Reliable Tuberculosis Detection Using Chest X-Ray With Deep Learning, Segmentation and Visualization
importance 10/10
ai assisted
COVID-19 diagnosis using AI TB diagnostics stage: research
Problem
Detecting tuberculosis accurately and early from chest X-ray images, since late diagnosis can be life-threatening.
Why it matters
TB is one of the top 10 leading causes of death, so early reliable detection saves lives.
What the AI did
Nine pre-trained deep CNNs were fine-tuned via transfer learning to classify TB vs. non-TB X-rays, plus two U-net models for lung segmentation.
Results
not stated in source
Impact
The study demonstrates deep-learning TB detection on a 7000-image dataset, but the abstract gives no final metrics.
Record
| Venue | IEEE Access |
|---|---|
| Published | 2020 |
| Work type | article |
| Authors | Tawsifur Rahman, Amith Abdullah Khandakar, Muhammad Abdul Kadir, Khandaker Reajul Islam, Khandakar Reajul Islam, Rashid Mazhar, Tahir Hamid, Mohammad Tariqul Islam et al. |
| Citations | 651 (OpenAlex, at import time) |
| References | 85 |
| Model or method | ResNet18/50/101, CheXNet, InceptionV3, VGG19, DenseNet201, SqueezeNet, MobileNet, U-Net |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1109/access.2020.3031384 |
| Open access | gold |
| Keywords | tuberculosis, chest X-ray, deep learning, CNN, segmentation, transfer learning |
Importance rating
10
existential / global scale
Abstract states TB is one of the top 10 leading causes of death; early reliable detection from X-rays saves lives.