Multimedia Journal of Metaverse in MEDICINE

The patient underwent thin-slice contrast-enhanced CT scanning. The imaging data were recorded onto a CD, and the DICOM-format images were imported into the reconstruction software. Manual annotation and identification were performed to complete the reconstruction. Initially, the starting points of each segmental pulmonary artery are manually identified. Subsequently, the computer automatically segments the lungs based on the pulmonary artery distribution and further divides them into subsegments according to the subsegmental pulmonary artery origins. The process primarily relies on the seed line method, with manual extraction serving as a supplementary approach. The lung parenchyma is maximally extracted and integrated with the segmented divisions to generate corresponding modules and assign distinct colors. Next, the CT images are adjusted to the pulmonary artery window, and the entire arterial tree is segmented using the seed point method. The same technique is applied to delineate the venous vascular tree. The bronchial reconstruction is then completed by adjusting the threshold to the bronchial range and using a combination of threshold segmentation and the seed point method. Finally, pulmonary nodules are segmented and reconstructed. Larger nodules are extracted using the threshold and seed point methods, while smaller nodules are manually delineated. By combining the 3D images of the pulmonary nodules with the 3D lung images and adjusting the transparency of the lung images, the precise location of the nodules and their corresponding pulmonary segment or subsegment can be clearly visualized.