[Paper Review] FCB-SwinV2 Transformer for Polyp Segmentation

Kerr Fitzgerald, B.K. Matuszewski|arXiv (Cornell University)|Feb 2, 2023

Colorectal Cancer Screening and Detection12 citations

TL;DR

This paper modifies FCBFormer by replacing its transformer with a SwinV2 Transformer-UNET and slight FCN tweaks, achieving state-of-the-art polyp segmentation on Kvasir-SEG and CVC-ClinicDB with better mDice scores and highlighting data-splitting issues.

ABSTRACT

Polyp segmentation within colonoscopy video frames using deep learning models has the potential to automate the workflow of clinicians. This could help improve the early detection rate and characterization of polyps which could progress to colorectal cancer. Recent state-of-the-art deep learning polyp segmentation models have combined the outputs of Fully Convolutional Network architectures and Transformer Network architectures which work in parallel. In this paper we propose modifications to the current state-of-the-art polyp segmentation model FCBFormer. The transformer architecture of the FCBFormer is replaced with a SwinV2 Transformer-UNET and minor changes to the Fully Convolutional Network architecture are made to create the FCB-SwinV2 Transformer. The performance of the FCB-SwinV2 Transformer is evaluated on the popular colonoscopy segmentation bench-marking datasets Kvasir-SEG and CVC-ClinicDB. Generalizability tests are also conducted. The FCB-SwinV2 Transformer is able to consistently achieve higher mDice scores across all tests conducted and therefore represents new state-of-the-art performance. Issues found with how colonoscopy segmentation model performance is evaluated within literature are also re-ported and discussed. One of the most important issues identified is that when evaluating performance on the CVC-ClinicDB dataset it would be preferable to ensure no data leakage from video sequences occurs during the training/validation/test data partition.

Motivation & Objective

• Motivate automated polyp segmentation to assist early colorectal cancer detection.
• Improve segmentation performance by integrating a SwinV2 Transformer-UNET into the FCBFormer framework.
• Assess generalization and data-splitting integrity on benchmark datasets.

Proposed method

• Replace the transformer architecture in FCBFormer with a SwinV2 Transformer-UNET.
• Apply minor modifications to the Fully Convolutional Network components.
• Evaluate the model on Kvasir-SEG and CVC-ClinicDB datasets.
• Perform generalizability tests across datasets and discuss data leakage concerns in evaluation.

Experimental results

Research questions

• RQ1 Can a SwinV2 Transformer-UNET integration improve polyp segmentation performance over the original FCBFormer?
• RQ2 Does the FCB-SwinV2 Transformer generalize across standard colonoscopy segmentation benchmarks?
• RQ3 What evaluation pitfalls (e.g., data leakage) affect reported performance on CVC-ClinicDB?
• RQ4 How do minor FCN modifications interact with transformer-based segmentation in this task?

Key findings

• The FCB-SwinV2 Transformer consistently achieves higher mDice scores than prior approaches on the tested benchmarks.
• The model demonstrates state-of-the-art performance on Kvasir-SEG and CVC-ClinicDB according to reported metrics.
• Generalizability tests support robust performance across datasets.
• The study identifies and discusses evaluation issues, including potential data leakage in CVC-ClinicDB assessments.