AI in Tuberculosis Detection

iDoc-X is a detection solution for TB. The software scans a Chest X-ray image and outputs the probability of TB. Also, localising (annotating) the areas of the image most indicative of abnormality. The solution is based on a convolutional neural network with more than 121 layers and advanced count of CNN layers.

 

Tuberculosis (TB) is one of the leading causes of death, having a worldwide view. An estimation depicts that around one-quarter of the world’s population is infected with TB. In 2017, more than 10 million cases of active TB had been reported which resulted in 1.6 million deaths, mostly in developing countries (around 95 %) such as in India (around 50%), Pakistan, Indonesia.

iDoc-X along with detection also comprises of Triage Service which allows iDoc-X to prioritise the CXR based on the presence of abnormality. The AI engine post-classification prioritise the detection on radiologist worklist. The solution delivers clinical augmentation. Also, non-clinical augmentation such as session time and cost improvement. Triage service is focuseded on prioritisation with focus on Time & Cost in the care delivery system.