRI.

Jin Jin, Lin Zhang, Ethan Leng, Gregory J Metzger, Joseph S Koopmeiners
Author Information
  1. Jin Jin: Department of Biostatistics, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD, USA.
  2. Lin Zhang: Devision of Biostatistics, School of Public Health, University of Minnesota, Minneapolis, MN, USA.
  3. Ethan Leng: Department of Biomedical Engineering, University of Minnesota, Minneapolis, MN, USA.
  4. Gregory J Metzger: Department of Radiology, University of Minnesota, Minneapolis, MN, USA.
  5. Joseph S Koopmeiners: Devision of Biostatistics, School of Public Health, University of Minnesota, Minneapolis, MN, USA.

Abstract

Multi-parametric MRI (mpMRI) is a critical tool in prostate cancer (PCa) diagnosis and management. To further advance the use of mpMRI in patient care, computer aided diagnostic methods are under continuous development for supporting/supplanting standard radiological interpretation. While voxel-wise PCa classification models are the gold standard, few if any approaches have incorporated the inherent structure of the mpMRI data, such as spatial heterogeneity and between-voxel correlation, into PCa classification. We propose a machine learning-based method to fill in this gap. Our method uses an ensemble learning approach to capture regional heterogeneity in the data, where classifiers are developed at multiple resolutions and combined using the super learner algorithm, and further account for between-voxel correlation through a Gaussian kernel smoother. It allows any type of classifier to be the base learner and can be extended to further classify PCa sub-categories. We introduce the algorithms for binary PCa classification, as well as for classifying the ordinal clinical significance of PCa for which a weighted likelihood approach is implemented to improve the detection of less prevalent cancer categories. The proposed method has shown important advantages over conventional modeling and machine learning approaches in simulations and application to our motivating patient data.

Keywords

References

Phys Med Biol. 2018 May 01;63(9):095004 [PMID: 29570456]
Med Phys. 2012 Jul;39(7):4093-103 [PMID: 22830742]
Stat Med. 2022 Feb 10;41(3):483-499 [PMID: 34747059]
Eur Urol. 2011 Apr;59(4):477-94 [PMID: 21195536]
Med Image Anal. 2021 Oct;73:102153 [PMID: 34246848]
J Magn Reson Imaging. 2018 Feb 22;: [PMID: 29469937]
Med Phys. 2017 Mar;44(3):1028-1039 [PMID: 28107548]
J Magn Reson Imaging. 2013 May;37(5):1035-54 [PMID: 23606141]
IEEE Trans Biomed Eng. 2021 Feb;68(2):374-383 [PMID: 32396068]
Histopathology. 2019 Jan;74(1):135-145 [PMID: 30565298]
J Urol. 2010 Feb;183(2):433-40 [PMID: 20006878]
Radiology. 2016 Jun;279(3):805-16 [PMID: 26761720]
Int J Comput Assist Radiol Surg. 2009 Jan;4(1):1-10 [PMID: 20033597]
IEEE Trans Image Process. 2010 Sep;19(9):2444-55 [PMID: 20716496]
Stat Appl Genet Mol Biol. 2007;6:Article25 [PMID: 17910531]
J Appl Clin Med Phys. 2020 Oct;21(10):179-191 [PMID: 32770600]
Proc Natl Acad Sci U S A. 2015 Nov 17;112(46):E6265-73 [PMID: 26578786]
Curr Opin Biomed Eng. 2017 Sep;3:20-27 [PMID: 29732440]
J Magn Reson Imaging. 2015 Apr;41(4):1104-14 [PMID: 24700476]
Comput Biol Med. 2015 May;60:8-31 [PMID: 25747341]
J Appl Clin Med Phys. 2019 Feb;20(2):146-153 [PMID: 30712281]
Abdom Imaging. 2015 Jan;40(1):134-42 [PMID: 25034558]
BMC Med Imaging. 2015 Aug 05;15:27 [PMID: 26242589]
J Magn Reson Imaging. 1996 Jul-Aug;6(4):603-7 [PMID: 8835953]
Nat Rev Urol. 2020 Jan;17(1):41-61 [PMID: 31316185]
BMC Med Imaging. 2019 Feb 28;19(1):22 [PMID: 30819131]
Am J Surg Pathol. 2005 Sep;29(9):1228-42 [PMID: 16096414]
IEEE Trans Med Imaging. 2014 May;33(5):1083-92 [PMID: 24770913]
J Magn Reson Imaging. 2014 Dec;40(6):1414-21 [PMID: 24243554]
Stat Med. 2018 Sep 30;37(22):3214-3229 [PMID: 29923345]
PLoS One. 2019 Jul 8;14(7):e0217702 [PMID: 31283771]
Phys Med Biol. 2012 Jun 21;57(12):3833-51 [PMID: 22640958]
Urology. 2000 Nov 1;56(5):823-7 [PMID: 11068310]
Histopathology. 2012 Jan;60(1):75-86 [PMID: 22212079]

Grants

  1. P30 CA077598/NCI NIH HHS
  2. P41 EB027061/NIBIB NIH HHS
  3. R01 CA155268/NCI NIH HHS
  4. R01 CA241159/NCI NIH HHS

Word Cloud

Similar Articles

Cited By