深度学习目标检测用于儿童腕部骨折识别
本研究利用395张儿童腕关节正位X线片训练Faster R-CNN目标检测模型,并将其用于辅助住院医师判读。模型识别任何骨折的AUC为0.92,敏感度88%、特异度89%,对桡骨远端扣带骨折的检出率为90%;住院医师在参考模型预测后,诊断准确率显著提高(任何骨折从80%升至93%,扣带骨折从69%升至92%)。
要点
- Faster R-CNN模型在儿童腕部骨折检测中AUC达0.92,可高敏感度识别扣带骨折。
- AI辅助显著提升住院医师诊断准确率,尤其对扣带骨折。
- 模型与医师意见不一致时,医师正确率显著高于AI,提示AI应作为辅助工具。
核心结论
- 基于Faster R-CNN的深度学习目标检测模型仅用数百例训练即可高准确度识别儿童腕部骨折,并提供预测框辅助住院医师提高诊断准确率。
研究设计与样本
- 回顾性研究,纳入395张儿童腕关节正位X线片(65%骨折阳性,30%为桡骨远端扣带骨折),分为训练集229张、调优集41张和测试集125张;训练Faster R-CNN目标检测模型,并由2名儿科和2名放射科住院医师在无AI辅助和有AI预测框辅助下分别判读。
诊断标准与关键数值
- Faster R-CNN模型识别任何骨折的AUC为0.92(95%CI 0.87-0.97),准确率88%(110/125,95%CI 81-93%),敏感度88%(70/80,95%CI 78-94%),特异度89%(40/45,95%CI 76-96%)。
- 模型对扣带骨折的检出率为90%(35/39,95%CI 76-97%)。
- 住院医师在参考AI预测后,诊断任何骨折的平均准确率从80%提高到93%(P<0.001),诊断扣带骨折从69%提高到92%(P<0.001)。
- 在医师与AI意见不一致的病例中,医师正确率为73%,显著高于AI的27%(P=0.002)。
临床应用与阅片要点
- AI模型可作为儿童腕部X线判读的辅助工具,尤其有助于低年资医师检出扣带骨折等细微骨折。
- 当AI与医师判断不一致时,应警惕AI假阳性或假阴性,医师需结合临床和影像特征综合判断。
局限性
- 样本量较小(共395张影像),来源单一,仅使用正位腕关节X线片。
- 缺乏外部验证,模型在其他机构或不同设备上的泛化能力未知。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
BACKGROUND: Missed fractures are the leading cause of diagnostic error in the emergency department, and fractures of pediatric bones, particularly subtle wrist fractures, can be misidentified because of their varying characteristics and responses to injury.
OBJECTIVE: This study evaluated the utility of an object detection deep learning framework for classifying pediatric wrist fractures as positive or negative for fracture, including subtle buckle fractures of the distal radius, and evaluated the performance of this algorithm as augmentation to trainee radiograph interpretation.
MATERIALS AND METHODS: We obtained 395 posteroanterior wrist radiographs from unique pediatric patients (65% positive for fracture, 30% positive for distal radial buckle fracture) and divided them into train (n = 229), tune (n = 41) and test (n = 125) sets. We trained a Faster R-CNN (region-based convolutional neural network) deep learning object-detection model. Two pediatric and two radiology residents evaluated radiographs initially without the artificial intelligence (AI) assistance, and then subsequently with access to the bounding box generated by the Faster R-CNN model.
RESULTS: The Faster R-CNN model demonstrated an area under the curve (AUC) of 0.92 (95% confidence interval [CI] 0.87-0.97), accuracy of 88% (n = 110/125; 95% CI 81-93%), sensitivity of 88% (n = 70/80; 95% CI 78-94%) and specificity of 89% (n = 40/45, 95% CI 76-96%) in identifying any fracture and identified 90% of buckle fractures (n = 35/39, 95% CI 76-97%). Access to Faster R-CNN model predictions significantly improved average resident accuracy from 80 to 93% in detecting any fracture (P < 0.001) and from 69 to 92% in detecting buckle fracture (P < 0.001). After accessing AI predictions, residents significantly outperformed AI in cases of disagreement (73% resident correct vs. 27% AI, P = 0.002).
CONCLUSION: An object-detection-based deep learning approach trained with only a few hundred examples identified radiographs containing pediatric wrist fractures with high accuracy. Access to model predictions significantly improved resident accuracy in diagnosing these fractures.
引用
- Zech, Carotenuto, Igbinoba, et al. Detecting pediatric wrist fractures using deep-learning-based object detection. Pediatr Radiol. 2023. PMID: 36650360. doi:10.1007/s00247-023-05588-8
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。