Automatic semantic segmentation for prediction of tuberculosis using lens-free microscopy images

TL;DR

This paper presents an automated method using a U-Net neural network to segment tuberculosis cords in lens-free microscopy images, aiming to improve TB diagnosis accessibility and ease of use.

Contribution

It introduces a novel application of U-Net for TB segmentation in lens-free microscopy images, enhancing diagnostic automation and potential deployment in resource-limited settings.

Findings

Initial results show promising segmentation accuracy.

Demonstrates feasibility of automated TB detection with lens-free microscopy.

Potential to assist untrained personnel in TB diagnosis.

Abstract

Tuberculosis (TB), caused by a germ called Mycobacterium tuberculosis, is one of the most serious public health problems in Peru and the world. The development of this project seeks to facilitate and automate the diagnosis of tuberculosis by the MODS method and using lens-free microscopy, due they are easier to calibrate and easier to use (by untrained personnel) in comparison with lens microscopy. Thus, we employ a U-Net network in our collected dataset to perform the automatic segmentation of the TB cords in order to predict tuberculosis. Our initial results show promising evidence for automatic segmentation of TB cords.