Abstract
Pulmonary embolism (PE) remains a significant cause of cardiovascular mortality, with untreated cases showing mortality rates of up to 30%. The evolution of computer-assisted detection (CAD) for PE has transformed dramatically over the past decades, progressing from simple pattern recognition to sophisticated deep learning approaches. Early CAD systems demonstrated modest performance, with sensitivity around 75% at 2–4 false positives per scan, whereas modern deep learning architectures achieve sensitivities of up to 92.9% at 0.15 false positives per scan. Significantly, the technological progression has evolved from basic patient-level classification to sophisticated voxel-level analysis. This review provides a comprehensive overview of the evolution of PE CAD systems, their clinical value, and future directions.