Preliminary study on artificial intelligence diagnosis of pulmonary embolism based on computer in-depth study

Area of Science:

• Radiology
• Artificial Intelligence
• Medical Imaging

Background:

• Pulmonary embolism (PE) diagnosis can be challenging.
• Accurate and efficient diagnostic tools are needed.

Purpose of the Study:

• To assess the feasibility of an AI computer-aided diagnosis (CAD) system for pulmonary embolism (PE).
• To localize and quantitatively diagnose PE using CT angiography (CTA).

Main Methods:

• Retrospective analysis of CTA data from 85 patients with PE.
• Development and application of a deep learning-based CAD system.
• Random allocation into a deep learning group (n=43) and an experimental group (n=42).

Main Results:

• The AI CAD system achieved 90.9% sensitivity for PE detection.
• An average of 2.0 false positives were recorded.
• Detection rate positively correlated with arterial grade.

Conclusions:

• AI CAD system effectively improves PE detection rates, particularly for high-grade embolisms (grade 3 and above).
• Further development with larger datasets is required to reduce misdetection rates for lower-grade embolisms.