Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

MD-Dose: A diffusion model based on the Mamba for radiation dose prediction

Linjie Fu, Xia Li, Xiuding Cai, Yingkai Wang, Xueyao Wang, Yali Shen, Yu Yao

TL;DR

MD-Dose presents a diffusion-model-based approach for radiation-dose prediction that leverages the Mamba architecture to balance global context and computational efficiency. By integrating a Mamba-UNet denoising network with a structure encoder, the method generates high-frequency dose details while respecting anatomical constraints between the PTV and OARs. Empirical results on 300 thoracic-tumor patients show MD-Dose achieves superior Dose Score, DVH Score, and HI compared with state-of-the-art diffusion models, with favorable inference times and fewer parameters. The work demonstrates the practical potential of Mamba-based diffusion models to accelerate radiotherapy planning and improve plan quality, with ablations confirming the importance of the Mamba blocks and structural guidance. Future work will broaden validation to other cancer sites and continue optimizing the Mamba-based components for clinical deployment.

Abstract

Radiation therapy is crucial in cancer treatment. Experienced experts typically iteratively generate high-quality dose distribution maps, forming the basis for excellent radiation therapy plans. Therefore, automated prediction of dose distribution maps is significant in expediting the treatment process and providing a better starting point for developing radiation therapy plans. With the remarkable results of diffusion models in predicting high-frequency regions of dose distribution maps, dose prediction methods based on diffusion models have been extensively studied. However, existing methods mainly utilize CNNs or Transformers as denoising networks. CNNs lack the capture of global receptive fields, resulting in suboptimal prediction performance. Transformers excel in global modeling but face quadratic complexity with image size, resulting in significant computational overhead. To tackle these challenges, we introduce a novel diffusion model, MD-Dose, based on the Mamba architecture for predicting radiation therapy dose distribution in thoracic cancer patients. In the forward process, MD-Dose adds Gaussian noise to dose distribution maps to obtain pure noise images. In the backward process, MD-Dose utilizes a noise predictor based on the Mamba to predict the noise, ultimately outputting the dose distribution maps. Furthermore, We develop a Mamba encoder to extract structural information and integrate it into the noise predictor for localizing dose regions in the planning target volume (PTV) and organs at risk (OARs). Through extensive experiments on a dataset of 300 thoracic tumor patients, we showcase the superiority of MD-Dose in various metrics and time consumption.

MD-Dose: A diffusion model based on the Mamba for radiation dose prediction

TL;DR

MD-Dose presents a diffusion-model-based approach for radiation-dose prediction that leverages the Mamba architecture to balance global context and computational efficiency. By integrating a Mamba-UNet denoising network with a structure encoder, the method generates high-frequency dose details while respecting anatomical constraints between the PTV and OARs. Empirical results on 300 thoracic-tumor patients show MD-Dose achieves superior Dose Score, DVH Score, and HI compared with state-of-the-art diffusion models, with favorable inference times and fewer parameters. The work demonstrates the practical potential of Mamba-based diffusion models to accelerate radiotherapy planning and improve plan quality, with ablations confirming the importance of the Mamba blocks and structural guidance. Future work will broaden validation to other cancer sites and continue optimizing the Mamba-based components for clinical deployment.

Abstract

Radiation therapy is crucial in cancer treatment. Experienced experts typically iteratively generate high-quality dose distribution maps, forming the basis for excellent radiation therapy plans. Therefore, automated prediction of dose distribution maps is significant in expediting the treatment process and providing a better starting point for developing radiation therapy plans. With the remarkable results of diffusion models in predicting high-frequency regions of dose distribution maps, dose prediction methods based on diffusion models have been extensively studied. However, existing methods mainly utilize CNNs or Transformers as denoising networks. CNNs lack the capture of global receptive fields, resulting in suboptimal prediction performance. Transformers excel in global modeling but face quadratic complexity with image size, resulting in significant computational overhead. To tackle these challenges, we introduce a novel diffusion model, MD-Dose, based on the Mamba architecture for predicting radiation therapy dose distribution in thoracic cancer patients. In the forward process, MD-Dose adds Gaussian noise to dose distribution maps to obtain pure noise images. In the backward process, MD-Dose utilizes a noise predictor based on the Mamba to predict the noise, ultimately outputting the dose distribution maps. Furthermore, We develop a Mamba encoder to extract structural information and integrate it into the noise predictor for localizing dose regions in the planning target volume (PTV) and organs at risk (OARs). Through extensive experiments on a dataset of 300 thoracic tumor patients, we showcase the superiority of MD-Dose in various metrics and time consumption.
Paper Structure (17 sections, 10 equations, 4 figures, 4 tables)

This paper contains 17 sections, 10 equations, 4 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Methodology
  3. Score-based Diffusion Generative Models
  4. Forward Process
  5. Reverse Process
  6. Objective Function
  7. Mamba-based Denoising Network
  8. SSM Layer
  9. Mamba Block
  10. Structure Encoder
  11. Experiments
  12. Datasets and Evaluation Metrics
  13. Training Details
  14. Comparison with State-Of-The-Art Methods
  15. Ablation Study
  16. ...and 2 more sections

Figures (4)

  • Figure 1: Demonstrate a radiation therapy plan using beam-shaped radiation.
  • Figure 2: The overview of the proposed MD-Dose, including (a) the overall structure of MD-Dose, encompassing both the forward and backward processes of the diffusion model; (b) the proposed Mamba-UNet; (c) the proposed Structure Encoder; (d) the holistic architecture of the Mamba Block.
  • Figure 3: Visual comparisons with state-of-the-art (SOTA) methods include two sets. The first and third rows illustrate predicted dose distribution maps, and the second and fourth rows display maps depicting dose errors. The last column represents the ground truth.
  • Figure 4: Visualize the DVH curves of all methods, encompassing curves for the PTV, Heart, Lung, and Spinal Cord.