TractOracle: towards an anatomically-informed reward function for RL-based tractography

TL;DR

TractOracle introduces an anatomically-informed reward function for reinforcement learning-based tractography, significantly improving true positive rates and reducing false positives by leveraging a streamline classification network.

Contribution

The paper presents TractOracle, a novel RL tractography system that incorporates an anatomical reward network for better streamline classification and early stopping, reducing false positives.

Findings

True positive ratios improved by nearly 20%

False positive ratios reduced by 3x

Number of true positive streamlines increased 2x to 7x

Abstract

Reinforcement learning (RL)-based tractography is a competitive alternative to machine learning and classical tractography algorithms due to its high anatomical accuracy obtained without the need for any annotated data. However, the reward functions so far used to train RL agents do not encapsulate anatomical knowledge which causes agents to generate spurious false positives tracts. In this paper, we propose a new RL tractography system, TractOracle, which relies on a reward network trained for streamline classification. This network is used both as a reward function during training as well as a mean for stopping the tracking process early and thus reduce the number of false positive streamlines. This makes our system a unique method that evaluates and reconstructs WM streamlines at the same time. We report an improvement of true positive ratios by almost 20\% and a reduction of 3x of false positive ratios on one dataset and an increase between 2x and 7x in the number true positive streamlines on another dataset.