AUTHOR=Andreassen Maren M. Sjaastad , Loubrie Stephane , Tong Michelle W. , Fang Lauren , Seibert Tyler M. , Wallace Anne M. , Zare Somaye , Ojeda-Fournier Haydee , Kuperman Joshua , Hahn Michael , Jerome Neil P. , Bathen Tone F. , Rodríguez-Soto Ana E. , Dale Anders M. , Rakow-Penner Rebecca 

TITLE=Restriction spectrum imaging with elastic image registration for automated evaluation of response to neoadjuvant therapy in breast cancer

JOURNAL=Frontiers in Oncology

VOLUME=Volume 13 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2023.1237720

DOI=10.3389/fonc.2023.1237720

ISSN=2234-943X

ABSTRACT=Neoadjuvant therapy of breast cancer is used to downstage tumors to allow for breastconserving surgery and to provide important information regarding drug efficacy prognosis. The current gold standard for monitoring treatment response is manual assessment of change in tumor size on dynamic contrast-enhanced MRI (DCE). However, DCE requires expert radiologist readers, administration of exogenous Gadolinium contrast with unknown sequela, and relatively high time and cost. Here we present an automatic cancer tissue classifier to predict response to neoadjuvant treatment of breast cancer. We here employ an advanced MRI model, three-component Restriction Spectrum Imaging (RSI3C), to generate an automatic quantification procedure and find that the derived tissue classifier indicated the ability to assess treatment response after only three weeks of therapy. The RSI3C classifier was also able to identify cases that correspond to residual tumor at surgery at the later phase of therapy. The results were similar to manual tumor size measurements by standard-of-care MRI by DCE. The classifier eliminates the need for manual user input as well as exogenous contrast agents that conventional MRI workflow methods require. Our results suggest that RSI3C may guide clinical decision-making and enable tailored treatment regimens and cost-efficient evaluation of neoadjuvant therapy of breast cancer.