Detection of Virus and Small Cell Patches in Foci Images Using Switchable Convolution and Feature Pyramid Networks

2026-05-21

AI Summary

This work addresses the challenge of detecting viral and small cellular plaques in focus-forming assay images, where significant variations in scale, density, contrast, and morphology hinder accurate identification. To overcome these difficulties, the authors propose an enhanced YOLOv2-based detector that integrates a Feature Pyramid Network (FPN) to strengthen multi-scale feature representation and incorporates a switchable atrous convolution mechanism to adaptively adjust the receptive field for fine-grained targets in high-density microscopic imagery. The proposed method substantially improves YOLOv2’s performance on biomedical images containing small, densely packed objects, achieving mean average precisions (mAP) of 68% and 40.5% at an IoU threshold of 0.25 for viral and small cellular plaque detection tasks, respectively, thereby demonstrating its effectiveness and innovation in specialized biomedical detection scenarios.

Abstract

Accurate detection and counting of virus patches in focus-forming unit (FFU) images, also known as foci images, are important for quantifying viral infection and analyzing cellular structures. This task is challenging because biomedical targets often vary substantially in size, density, contrast, and shape. In this paper, we propose an enhanced YOLOv2-based detector that integrates a Feature Pyramid Network (FPN) to improve multi-scale feature representation. We also incorporate a switchable atrous convolution mechanism to adapt the receptive field for fine-grained targets in dense microscopy images. The proposed method is evaluated on biomedical foci image datasets for virus patch and small cell patch detection. For small cell patch detection, the model achieves a mean average precision (mAP) of 40.5% at a 25% Intersection over Union (IoU) threshold. For FFU virus patch detection, the model achieves an mAP of 68%. These results indicate that combining FPN-based feature fusion with switchable convolution improves the suitability of YOLOv2 for specialized biomedical object detection tasks

Problem

• virus detection
• small cell patches
• foci images
• biomedical object detection
• multi-scale targets

Innovation

• Switchable Atrous Convolution
• Feature Pyramid Network
• YOLOv2
• Multi-scale Feature Representation
• Biomedical Object Detection