Automatic Polyp Segmentation Using Modified Recurrent Residual Unet Network

Abstract

Colorectal cancer is a dangerous disease with a high mortality rate. To increase the likelihood of successful treatment, early detection of polyps is a useful solution. The Unet-architecture network model is showing success in medical image segmentation including analysis of polyps from colonoscopy images. Traditional Unet and Unet-based models are often huge, requiring training and deployment with a highperformance system. Designing models with compact size and high performance would be an important goal. In this study, we proposed to modify the Residual Recurrent Unet architecture to improve the size of the model while ensuring the model performance. The proposed model has flexibility in changing the number of filters in convolution units. By taking advantage of the strengths of residual and recurrent structures in terms of reuse of convolutional functions, the new model, therefore, was not only smaller in size but also has superior performance compared to the traditional Unet model and the others. The evaluations were performed on three public Colonoscopy image datasets: CVC-ClinicDB, ETIS-LaribPolypDB, and CVC-ColonDB. The Dice score on CVC-ClinicDB reached 94.59%, ETIS-LaribPolypDB reached 92.73% and 93.31% on CVC-ColonDB dataset. The experimental results obtained from the proposed network on datasets were better than those in recent related studies. The introduced model has a smaller size than the traditional model nevertheless has outstanding performance, therefore, it would be extremely productive for developing applications on low-performance devices.

MGF‐net: Multi‐channel group fusion enhancing boundary attention for polyp segmentation

“…In order to further quantitatively analyze the segmentation performance of MGF‐Net, this chapter also introduced ResUnet++, 16 SFANet, 39 ACS‐Net, 40 EU‐Net, 41 TMDUNet, 42 and APSUNet 43 for numerical comparison. Table 1 presents the quantitative results of MGF‐Net and mainstream methods on the Kvasir and CVC‐ClinicDB datasets, where the best metric is marked in red.…”

CRPU-NET: a deep learning model based semantic segmentation for the detection of colorectal polyp in lower gastrointestinal tract

“…Furthermore, the IARC estimates that the number of deaths resulting from CRC will also experience a substantial rise, reaching approximately 1.6 million fatalities per year. This represents a significant increase of 73% CRC rate compared to the current statistics [2,3].…”

“…The CRC mostly originate from innocuous cell growth called polyp on the mucosal surface of the colon or rectum with the symptoms of abdominal pain or cramping, bloating or distention, nausea or vomiting, diarrhoea, constipation, unintended weight loss, rectal bleeding or blood in stools, difficulty swallowing and frequent heartburn or acid reflux [3,4]. The colorectal polyp (CRP) can exhibit two distinct forms of attachment: sessile, where the base is directly connected to the mucous, or pedunculated, where it is attached via a stalk.…”

Automated Detection of Colorectal Polyp Utilizing Deep Learning Methods With Explainable AI

“…A substantial amount of effort has been devoted to creating practical algorithms and methodologies. At first, conventional machine-learning techniques and manual approaches were employed to identify polyps through color and texture analysis [9]- [13]. Furthermore, most research studies have focused solely on the segmentation phase, with little emphasis on the localization phase.…”