Towards Automatic Identification of Missing Tissues using a Geometric-Learning Correspondence Model

TL;DR

This paper introduces a novel geometric-learning pipeline that detects missing tissues in intra-patient structures by analyzing prediction discrepancies, achieving high accuracy in simulated data and promising results in real cases.

Contribution

It is the first to use geometric-learning models for identifying missing points in anatomical structures, improving dose mapping in reirradiation.

Findings

Balanced accuracy of 0.883 on simulated data

Effective in cases with ~25% tissue removal

Failed on cases with ~50% tissue removal

Abstract

Missing tissue presents a big challenge for dose mapping, e.g., in the reirradiation setting. We propose a pipeline to identify missing tissue on intra-patient structure meshes using a previously trained geometric-learning correspondence model. For our application, we relied on the prediction discrepancies between forward and backward correspondences of the input meshes, quantified using a correspondence-based Inverse Consistency Error (cICE). We optimised the threshold applied to cICE to identify missing points in a dataset of 35 simulated mandible resections. Our identified threshold, 5.5 mm, produced a balanced accuracy score of 0.883 in the training data, using an ensemble approach. This pipeline produced plausible results for a real case where ~25% of the mandible was removed after a surgical intervention. The pipeline, however, failed on a more extreme case where ~50% of the mandible was removed. This is the first time geometric-learning modelling is proposed to identify missing points in corresponding anatomy.