Blog - Latest News

Article of the week: Deep learning computer vision algorithm for detecting kidney stone composition

Every week, the Editor-in-Chief selects an Article of the Week from the current issue of BJUI. The abstract is reproduced below and you can click on the button to read the full article, which is freely available to all readers for at least 30 days from the time of this post.

If you only have time to read one article this week, we recommend this one. 

Deep learning computer vision algorithm for detecting kidney stone composition

Kristian M. Black*, Hei Law, Ali Aldoukhi*, Jia Deng and Khurshid R. Ghani*

*Department of Urology, University of Michigan, Ann Arbor, MI, and Department of Computer Science, Princeton University, Princeton, NJ, USA



To assess the recall of a deep learning (DL) method to automatically detect kidney stones composition from digital photographs of stones.

Materials and Methods

A total of 63 human kidney stones of varied compositions were obtained from a stone laboratory including calcium oxalate monohydrate (COM), uric acid (UA), magnesium ammonium phosphate hexahydrate (MAPH/struvite), calcium hydrogen phosphate dihydrate (CHPD/brushite), and cystine stones. At least two images of the stones, both surface and inner core, were captured on a digital camera for all stones. A deep convolutional neural network (CNN), ResNet‐101 (ResNet, Microsoft), was applied as a multi‐class classification model, to each image. This model was assessed using leave‐one‐out cross‐validation with the primary outcome being network prediction recall.

Fig.2. Representative samples for each stone composition prior to cropping. A total of 63 stones were used in this study including: 17 UA, 21 COM, seven struvite, four cystine, and 14 brushite stones consisting of a total of 127 images. Automatic stone composition recall was highest for UA stones at 94%.


The composition prediction recall for each composition was as follows: UA 94% ( = 17), COM 90% ( = 21), MAPH/struvite 86% ( = 7), cystine 75% ( = 4), CHPD/brushite 71% ( = 14). The overall weighted recall of the CNNs composition analysis was 85% for the entire cohort. Specificity and precision for each stone type were as follows: UA (97.83%, 94.12%), COM (97.62%, 95%), struvite (91.84%, 71.43%), cystine (98.31%, 75%), and brushite (96.43%, 75%).


Deep CNNs can be used to identify kidney stone composition from digital photographs with good recall. Future work is needed to see if DL can be used for detecting stone composition during digital endoscopy. This technology may enable integrated endoscopic and laser systems that automatically provide laser settings based on stone composition recognition with the goal to improve surgical efficiency.

1 reply
  1. Cblm Holy Family Hospital
    Cblm Holy Family Hospital says:

    You have written quite an impressive and well-detailed article for the kidney patient. The points that have been mentioned here will surely help the kidney patient enhance their self confidence with the knowledge which you have mentioned in above blog.
    A kidney stone is a condition that results due to excess building up of dissolved minerals in the inner lining of kidneys. These with time grow in size forming a crystalline structure. CBLM is one of the top listed multispecialty hospitals in Jaipur which offers world-class treatment and surgery for kidney stones in India. Backed by a team of expert neurologists the hospital assists with appropriate diagnosis and treatment.

Comments are closed.

© 2024 BJU International. All Rights Reserved.