CLChun-Chung Lui

Chun-Chung Lui

*This is an algorithmically generated author record based on publicly available information.

Chun-Chung Lui has a total of 20 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Malignant Solitary Fibrous Tumor in Lacrimal Sac

    by Tse-Jen Huang, Chun-Chung Lui, Kai-Jen Lin, Chao-Tien Hsu

  • Solitary Fibrous Tumor in the Central Nervous System

    by Kuo-Wei Wang, Chun-Chung Lui, Tsung-Heng Tsai, Chao-Tien Hsu

  • Open Access

    Generative Adversarial Network (GAN) for Automatic Reconstruction of the 3D Spine Structure by Using Simulated Bi-Planar X-ray Images

    by Ching-Juei Yang, Cheng-Li Lin, Chien-Kuo Wang, Jing-Yao Wang, Chih-Chia Chen, Fong-Chin Su, Yin-Ju Lee, Chun-Chung Lui, Lee-Ren Yeh, Yu-Hua Dean Fang

    In this study, we modified the previously proposed X2CT-GAN to build a 2Dto3D-GAN of the spine. This study also incorporated the radiologist’s perspective in the adjustment of input signals to prove the feasibility of the automatic production of three-dimensional (3D) structures of the spine from simulated bi-planar two-dimensional (2D) X-ray images. Data from 1012 computed tomography (CT) studies of 984 patients were retrospectively collected. We tested this model under different dataset sizes (333, 666, and 1012) with different bone signal conditions to observe the training performance. A 10-fold cross-validation and five metrics—Dice similarity coefficient (DSC) value, Jaccard similarity coefficient (JSC), overlap volume (OV), and structural similarity index (SSIM)—were applied for model evaluation. The optimal mean values for DSC, JSC, OV, SSIM_anteroposterior (AP), and SSIM_Lateral (Lat) were 0.8192, 0.6984, 0.8624, 0.9261, and 0.9242, respectively. There was a significant improvement in the training performance under empirically enhanced bone signal conditions and with increasing training dataset sizes. These results demonstrate the potential of the clinical implantation of GAN for automatic production of 3D spine images from 2D images. This prototype model can serve as a foundation in future studies applying transfer learning for the development of advanced medical diagnostic techniques.


  • Open Access

    Treatment, Outcome, and Relapse of Spontaneous and Nonspontaneous Cerebrospinal Fluid Leak

    by Yi-Cheng Tai, Yi-Sheng Tai, Chang-Hsien Ou, Chun-Chung Lui, Hao-Kuang Wang, Hung-Chang Kuo, Shih-Pin Hsu

    Cerebrospinal fluid (CSF) leak can be spontaneous or nonspontaneous. The management options include conservative treatments, blood patch, and surgical repairs. We compared clinical symptoms, image findings, management options, hospitalization, and relapse rates among different causes of CSF leaks. Eighty-one patients were recruited: 20 with spontaneous and 61 with nonspontaneous CSF leaks. Nonspontaneous causes included lumbar puncture, surgery, and trauma. Surgery sites comprised sphenoid, spine, skull base, and calvaria. Spontaneous CSF leak came from the sphenoid or spine. Age, gender, body mass index, initial symptoms, hospitalization, treatment courses, and recurrence rates showed no difference between the groups. The spontaneous group had higher CSF accumulations on their MRIs. MRI pachymeninge enhancement showed the highest sensitivity (78.6%) for intracranial hypotension. Meningitis occurred in 1/3 of sphenoid, skull base, and calvarian surgeries. Earlier reoperation was correlated with shorter hospitalization (r = 0.651), but the recurrence rates were similar. Longer intervals between surgery and CSF leak encouraged reoperation. Among the spontaneous spine and lumbar puncture-related CSF leaks, 57.1% of them responded to 4 days of conservative treatment. Among the trauma-related CSF leaks, 90.9% of them required surgical repair. The demographic data and symptoms were similar in various groups of CSF leak. The symptom onset durations and treatment strategies were different. However, the recurrence rates were similar.