방사선의학 웹진

바로가기  >

Abstract
This study investigated the segmentation of cerebral arteries using single-exposure computed tomographic angiography (CTA) based on the projection domain. The proposed method employs the DeepLab V3+ model to segment brain vessels directly in the projection. This approach eliminates the risk of double exposure and motion artifacts while preserving clinical information. Additionally, it minimizes beam-hardening artifacts from high-density materials and reduces the operator's workload. The cerebral artery images reconstructed using the proposed method were quantitatively compared to those of the labeled images, and the intersection over union, Dice similarity coefficient, bidirectional Hausdorff distance, bfscore, F1-score, and precision were measured to be approximately 0.89, 0.90, 219.12, 0.90, 0.90, and 0.89, respectively. All parameters were statistically significant (p-value <0.005). In addition, the proposed method confirmed that reconstructing cerebral arteries and metallic implant components can yield meaningful image information for clinical diagnosis with minimal information loss. These results indicate that the proposed method can provide accurate 3D images of cerebral arteries in CTA imaging systems, thereby preventing misdiagnosis and improving image quality.

Affiliation

Kang-Hyeon Seo1,2 Hajin Kim2 Jina Shim3,4 Caterina Battaglia5 Kyuseok Kim6 Youngjin Lee7

1Department of Radiology, Hallym Hospital,
Gyeyang-gu, Incheon, Republic of Korea
2Department of Health Science, General
Graduate School of Gachon University,
Yeonsu-gu, Incheon, Republic of Korea
3Department of Radiology, Severance
Hospital, Seodaemun-gu, Seoul, Republic of
Korea
4Department of Radiological Science,
Wonkwang University, Iksan-Si, Republic of
Korea
5Department of Experimental and Clinical
Medicine, Magna Graecia University,
Catanzaro, Italy
6Institute of Human Convergence Health
Science, Gachon University, Yeonsu-gu,
Incheon, Republic of Korea
7Department of Radiological Science,
Gachon University, Yeonsu-gu, Incheon,
Republic of Korea