The first version of the system is trained on a small number of images. Only frontal projection of patient images was taken into account.
The accuracy of the system will be GREATLY improved with the use of close and detailed datasets (pictures in two projections) and patient medical test data. The system can be trained to identify signs of other diseases. This will provide a doctor with the widest possible range of pathologies and help take effective measures to confirm and treat diseases.
Further development involves the following promising areas:
- Recognition of MRI and CT scans
- Recognition of ultrasound pictures
- Medical history analysis
- DNA analysis