Classification of Informative Frames in Colonoscopy Videos Using Convolutional Neural Networks with Binarized Weights

TL;DR

This paper introduces a CNN-based method with binarized weights for classifying informative frames in colonoscopy videos, aiming to improve early polyp detection for colorectal cancer treatment.

Contribution

It presents a novel CNN architecture with binarized weights for efficient classification of colonoscopy frames, suitable for medical hardware implementation.

Findings

Achieved a dice score of 71.20% on the dataset.

Attained over 90% accuracy in frame classification.

Reduced CNN size for hardware deployment.

Abstract

Colorectal cancer is one of the common cancers in the United States. Polyp is one of the main causes of the colonic cancer and early detection of polyps will increase chance of cancer treatments. In this paper, we propose a novel classification of informative frames based on a convolutional neural network with binarized weights. The proposed CNN is trained with colonoscopy frames along with the labels of the frames as input data. We also used binarized weights and kernels to reduce the size of CNN and make it suitable for implementation in medical hardware. We evaluate our proposed method using Asu Mayo Test clinic database, which contains colonoscopy videos of different patients. Our proposed method reaches a dice score of 71.20% and accuracy of more than 90% using the mentioned dataset.