STOP researchers team up with Spanish start up SpotLab to improve the diagnosis of soil-transmitted helminthiasis

Using smartphhones to detect helminths


Over 1.5 billion people worldwide are estimated to be infected by at least one STH species, which include roundworm (Ascaris lumbricoides), whipworm (Trichuris trichiura) and hookworm (Ancylostoma duodenale and Necator americanus). The diagnosis of STH infections is mostly performed by analysing stool samples under the microscope using the so-called Kato-Katz technique, but this procedure has several disadvantages: egg detection depends on the experience of the person performing the analysis, and the samples must be read within 30 minutes of their preparation.

In this study, a team at SpotLab (a Spanish company dedicated to medical diagnosis) joined efforts with STOP researchers in order to develop and validate an artificial intelligence algorithm capable of detecting and quantifying these intestinal parasites. The proposed system first digitizes microscopy samples using an affordable 3D-printed adapter and smartphones. Then, the digitized images are uploaded to a telemedicine platform to enable remote diagnosis. Lastly, the digitized images are automatically analysed by an artificial intelligence algorithm fully integrated in both the telemedicine platform and mobile app to automatically and objectively count different types of STH eggs. The deep learning algorithm was first trained and tested on 51 slides of stool samples containing 949 Trichuris spp. eggs from 6 different subjects. The results show that the AI algorithm performed very well, with a precision of 98.44% (98% of the selected elements were true positives) and a recall of almost 81% (it recognised 81% of all relevant elements).

“The results suggest that remote and AI-assisted analysis of digitized images allows to detect more eggs compared to the conventional procedure, since the analysis can be performed in a more exhaustive manner,” says Elena Dacal, researcher at SpotLab and first co-author of the study.

“This means that these tools would be potentially useful not only for diagnosing STH at the individual level, but also for evaluating the effectiveness of mass drug administration programmes,” says Jose Muñoz, leader of the STOP project.



Dacal E, Bermejo-Peláez D, Lin L, et al. Mobile microscopy and telemedicine platform assisted by deep learning for the quantification of Trichuris trichiura infection. PLoS Negl Trop Dis 15(9):e0009677.