Diffusion-based Radiotherapy Dose Prediction Guided by Inter-slice Aware Structure Encoding

TL;DR

This paper introduces DiffDose, a diffusion model-based approach for radiotherapy dose prediction that overcomes over-smoothing issues of traditional methods, leading to more accurate and detailed dose distribution maps.

Contribution

The novel DiffDose model applies diffusion processes to radiotherapy dose prediction, improving detail preservation over existing deep learning methods.

Findings

Outperforms traditional methods in dose prediction accuracy

Reduces over-smoothing artifacts in predicted dose maps

Demonstrates effectiveness on clinical datasets

Abstract

Deep learning (DL) has successfully automated dose distribution prediction in radiotherapy planning, enhancing both efficiency and quality. However, existing methods suffer from the over-smoothing problem for their commonly used L1 or L2 loss with posterior average calculations. To alleviate this limitation, we propose a diffusion model-based method (DiffDose) for predicting the radiotherapy dose distribution of cancer patients. Specifically, the DiffDose model contains a forward process and a reverse process. In the forward process, DiffDose transforms dose distribution maps into pure Gaussian noise by gradually adding small noise and a noise predictor is simultaneously trained to estimate the noise added at each timestep. In the reverse process, it removes the noise from the pure Gaussian noise in multiple steps with the well-trained noise predictor and finally outputs the predicted dose distribution maps...