CAD4TB from Delft Imaging
AI that evaluates chest X-rays for signs of tuberculosis, even in areas with few available radiologists.
About
CAD4TB provides a risk score for a digital chest X-ray and highlights suspicious areas. This supports large screening programs and mobile clinics without continuous access to a radiologist. In an independent comparison of twelve systems, CAD4TB version 7 performed about as well as an experienced tuberculosis reader on the dataset studied. A laboratory test and clinician are still required for diagnosis.
AI factor
The model can assess many chest X-rays in succession and prioritize suspicious images. A healthcare professional decides who receives additional testing and makes the diagnosis based on more than just the X-ray.
Sources
Inspiration sourced from AI For Good. Not an Arnaldus customer case.