Deep ensemble learning for segmenting tuberculosis-consistent manifestations in chest radiographs

TL;DR

This paper explores the use of ensemble deep learning models, specifically stacking, with fine-grained annotations to improve the accuracy of segmenting tuberculosis lesions in chest X-rays, outperforming individual models.

Contribution

It introduces the first application of ensemble learning, particularly stacking, for fine-grained TB lesion segmentation in chest radiographs, demonstrating improved performance over single models.

Findings

Stacking ensemble achieved the highest Dice score of 0.5743.

Fine-grained annotations improve segmentation accuracy.

Ensemble methods outperform individual U-Net variants.

Abstract

Automated segmentation of tuberculosis (TB)-consistent lesions in chest X-rays (CXRs) using deep learning (DL) methods can help reduce radiologist effort, supplement clinical decision-making, and potentially result in improved patient treatment. The majority of works in the literature discuss training automatic segmentation models using coarse bounding box annotations. However, the granularity of the bounding box annotation could result in the inclusion of a considerable fraction of false positives and negatives at the pixel level that may adversely impact overall semantic segmentation performance. This study (i) evaluates the benefits of using fine-grained annotations of TB-consistent lesions and (ii) trains and constructs ensembles of the variants of U-Net models for semantically segmenting TB-consistent lesions in both original and bone-suppressed frontal CXRs. We evaluated segmentation performance using several ensemble methods such as bitwise AND, bitwise-OR, bitwise-MAX, and stacking. We observed that the stacking ensemble demonstrated superior segmentation performance (Dice score: 0.5743, 95% confidence interval: (0.4055,0.7431)) compared to the individual constituent models and other ensemble methods. To the best of our knowledge, this is the first study to apply ensemble learning to improve fine-grained TB-consistent lesion segmentation performance.