通过受训技师构建用于人工智能应用的三维胰腺分割CT数据集:COVID-19防控阶段的一项研究
本研究评估受训技师与放射科医生进行三维胰腺分割的表现,并观察补充培训对技师表现的影响。22名技师在放射科医生主导的互动视频会议培训后分割188例门静脉期CT,补充培训(针对常见错误的多媒体视频)后再分割159例;两批分割的Dice-Sorensen系数(DSC)均为0.63,差异无统计学意义(p > 0.05),但补充培训后胰腺体积差范围缩小。结论认为受训技师能以合理精度完成三维胰腺分割,补充培训进一步缩小了体积差范围,投资于技师培训可扩充和加速用于人工智能的体部影像数据集开发。
要点
- 受训技师在门静脉期CT上可完成三维胰腺分割,两批DSC均为0.63。
- 首批38%病例需放射科医生修正,补充培训后第二批升至48%。
- 补充培训后胰腺体积差范围缩小,但DSC等指标无显著差异(p > 0.05)。
诊断标准与关键数值
- 首批分割需修正71例(38%),其中过度分割26例(37%)、分割不足45例(63%)。
- 第二批分割需修正77例(48%),其中过度分割12例(16%)、分割不足65例(84%)。
- 首批与第二批的DSC分别为0.63(0.15)与0.63(0.16),JC分别为0.48(0.15)与0.48(0.15),差异均无统计学意义(p > 0.05)。
- 首批与第二批的假阳性分别为0.29(0.21)与0.21(0.10),假阴性分别为0.36(0.20)与0.43(0.19)。
- 首批与第二批的平均胰腺体积差分别为-2.74 cc(最小-92.96 cc,最大87.47 cc)与-23.57 cc(最小-77.32 cc,最大30.19 cc),第二批范围缩小。
临床应用与阅片要点
- 放射科医生对技师三维胰腺分割的修正率为38%至48%,阅片时可将技师分割作为初始轮廓再行修正。
- 补充培训后分割不足占比由63%升至84%,过度分割占比由37%降至16%。
局限性
- 补充培训后DSC、JC、假阳性、假阴性差异均无统计学意义(p > 0.05)。
- 补充培训后修正病例比例由38%升至48%。
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
PURPOSE: To evaluate the performance of trained technologists vis-à-vis radiologists for volumetric pancreas segmentation and to assess the impact of supplementary training on their performance.
METHODS: In this IRB-approved study, 22 technologists were trained in pancreas segmentation on portal venous phase CT through radiologist-led interactive videoconferencing sessions based on an image-rich curriculum. Technologists segmented pancreas in 188 CTs using freehand tools on custom image-viewing software. Subsequent supplementary training included multimedia videos focused on common errors, which were followed by second batch of 159 segmentations. Two radiologists reviewed all cases and corrected inaccurate segmentations. Technologists' segmentations were compared against radiologists' segmentations using Dice-Sorenson coefficient (DSC), Jaccard coefficient (JC), and Bland-Altman analysis.
RESULTS: Corrections were made in 71 (38%) cases from first batch [26 (37%) oversegmentations and 45 (63%) undersegmentations] and in 77 (48%) cases from second batch [12 (16%) oversegmentations and 65 (84%) undersegmentations]. DSC, JC, false positive (FP), and false negative (FN) [mean (SD)] in first versus second batches were 0.63 (0.15) versus 0.63 (0.16), 0.48 (0.15) versus 0.48 (0.15), 0.29 (0.21) versus 0.21 (0.10), and 0.36 (0.20) versus 0.43 (0.19), respectively. Differences were not significant (p > 0.05). However, range of mean pancreatic volume difference reduced in the second batch [- 2.74 cc (min - 92.96 cc, max 87.47 cc) versus - 23.57 cc (min - 77.32, max 30.19)].
CONCLUSION: Trained technologists could perform volumetric pancreas segmentation with reasonable accuracy despite its complexity. Supplementary training further reduced range of volume difference in segmentations. Investment into training technologists could augment and accelerate development of body imaging datasets for AI applications.
引用
- Suman G, Panda A, Korfiatis P, et al. Development of a volumetric pancreas segmentation CT dataset for AI applications through trained technologists: a study during the COVID 19 containment phase. Abdom Radiol (NY). 2020. PMID: 32939632. doi:10.1007/s00261-020-02741-x
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。