Policy to Assist Iteratively Local Segmentation: Optimising Modality and Location Selection for Prostate Cancer Localisation

TL;DR

This paper introduces a policy network that dynamically recommends optimal imaging modalities and regions to improve prostate cancer segmentation accuracy, surpassing standard methods and potentially assisting radiologists.

Contribution

A novel policy network framework that iteratively selects imaging modalities and regions to enhance machine learning-based prostate cancer segmentation.

Findings

The approach improves segmentation accuracy over standard networks.

The trained agent develops its own optimal strategy, sometimes differing from radiologist guidelines.

Demonstrated effectiveness on 1325 multiparametric MRI images.

Abstract

Radiologists often mix medical image reading strategies, including inspection of individual modalities and local image regions, using information at different locations from different images independently as well as concurrently. In this paper, we propose a recommend system to assist machine learning-based segmentation models, by suggesting appropriate image portions along with the best modality, such that prostate cancer segmentation performance can be maximised. Our approach trains a policy network that assists tumor localisation, by recommending both the optimal imaging modality and the specific sections of interest for review. During training, a pre-trained segmentation network mimics radiologist inspection on individual or variable combinations of these imaging modalities and their sections - selected by the policy network. Taking the locally segmented regions as an input for the next step, this dynamic decision making process iterates until all cancers are best localised. We validate our method using a data set of 1325 labelled multiparametric MRI images from prostate cancer patients, demonstrating its potential to improve annotation efficiency and segmentation accuracy, especially when challenging pathology is present. Experimental results show that our approach can surpass standard segmentation networks. Perhaps more interestingly, our trained agent independently developed its own optimal strategy, which may or may not be consistent with current radiologist guidelines such as PI-RADS. This observation also suggests a promising interactive application, in which the proposed policy networks assist human radiologists.