Automatic localization and deep convolutional generative adversarial network‐based classification of focal liver lesions in computed tomography images: A preliminary study
Pushpanjali Gupta, Yao‐Chun Hsu, Liang Li, Yuan-Chia Chu, Chia‐Sheng Chu, Jaw‐Liang Wu, Jian‐An Chen, Wei‐Hsiu Tseng, Y. Yang, Teng‐Yu Lee, Che‐Lun Hung, Chun‐Ying Wu · Journal of Gastroenterology and Hepatology · 2024
BACKGROUND AND AIM: Computed tomography of the abdomen exhibits subtle and complex features of liver lesions, subjectively interpreted by physicians. We developed a deep learning-based localization and classification (DLLC) system for focal liver lesions (FLLs) in computed tomography imaging that could assist physicians in more robust clinical decision-making. METHODS: We conducted a retrospective study (approval no. EMRP-109-058) on 1589 patients with 17 335 slices with 3195 FLLs using data from January 2004 to December 2020. The training set included 1272 patients (male: 776, mean age 62 ± 10.9), and the test set included 317 patients (male: 228, mean age 57 ± 11.8). The slices were annotated by annotators with different experience levels, and the DLLC system was developed using generative adversarial networks for data augmentation. A comparative analysis was performed for the DLLC system versus physicians using external data. RESULTS: Our DLLC system demonstrated mean average precision at 0.81 for localization. The system's overall accuracy for multiclass classifications was 0.97 (95% confidence interval [CI]: 0.95-0.99). Considering FLLs ≤ 3 cm, the system achieved an accuracy of 0.83 (95% CI: 0.68-0.98), and for size > 3 cm, the accuracy was 0.87 (95% CI: 0.77-0.97) for localization. Furthermore, during classification, the accuracy was 0.95 (95% CI: 0.92-0.98) for FLLs ≤ 3 cm and 0.97 (95% CI: 0.94-1.00) for FLLs > 3 cm. CONCLUSION: This system can provide an accurate and non-invasive method for diagnosing liver conditions, making it a valuable tool for hepatologists and radiologists.