FCN-Transformer Feature Fusion for Polyp Segmentation

TL;DR

This paper introduces a novel FCN-Transformer architecture for full-size polyp segmentation in colonoscopy images, combining transformer and convolutional features to improve accuracy and generalization over existing methods.

Contribution

The proposed architecture effectively fuses transformer and convolutional features for full-size segmentation, achieving state-of-the-art results and better generalization in polyp segmentation tasks.

Findings

State-of-the-art performance on Kvasir-SEG and CVC-ClinicDB datasets.

Superior generalization demonstrated by cross-dataset evaluation.

Improved metrics such as mDice, mIoU, mPrecision, and mRecall.

Abstract

Colonoscopy is widely recognised as the gold standard procedure for the early detection of colorectal cancer (CRC). Segmentation is valuable for two significant clinical applications, namely lesion detection and classification, providing means to improve accuracy and robustness. The manual segmentation of polyps in colonoscopy images is time-consuming. As a result, the use of deep learning (DL) for automation of polyp segmentation has become important. However, DL-based solutions can be vulnerable to overfitting and the resulting inability to generalise to images captured by different colonoscopes. Recent transformer-based architectures for semantic segmentation both achieve higher performance and generalise better than alternatives, however typically predict a segmentation map of $\frac{h}{4}\times\frac{w}{4}$ spatial dimensions for a $h\times w$ input image. To this end, we propose a new architecture for full-size segmentation which leverages the strengths of a transformer in extracting the most important features for segmentation in a primary branch, while compensating for its limitations in full-size prediction with a secondary fully convolutional branch. The resulting features from both branches are then fused for final prediction of a $h\times w$ segmentation map. We demonstrate our method's state-of-the-art performance with respect to the mDice, mIoU, mPrecision, and mRecall metrics, on both the Kvasir-SEG and CVC-ClinicDB dataset benchmarks. Additionally, we train the model on each of these datasets and evaluate on the other to demonstrate its superior generalisation performance.