文献速递骨肌技术与安全DR

三种商业AI算法在急诊真实场景中检测急性外周骨折的性能

Artificial Intelligence for Detecting Acute Fractures in Patients Admitted to an Emergency Department: Real-Life Performance of Three Commercial Algorithms.
PubMed 文献 · Acad Radiol · 2023-07-18

本研究纳入急诊连续1210例骨骼创伤患者共1500套X线片,以四位肌骨放射科医生判定为参考标准,评估SmartUrgence、Rayvolve和BoneView三种AI算法的诊断性能。结果显示SmartUrgence和BoneView的准确度和特异度显著高于Rayvolve,但三者敏感度无显著差异;AI性能与算法、患者年龄和检查部位相关。

要点

  • 三种AI算法敏感度均较高(90.2%~92.6%),但Rayvolve特异度仅70.4%,假阳性较多。
  • SmartUrgence和BoneView整体准确度约90%,显著优于Rayvolve(71.0%)。
  • AI性能受检查部位影响,膝部准确度和特异度高于足部。

核心结论

  • 在急诊日常放射实践中,三种商业AI算法检测急性外周骨折的准确度良好至较高,性能与AI算法、患者年龄和检查部位相关。

研究设计与样本

  • 单中心前瞻性收集连续急诊骨骼创伤患者,共1210例(平均年龄41.3±18.5岁,男性742例[61.3%]),对应1500套X线片。
  • 以四位肌骨放射科医生判定的骨折为参考标准,计算三种AI算法在X线套水平的诊断性能。

诊断标准与关键数值

  • 骨折在X线套中的患病率为23.7%(356/1500)。
  • SmartUrgence、Rayvolve、BoneView的准确度分别为90.1%、71.0%、88.8%;敏感度分别为90.2%、92.6%、91.3%;特异度分别为92.5%、70.4%、90.5%。
  • SmartUrgence和BoneView的准确度和特异度在全人群及亚组中均显著高于Rayvolve(P < .0001),三者敏感度无显著差异(P = .27)。
  • Rayvolve在27-36岁组的准确度和特异度显著高于≥53岁组(P = .0029和P = .0019),膝部特异度高于足部(P = .0149)。
  • BoneView膝部准确度显著高于足部(P = .0006)和腕/手部(P = .0228),膝部特异度显著高于足部(P = .0003),踝部特异度显著高于足部(P = .0195)。

临床应用与阅片要点

  • AI可作为急诊骨折筛查的辅助工具,但需注意Rayvolve特异度较低可能增加假阳性复核负担。
  • 阅片时应结合患者年龄和检查部位,对足部等性能相对较低的部位保持警惕。

局限性

  • 单中心研究,样本来自同一急诊科2个月连续患者,外推性有限。

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

RATIONALE AND OBJECTIVES: Interpreting radiographs in emergency settings is stressful and a burden for radiologists. The main objective was to assess the performance of three commercially available artificial intelligence (AI) algorithms for detecting acute peripheral fractures on radiographs in daily emergency practice.

MATERIALS AND METHODS: Radiographs were collected from consecutive patients admitted for skeletal trauma at our emergency department over a period of 2 months. Three AI algorithms-SmartUrgence, Rayvolve, and BoneView-were used to analyze 13 body regions. Four musculoskeletal radiologists determined the ground truth from radiographs. The diagnostic performance of the three AI algorithms was calculated at the level of the radiography set. Accuracies, sensitivities, and specificities for each algorithm and two-by-two comparisons between algorithms were obtained. Analyses were performed for the whole population and for subgroups of interest (sex, age, body region).

RESULTS: A total of 1210 patients were included (mean age 41.3 ± 18.5 years; 742 [61.3%] men), corresponding to 1500 radiography sets. The fracture prevalence among the radiography sets was 23.7% (356/1500). Accuracy was 90.1%, 71.0%, and 88.8% for SmartUrgence, Rayvolve, and BoneView, respectively; sensitivity 90.2%, 92.6%, and 91.3%, with specificity 92.5%, 70.4%, and 90.5%. Accuracy and specificity were significantly higher for SmartUrgence and BoneView than Rayvolve for the whole population (P < .0001) and for subgroups. The three algorithms did not differ in sensitivity (P = .27). For SmartUrgence, subgroups did not significantly differ in accuracy, specificity, or sensitivity. For Rayvolve, accuracy and specificity were significantly higher with age 27-36 than ≥53 years (P = .0029 and P = .0019). Specificity was higher for the subgroup knee than foot (P = .0149). For BoneView, accuracy was significantly higher for the subgroups knee than foot (P = .0006) and knee than wrist/hand (P = .0228). Specificity was significantly higher for the subgroups knee than foot (P = .0003) and ankle than foot (P = .0195).

CONCLUSION: The performance of AI detection of acute peripheral fractures in daily radiological practice in an emergency department was good to high and was related to the AI algorithm, patient age, and body region examined.

来源:Acad Radiol,查看原文。版权归原作者所有。

引用

  1. Bousson V, Attané G, Benoist N, et al. Artificial Intelligence for Detecting Acute Fractures in Patients Admitted to an Emergency Department: Real-Life Performance of Three Commercial Algorithms. Acad Radiol. 2023. PMID: 37468377. doi:10.1016/j.acra.2023.06.016

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。