A newly developed filtering scheme for confocal laser endomicroscopy (CLE) video sequences is designed to enhance self-supervised learning, facilitating the interpretation of complex images by physicians. This approach focuses on improving the analysis of mucous structures captured in CLE videos, which are often challenging to assess. By leveraging machine learning techniques, the filtering scheme aims to streamline the diagnostic process, potentially making it more efficient and accurate. The innovation is expected to assist medical professionals by providing clearer, more interpretable visual data. According to the available evidence, this method plausibly improves the diagnostic workflow in clinical settings. Overall, the scheme represents a promising advancement in the application of artificial intelligence to medical imaging, particularly in endomicroscopy.