Pasa, F., Golkov, V., Pfeiffer, F., Cremers, D., & Pfeiffer, D. Efficient deep network architectures for fast chest X-ray tuberculosis screening and visualization. Scientific reports, 9(1), 1-9 (2019) https://doi.org/10.1038/s41598-019-42557-4
Purpose: In this research project, a deep learning architecture tailored to tuberculosis diagnosis is presented. With this approach, the computational and memory requirement reduces significantly, without sacrificing the classification performance. Results of the training through the use of saliency maps and grad-CAMs are further discuss . These techniques, which, to the best our knowledge, were never applied to this problem, provide an approximate visual diagnosis that might be a useful additional tool for clinicians.
Figure: Saliency map with overlay for two correctly classified cases.