Hepatocellular Carcinoma Segmentation from Digital Subtraction Angiography Videos using Learnable Temporal Difference

TL;DR

This paper introduces DSA-LTDNet, a novel deep learning model that leverages temporal difference learning for improved segmentation of hepatocellular carcinoma in DSA videos, addressing motion artifacts and ambiguous boundaries.

Contribution

The paper presents a new segmentation network with a temporal difference learning module and a dedicated liver region sub-network, tailored for DSA video analysis.

Findings

DSA-LTDNet improves DICE score by nearly 4% over U-Net baseline.

The model effectively captures motion information in DSA videos.

Experimental results validate the approach on a self-collected dataset.

Abstract

Automatic segmentation of hepatocellular carcinoma (HCC) in Digital Subtraction Angiography (DSA) videos can assist radiologists in efficient diagnosis of HCC and accurate evaluation of tumors in clinical practice. Few studies have investigated HCC segmentation from DSA videos. It shows great challenging due to motion artifacts in filming, ambiguous boundaries of tumor regions and high similarity in imaging to other anatomical tissues. In this paper, we raise the problem of HCC segmentation in DSA videos, and build our own DSA dataset. We also propose a novel segmentation network called DSA-LTDNet, including a segmentation sub-network, a temporal difference learning (TDL) module and a liver region segmentation (LRS) sub-network for providing additional guidance. DSA-LTDNet is preferable for learning the latent motion information from DSA videos proactively and boosting segmentation performance. All of experiments are conducted on our self-collected dataset. Experimental results show that DSA-LTDNet increases the DICE score by nearly 4% compared to the U-Net baseline.