Computational imaging reveals shape differences between normal and malignant prostates on MRI

Mirabela Rusu, Andrei S. Purysko, Sadhna Verma, Jonathan Kiechle, Jay Gollamudi, Soumya Ghose, Karin Herrmann, Vikas Gulani, Raj Paspulati, Lee Ponsky, Maret Bohm, Anne Maree Haynes, Daniel Moses, Ron Shnier, Warick Delprado, James T. Thompson, Phillip D. Stricker, Anant Madabhushi

Research output: Contribution to journalArticlepeer-review

8 Citations (Scopus)
90 Downloads (Pure)

Abstract

We seek to characterize differences in the shape of the prostate and the central gland (combined central and transitional zones) between men with biopsy confirmed prostate cancer and men who were identified as not having prostate cancer either on account of a negative biopsy or had pelvic imaging done for a non-prostate malignancy. T2w MRI from 70 men were acquired at three institutions. The cancer positive group (PCa+) comprised 35 biopsy positive (Bx+) subjects from three institutions (Gleason scores: 6-9, Stage: T1-T3). The negative group (PCa-) combined 24 biopsy negative (Bx-) from two institutions and 11 subjects diagnosed with rectal cancer but with no clinical or MRI indications of prostate cancer (Cl-). The boundaries of the prostate and central gland were delineated on T2w MRI by two expert raters and were used to construct statistical shape atlases for the PCa+, Bx- and Cl- prostates. An atlas comparison was performed via per-voxel statistical tests to localize shape differences (significance assessed at p <0.05). The atlas comparison revealed central gland hypertrophy in the Bx- subpopulation, resulting in significant volume and posterior side shape differences relative to PCa+ group. Significant differences in the corresponding prostate shapes were noted at the apex when comparing the Cl- and PCa+ prostates.

Original languageEnglish
Article number41261
Pages (from-to)1-11
Number of pages11
JournalScientific Reports
Volume7
DOIs
Publication statusPublished - 1 Feb 2017
Externally publishedYes

Bibliographical note

Copyright the Author(s) 2017. Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher.

Fingerprint

Dive into the research topics of 'Computational imaging reveals shape differences between normal and malignant prostates on MRI'. Together they form a unique fingerprint.

Cite this