Deep learning for automated detection of breast cancer in deep ultraviolet fluorescence images with diffusion probabilistic model

TL;DR

This paper demonstrates that using a diffusion probabilistic model to augment deep ultraviolet fluorescence images significantly enhances breast cancer detection accuracy during intraoperative margin assessment.

Contribution

The study introduces the application of DPM for dataset augmentation in medical imaging, improving classification accuracy over traditional augmentation methods.

Findings

Augmentation with DPM increased accuracy from 93% to 97%.

DPM outperformed Affine transformations and ProGAN in dataset augmentation.

Patch-level features combined with Grad-CAM++ improved whole surface prediction.

Abstract

Data limitation is a significant challenge in applying deep learning to medical images. Recently, the diffusion probabilistic model (DPM) has shown the potential to generate high-quality images by converting Gaussian random noise into realistic images. In this paper, we apply the DPM to augment the deep ultraviolet fluorescence (DUV) image dataset with an aim to improve breast cancer classification for intraoperative margin assessment. For classification, we divide the whole surface DUV image into small patches and extract convolutional features for each patch by utilizing the pre-trained ResNet. Then, we feed them into an XGBoost classifier for patch-level decisions and then fuse them with a regional importance map computed by Grad-CAM++ for whole surface-level prediction. Our experimental results show that augmenting the training dataset with the DPM significantly improves breast cancer detection performance in DUV images, increasing accuracy from 93% to 97%, compared to using Affine transformations and ProGAN.