Tseng Wen-Chang, Wang Yung-Cheng, Chen Wei-Chi, Lin Kang-Ping
Department of Radiology, Cathay General Hospital, Taipei 106, Taiwan.
Department of Electrical Engineering, Chung Yuan Christian University, Taoyuan 320, Taiwan.
Eur J Radiol Open. 2025 Jun 10;14:100664. doi: 10.1016/j.ejro.2025.100664. eCollection 2025 Jun.
This study develops an AI-assisted pneumothorax diagnosis system using deep learning and chest X-ray images to enhance diagnostic efficiency and accuracy, reduce radiologists' workload, and provide timely treatment. The system addresses limitations of traditional methods, which rely on subjective interpretation and are vulnerable to fatigue or inexperience.
The DenseNet121 model was employed using a chest X-ray dataset from a medical center in northern Taiwan, with a total of 6888 images' divided into training (64 %), validation (16 %), and testing (20 %) sets. Image preprocessing involved normalization, data augmentation (rotation, translation, scaling, brightness adjustment), and standardization. The model was trained using stochastic gradient descent with an initial learning rate of 0.0016 for 150 epochs. Performance evaluation included accuracy, sensitivity, specificity, and AUROC, integrating with the hospital's PACS for real-time analysis.
Initial testing yielded AUROC values of 94.52 % and 97.21 % for pneumothorax and mild pneumothorax groups. However, when applied to 6888 clinical images, the AUROC dropped to 62.55 %, resulting in 4294 false positives. Adjusting the dataset split and retraining with 1000 false positive images improved the AUROC from 62.55 % to 85.53 %.
The AI model shows potential in pneumothorax detection, but performance is influenced by data diversity, image quality, and clinical complexity. The model struggles to identify key areas in complex cases, indicating a need for attention mechanisms or region proposal networks (RPN). Expanding the dataset, optimizing preprocessing, and training separate models for different image locations could enhance performance further.
本研究开发一种利用深度学习和胸部X光图像的人工智能辅助气胸诊断系统，以提高诊断效率和准确性，减轻放射科医生的工作量，并提供及时治疗。该系统解决了传统方法的局限性，传统方法依赖主观解读，容易受到疲劳或经验不足的影响。
采用DenseNet121模型，使用来自台湾北部一家医疗中心的胸部X光数据集，共有6888张图像，分为训练集（64%）、验证集（16%）和测试集（20%）。图像预处理包括归一化、数据增强（旋转、平移、缩放、亮度调整）和标准化。模型使用随机梯度下降进行训练，初始学习率为0.0016，训练150个轮次。性能评估包括准确率、灵敏度、特异性和曲线下面积（AUROC），并与医院的图像存档与通信系统（PACS）集成以进行实时分析。
气胸组和轻度气胸组的初始测试中，曲线下面积（AUROC）值分别为94.52%和97.21%。然而，当应用于6888张临床图像时，曲线下面积（AUROC）降至62.55%，导致4294例假阳性。调整数据集划分并使用1000例假阳性图像重新训练后，曲线下面积（AUROC）从62.55%提高到85.53%。
人工智能模型在气胸检测中显示出潜力，但性能受数据多样性、图像质量和临床复杂性影响。该模型在复杂病例中难以识别关键区域，表明需要注意力机制或区域建议网络（RPN）。扩大数据集、优化预处理以及针对不同图像位置训练单独的模型可能会进一步提高性能。
