Intentional Deep Overfit Learning (IDOL): A Novel Deep Learning Strategy for Adaptive Radiation Therapy

TL;DR

This paper introduces IDOL, a deep learning framework that intentionally overfits models to patient-specific data to improve performance in adaptive radiation therapy tasks such as auto-contouring, MRI super-resolution, and synthetic CT reconstruction.

Contribution

The study presents a novel, task-agnostic deep learning approach that enhances patient-specific accuracy in radiotherapy workflows by intentionally overfitting models to personalized data.

Findings

Auto-contouring Dice coefficient improved from 0.847 to 0.935.

MRI super-resolution MAE reduced by 40%.

Synthetic CT MAE decreased from 68 to 22 HU.

Abstract

In this study, we propose a tailored DL framework for patient-specific performance that leverages the behavior of a model intentionally overfitted to a patient-specific training dataset augmented from the prior information available in an ART workflow - an approach we term Intentional Deep Overfit Learning (IDOL). Implementing the IDOL framework in any task in radiotherapy consists of two training stages: 1) training a generalized model with a diverse training dataset of N patients, just as in the conventional DL approach, and 2) intentionally overfitting this general model to a small training dataset-specific the patient of interest (N+1) generated through perturbations and augmentations of the available task- and patient-specific prior information to establish a personalized IDOL model. The IDOL framework itself is task-agnostic and is thus widely applicable to many components of the ART workflow, three of which we use as a proof of concept here: the auto-contouring task on re-planning CTs for traditional ART, the MRI super-resolution (SR) task for MRI-guided ART, and the synthetic CT (sCT) reconstruction task for MRI-only ART. In the re-planning CT auto-contouring task, the accuracy measured by the Dice similarity coefficient improves from 0.847 with the general model to 0.935 by adopting the IDOL model. In the case of MRI SR, the mean absolute error (MAE) is improved by 40% using the IDOL framework over the conventional model. Finally, in the sCT reconstruction task, the MAE is reduced from 68 to 22 HU by utilizing the IDOL framework.