Assessing deep learning methods for the identification of kidney stones in endoscopic images

TL;DR

This paper compares deep learning and traditional methods for identifying kidney stone types from in vivo endoscopic images, aiming to enable real-time, non-invasive diagnosis during procedures.

Contribution

It evaluates and compares five classification methods, demonstrating that a well-designed XGBoost approach can nearly match deep learning performance with limited data.

Findings

DCNN achieved 98% precision and 97% recall

XGBoost closely approaches DCNN performance

Traditional methods can be effective with limited data

Abstract

Knowing the type (i.e., the biochemical composition) of kidney stones is crucial to prevent relapses with an appropriate treatment. During ureteroscopies, kidney stones are fragmented, extracted from the urinary tract, and their composition is determined using a morpho-constitutional analysis. This procedure is time consuming (the morpho-constitutional analysis results are only available after some days) and tedious (the fragment extraction lasts up to an hour). Identifying the kidney stone type only with the in-vivo endoscopic images would allow for the dusting of the fragments, while the morpho-constitutional analysis could be avoided. Only few contributions dealing with the in vivo identification of kidney stones were published. This paper discusses and compares five classification methods including deep convolutional neural networks (DCNN)-based approaches and traditional (non DCNN-based) ones. Even if the best method is a DCCN approach with a precision and recall of 98% and 97% over four classes, this contribution shows that a XGBoost classifier exploiting well-chosen feature vectors can closely approach the performances of DCNN classifiers for a medical application with a limited number of annotated data.