Clinical Evaluation of AI100 with Shonit™ versus Manual Microscopy

Area of Science:

• Hematology
• Medical Diagnostics
• Artificial Intelligence in Medicine

Background:

• Peripheral blood smear (PBS) analysis is crucial for medical diagnosis.
• Current methods rely on manual microscopy, which can be time-consuming.
• Automated systems offer potential for improved efficiency and accuracy.

Purpose of the Study:

• To evaluate the clinical performance of the AI100 with Shonit™ device.
• To establish the equivalence of AI100 with Shonit™ to manual microscopy for PBS analysis.
• To assess the device's utility for white blood cell (WBC) differential count, red blood cell (RBC) and platelet morphology precharacterization.

Main Methods:

• A retrospective, multicenter study involving 882 samples across four sites (India and USA).
• Independent analysis of samples using manual microscopy and the AI100 with Shonit™ by qualified reviewers.
• Comparison of post-review results between the automated device and manual microscopy.

Main Results:

• High overall agreement for WBC abnormality (95.03%), RBC anisocytosis (94.7%), and RBC poikilocytosis (92.1%).
• Excellent sensitivity and specificity for platelet morphology (97.9% and 94.6%, respectively).
• Demonstrated repeatability and reproducibility of the AI100 with Shonit™ system.

Conclusions:

• The AI100 with Shonit™ system demonstrates substantial equivalence to manual microscopy.
• The device is suitable for diagnostic purposes in peripheral blood smear analysis.
• AI-powered automation shows promise for efficient and accurate hematological diagnostics.