Precision Cell Detection and Counting In Adhesion Experiments: A Deep Learning Perspective With YOLO. Annotated Dataset

Abstract

Cell adhesion is a fundamental biological process underpinning various physiological and pathological phenomena, including tissue repair and cancer metastasis. While straightforward, traditional assays for assessing cell adhesion suffer from poor reproducibility and low throughput. This study introduces a deep learning-based approach using the You Only Look Once (YOLO) convolutional neural networks to automate cell detection and counting, even in real-time, thereby improving the speed and efficiency of cell adhesion assays. Our methodology involved the analysis of AQP2-RCCD1 cell adhesion assays with the data captured and processed using the YOLO models. These models were trained on various image resolutions to assess the trade-offs between image quality and computational efficiency, significantly optimizing the detection process. Employing the YOLOv3, YOLOv5, YOLOv8, and YOLOv9 architectures, we address the challenges of variability in cell density and illumination within adhesion experiments. In commitment to open science principles, the source code, the trained models, and our real-time webcam analysis approach are shared to foster innovation and collaboration. Our findings highlight the potential of using YOLO models for efficient and accurate cell analysis, making advanced image processing accessible to a broader range of researchers.