CNN-Based Automatic Urinary Particles Recognition

TL;DR

This paper applies CNN-based object detection methods, Faster R-CNN and SSD, to automatically recognize urine particles in microscopic images, achieving high accuracy and efficiency for medical diagnostics.

Contribution

It introduces a CNN-based end-to-end approach for urine particle recognition, replacing traditional hand-crafted feature methods with deep learning techniques.

Findings

Achieved a maximum mAP of 84.1% on urine particle dataset.

Processed images in 72 ms on NVIDIA Titan X GPU.

Compared and evaluated different CNN-based detection methods.

Abstract

The urine sediment analysis of particles in microscopic images can assist physicians in evaluating patients with renal and urinary tract diseases. Manual urine sediment examination is labor-intensive, subjective and time-consuming, and the traditional automatic algorithms often extract the hand-crafted features for recognition. Instead of using the hand-crafted features, in this paper, we exploit CNN to learn features in an end-to-end manner to recognize the urine particles. We treat the urine particles recognition as object detection and exploit two state-of-the-art CNN-based object detection methods, Faster R-CNN and SSD, as well as their variants for urine particles recognition. We further investigate different factors involving these CNN-based object detection methods for urine particles recognition. We comprehensively evaluate these methods on a dataset consisting of 5,376 annotated images corresponding to 7 categories of urine particles, i.e., erythrocyte, leukocyte, epithelial cell, crystal, cast, mycete, epithelial nuclei, and obtain a best mAP (mean average precision) of 84.1% while taking only 72 ms per image on a NVIDIA Titan X GPU.