Winners of the 2024 Tuberculosis Detection Challenge

Congratulations to the winners of the 2024 Detecting Active Tuberculosis Bacilli Data Challenge. The goal of this challenge is to detect the presence of TB bacilli (Mycobacterium tuberculosis) from digitized images of sputum smear slides. The winning solutions are published as open source by their authors to promote collaborative and repeatable research.

‍

Winner, $3,000 cash prize

Team name: NeuraBB
‍Team members: Zsolt Bedőházi^1,2, András M. Biricz²
Organization:
¹ ELTE Eötvös Loránd University, Budapest, Hungary, Doctoral School of Informatics,
² ELTE Eötvös Loránd University, Budapest, Hungary, Department of Complex Systems in Physics
‍Code on GitHub‍

‍

Second Place, $2,000 cash prize

Team name: Agaldran-tuberculosis-detection
Team Members: Adrian Galdran
Organization: Universitat Pompeu Fabra, Barcelona, Spain
‍Code on GitHub

‍

Honorable Mentions

Third Place

Team name: Bonaventure F. P. Dossou
Team members: Bonaventure Dossou
Organization: McGill University, Mila Quebec AI Institute
‍Code on GitHub

Other Qualifying Teams

The following teams had submissions that surpassed the baseline score for the contest.

PKU-Edinburgh
TB_abnormal_docs

‍

Thanks to the data provider ‍Wellgen Medical and to Gordon and Betty Moore Foundation's Diagnostic Excellence Initiative for the funding support to publish this dataset.

Congratulations to the winners!

‍