A Study on the Performance of U-Net Modifications in Retroperitoneal Tumor Segmentation
Moein Heidari, Ehsan Khodapanah Aghdam, Alexander Manzella, Daniel Hsu, Rebecca Scalabrino, Wenjin Chen, David J. Foran, Ilker Hacihaliloglu
2025-02-04
Summary
This paper talks about improving how computers can automatically find and measure tumors in a tricky part of the body called the retroperitoneum. The researchers tested different versions of a popular AI tool called U-Net to see which one works best for this job.
What's the problem?
Doctors need to measure tumors in the retroperitoneum accurately, but these tumors are often oddly shaped and close to important organs. Doing this by hand takes a long time, and while some AI methods can help, they often need a lot of computer power to work well.
What's the solution?
The researchers tried out several improved versions of U-Net, including ones that use new AI techniques like Vision Transformers, Mamba State Space Models, and Extended Long-Short Term Memory. They created a new version called ViLU-Net that combines different AI methods. They tested these on both their own collection of CT scans and a public dataset of organ images to see which one worked best.
Why it matters?
This research matters because finding better ways to measure tumors can help doctors plan treatments more accurately and potentially save lives. By making the AI tools more efficient, hospitals might be able to use this technology more easily, even if they don't have super powerful computers. This could lead to faster and more accurate diagnoses for patients with these rare and challenging tumors.
Abstract
The retroperitoneum hosts a variety of tumors, including rare benign and malignant types, which pose diagnostic and treatment challenges due to their infrequency and proximity to vital structures. Estimating tumor volume is difficult due to their irregular shapes, and manual segmentation is time-consuming. Automatic segmentation using U-Net and its variants, incorporating Vision Transformer (ViT) elements, has shown promising results but struggles with high computational demands. To address this, architectures like the Mamba State Space Model (SSM) and Extended Long-Short Term Memory (xLSTM) offer efficient solutions by handling long-range dependencies with lower resource consumption. This study evaluates U-Net enhancements, including CNN, ViT, Mamba, and xLSTM, on a new in-house CT dataset and a public organ segmentation dataset. The proposed ViL<PRE_TAG>U-Net</POST_TAG> model integrates Vi-blocks for improved segmentation. Results highlight xLSTM's efficiency in the U-Net framework. The code is publicly accessible on GitHub.