TOD-CNN: An effective convolutional neural network for tiny object detection in sperm videos

Comput Biol Med. 2022 Jul:146:105543. doi: 10.1016/j.compbiomed.2022.105543. Epub 2022 Apr 22.

Abstract

The detection of tiny objects in microscopic videos is a problematic point, especially in large-scale experiments. For tiny objects (such as sperms) in microscopic videos, current detection methods face challenges in fuzzy, irregular, and precise positioning of objects. In contrast, we present a convolutional neural network for tiny object detection (TOD-CNN) with an underlying data set of high-quality sperm microscopic videos (111 videos, > 278,000 annotated objects), and a graphical user interface (GUI) is designed to employ and test the proposed model effectively. TOD-CNN is highly accurate, achieving 85.60% AP50 in the task of real-time sperm detection in microscopic videos. To demonstrate the importance of sperm detection technology in sperm quality analysis, we carry out relevant sperm quality evaluation metrics and compare them with the diagnosis results from medical doctors.

Keywords: Convolutional neural network; Image analysis; Object detection; Sperm microscopy video.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Face*
  • Humans
  • Male
  • Neural Networks, Computer*
  • Spermatozoa