Adaptive Frequency Learning Network With Anti-Aliasing Complex Convolutions for Colon Diseases Subtypes

The automatic and dependable identification of colonic disease subtypes by colonoscopy is crucial. Once successful, it will facilitate clinically more in-depth disease staging analysis and the formulation of more tailored treatment plans. However, inter-class confusion and brightness imbalance are major obstacles to colon disease subtyping. Notably, the Fourier-based image spectrum, with its distinctive frequency features and brightness insensitivity, offers a potential solution. To effectively leverage its advantages to address the existing challenges, this article proposes a framework capable of thorough learning in the frequency domain based on four core designs: the position consistency module, the high-frequency self-supervised module, the complex number arithmetic model, and the feature anti-aliasing module. The position consistency module enables the generation of spectra that preserve local and positional information while compressing the spectral data range to improve training stability. Through band masking and supervision, the high-frequency autoencoder module guides the network to learn useful frequency features selectively. The proposed complex number arithmetic model allows direct spectral training while avoiding the loss of phase information caused by current general-purpose real-valued operations. The feature anti-aliasing module embeds filters in the model to prevent spectral aliasing caused by down-sampling and improve performance. Experiments are performed on the collected five-class dataset, which contains 4591 colorectal endoscopic images. The outcomes show that our proposed method produces state-of-the-art results with an accuracy rate of 89.82%.

I. Introduction

The automated diagnosis of multi-colonic diseases is crucial in promoting effective treatment to decrease mortality and improve prognosis [1]. Colorectal cancer often has few obvious symptoms in its early stages, leading to late diagnoses and high mortality rates [2]. for diagnosing colorectal lesions in clinical practice, and if precancerous lesions can be identified and removed through endoscopic mucosal resection, the incidence of colon cancer can be significantly reduced [3]. However, manual lesion classification by endoscopists is frequently subjective and time-consuming. In this regard, the automatic classification of colorectal lesions from colonoscopy images is essential for clinical analysis because it: 1) assists physicians in identifying the type of colonic disease; 2) determines the most effective treatment.