Care Mentor AI

Computer vision system for the detection of focal lung lesions based on the results of computed tomography "CT Onco Mentor"

The Care Mentor AI neural network system for diagnosing cancer from computed tomography analyzes depersonalized CT scans of the chest organs to detect signs of lung cancer (solid and subsolid lesions).

How it works?

System for the analysis of CT studies: "Onco Mentor":
• Analyzes images of chest CT slices
• Segments findings of diagnostic value
• Calculates by lung lobes
• Provides the diagnostician with a CT scan analysis protocol, which includes:

I. description of the changes found (left-right and further by shares),
II. visualization of diagnostic findings on sections in the form of masks of various colors,
III. conclusion

The metrics of diagnostic accuracy in relation to the problem of binary classification for the presence / absence of signs of cancer are:

• Sensitivity> 0.86
• Specificity> 0.92
• Accuracy (overall validity)> 0.91
• Area under - curve (AUC)> 0.93

Our partners

Developed with financial support