Convolutional Neural Network for Early Pulmonary Embolism Detection via Computed Tomography Pulmonary Angiography
This work addresses a critical medical issue for patients at risk of PE by providing an incremental improvement in triage and lesion labeling to expedite diagnosis.
The study tackled the problem of reducing death rates during the waiting period for pulmonary embolism (PE) diagnosis by developing a computer-aided detection system using computed tomography pulmonary angiography, achieving an accuracy of 0.85 for classification and a mean intersection over union of 0.689 for segmentation.
This study was conducted to develop a computer-aided detection (CAD) system for triaging patients with pulmonary embolism (PE). The purpose of the system was to reduce the death rate during the waiting period. Computed tomography pulmonary angiography (CTPA) is used for PE diagnosis. Because CTPA reports require a radiologist to review the case and suggest further management, this creates a waiting period during which patients may die. Our proposed CAD method was thus designed to triage patients with PE from those without PE. In contrast to related studies involving CAD systems that identify key PE lesion images to expedite PE diagnosis, our system comprises a novel classification-model ensemble for PE detection and a segmentation model for PE lesion labeling. The models were trained using data from National Cheng Kung University Hospital and open resources. The classification model yielded 0.73 for receiver operating characteristic curve (accuracy = 0.85), while the mean intersection over union was 0.689 for the segmentation model. The proposed CAD system can distinguish between patients with and without PE and automatically label PE lesions to expedite PE diagnosis