Dados do Trabalho

Título

Pulmonary Tuberculosis Screening from Radiological Signs on Chest X-Ray Images Using Deep Models

Introdução

The World Health Organization has recently recommended the use of computer-aided detection (CAD) systems for screening pulmonary tuberculosis (PTB) in Chest X-Ray images. Previous CAD models are based on direct image to probability detection techniques – and do not generalize well (from training to validation databases).

Objetivo(s)

We propose a method that overcomes these limitations using radiological signs as intermediary proxies for PTB detection.

Material e Métodos

We developed an open-source multi-class deep learning model, mapping images to 14 radiological signs such as cavities, infiltration, nodules, and fibrosis, using the NIH CXR14 dataset, which contains 112,120 images. Using three public PTB datasets (Montgomery County–MC, Shenzen–CH, and Indian–IN), totalizing 955 images, we developed a second model mapping radiological finding probabilities to PTB diagnosis (binary labels). We evaluated this approach for its generalization capabilities against direct models, learnt directly from PTB training data or by transfer learning via crossfolding and cross-database experiments. The performance of each approach was evaluated by the area under the specificity vs. sensitivity curve (AUC).

Resultados e Conclusão

The AUC for intra-dataset tests baseline direct detection deep models achieved 0.95 (MC), 0.95 (CH) and 0.91 (IN), with up to 35% performance drop on a cross-dataset evaluation scenario. Our proposed approach achieved AUC of 0.97 (MC), 0.90 (CH), and 0.93 (IN), with at most 11% performance drop on a cross-dataset evaluation. In most tests, the difference was less than 5%.
Conclusions:
A two-step CAD model based on radiological signs offers an adequate base for the development of PTB screening systems and is more generalizable than a direct model. Unlike commercially available CADs, our model is completely reproducible and available open-source at https://pypi.org/project/bob.med.tb/.

Palavras-chave

computer-aided detection
screening
tuberculosis