The current screening pipeline for Tuberculosis(TB) screening has many challenges including the requirement of huge infrastructure for X-ray testing, delay due to different tests and drop-off due to the complicated process. It is also not pointof- care and requires involvement of experts. In this work, we propose to build a much safer and portable ultrasound-based point-of-care screening tool using AI. Our proposed AI method not only surpasses radiologists in classifying the ultrasound-videos with a large margin but also provides a proper explanation to its outcomes.
Supplementary notes can be added here, including code, math, and images.