文献速递心脏核医学

减轻深度学习从心肌灌注SPECT图像诊断冠状动脉疾病中的偏倚

Mitigating bias in deep learning for diagnosis of coronary artery disease from myocardial perfusion SPECT images.
PubMed 文献 · Eur J Nucl Med Mol Imaging · 2022-10-04

研究用3个中心828例患者的MPI与6个月内侵入性血管造影数据训练深度学习模型,比较无数据增强、对无阻塞性CAD患者增强采样、对无CAD且TPD<2%患者增强采样三种策略,并在外部人群(造影者332例、低可能性CAD者179例)测试。模型3校准最佳(Brier评分0.104 vs 0.121,p<0.01),女性中改善尤为明显(0.084 vs 0.124,p<0.01);外部测试AUC模型3为0.930,高于U-TPD的0.897和S-TPD的0.900(均p<0.01)。

要点

  • 对低风险患者增强采样训练可改善AI模型校准,更准确估计CAD概率。
  • 模型3校准最佳,Brier评分0.104 vs 0.121(p<0.01),女性中0.084 vs 0.124(p<0.01)。
  • 外部测试AUC模型3为0.930,高于U-TPD的0.897和S-TPD的0.900(均p<0.01)。

核心结论

  • 使用低风险患者增强采样训练AI模型可改善识别CAD的模型校准,使疾病概率赋值更准确。
  • 这在较低风险人群和女性中尤为重要,因为高估疾病概率可能显著影响下游患者管理。

研究设计与样本

  • 深度学习训练使用来自3个中心的828例患者,均有MPI及6个月内侵入性血管造影。
  • 灌注评估采用直立位总灌注缺损(U-TPD)和仰卧位总灌注缺损(S-TPD)。
  • 模型1为无数据增强训练,模型2为对无阻塞性CAD患者增强采样训练,模型3为对无CAD且TPD<2%患者增强采样训练。
  • 所有模型在外部人群测试:6个月内侵入性血管造影者332例或CAD低可能性者179例,合计511例。

主要发现

  • 模型3校准最佳,Brier评分0.104 vs 0.121(p<0.01)。
  • 校准改善在女性中尤为明显,Brier评分0.084 vs 0.124(p<0.01)。
  • 外部测试中模型3的AUC为0.930,高于U-TPD的0.897和S-TPD的0.900(均p<0.01)。

关键数值与推荐

  • 模型3 Brier评分0.104 vs 0.121(p<0.01)。
  • 女性中模型3 Brier评分0.084 vs 0.124(p<0.01)。
  • 外部测试AUC:模型3为0.930,U-TPD为0.897,S-TPD为0.900(均p<0.01)。

临床怎么用

  • 原文未据此给出推荐。

局限性

  • 原文未报告局限性。

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)。

原文

PURPOSE: Artificial intelligence (AI) has high diagnostic accuracy for coronary artery disease (CAD) from myocardial perfusion imaging (MPI). However, when trained using high-risk populations (such as patients with correlating invasive testing), the disease probability can be overestimated due to selection bias. We evaluated different strategies for training AI models to improve the calibration (accurate estimate of disease probability), using external testing. METHODS: Deep learning was trained using 828 patients from 3 sites, with MPI and invasive angiography within 6 months. Perfusion was assessed using upright (U-TPD) and supine total perfusion deficit (S-TPD). AI training without data augmentation (model 1) was compared to training with augmentation (increased sampling) of patients without obstructive CAD (model 2), and patients without CAD and TPD < 2% (model 3). All models were tested in an external population of patients with invasive angiography within 6 months (n = 332) or low likelihood of CAD (n = 179). RESULTS: Model 3 achieved the best calibration (Brier score 0.104 vs 0.121, p < 0.01). Improvement in calibration was particularly evident in women (Brier score 0.084 vs 0.124, p < 0.01). In external testing (n = 511), the area under the receiver operating characteristic curve (AUC) was higher for model 3 (0.930), compared to U-TPD (AUC 0.897) and S-TPD (AUC 0.900, p < 0.01 for both). CONCLUSION: Training AI models with augmentation of low-risk patients can improve calibration of AI models developed to identify patients with CAD, allowing more accurate assignment of disease probability. This is particularly important in lower-risk populations and in women, where overestimation of disease probability could significantly influence down-stream patient management.

来源:Eur J Nucl Med Mol Imaging,查看原文。版权归原作者所有。

引用

  1. Miller RJH, Singh A, Otaki Y, et al. Mitigating bias in deep learning for diagnosis of coronary artery disease from myocardial perfusion SPECT images. Eur J Nucl Med Mol Imaging. 2022. PMID: 36194270. doi:10.1007/s00259-022-05972-w

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。