Artificial Intelligence for Medical Systems (AIMS) Lab, Oregon Health & Science University, Portland, OR, USA.
Harold Schnitzer Diabetes Health Center, Oregon Health & Science University, Portland, OR, USA.
Nat Metab. 2020 Jul;2(7):612-619. doi: 10.1038/s42255-020-0212-y. Epub 2020 Jun 1.
Type 1 diabetes (T1D) is characterized by pancreatic beta cell dysfunction and insulin depletion. Over 40% of people with T1D manage their glucose through multiple injections of long-acting basal and short-acting bolus insulin, so-called multiple daily injections (MDI). Errors in dosing can lead to life-threatening hypoglycaemia events (<70 mg dl) and hyperglycaemia (>180 mg dl), increasing the risk of retinopathy, neuropathy, and nephropathy. Machine learning (artificial intelligence) approaches are being harnessed to incorporate decision support into many medical specialties. Here, we report an algorithm that provides weekly insulin dosage recommendations to adults with T1D using MDI therapy. We employ a unique virtual platform to generate over 50,000 glucose observations to train a k-nearest neighbours decision support system (KNN-DSS) to identify causes of hyperglycaemia or hypoglycaemia and determine necessary insulin adjustments from a set of 12 potential recommendations. The KNN-DSS algorithm achieves an overall agreement with board-certified endocrinologists of 67.9% when validated on real-world human data, and delivers safe recommendations, per endocrinologist review. A comparison of inter-physician-recommended adjustments to insulin pump therapy indicates full agreement of 41.2% among endocrinologists, which is consistent with previous measures of inter-physician agreement (41-45%). In silico benchmarking using a platform accepted by the United States Food and Drug Administration for evaluation of artificial pancreas technologies indicates substantial improvement in glycaemic outcomes after 12 weeks of KNN-DSS use. Our data indicate that the KNN-DSS allows for early identification of dangerous insulin regimens and may be used to improve glycaemic outcomes and prevent life-threatening complications in people with T1D.
1 型糖尿病（T1D）的特征是胰岛β细胞功能障碍和胰岛素耗竭。超过 40%的 T1D 患者通过多次注射长效基础胰岛素和短效餐时胰岛素来控制血糖，即所谓的多次每日注射（MDI）。剂量错误可能导致危及生命的低血糖事件（<70mg/dl）和高血糖（>180mg/dl），增加视网膜病变、神经病变和肾病的风险。机器学习（人工智能）方法正被用于将决策支持纳入许多医学专业领域。在这里，我们报告了一种算法，该算法使用 MDI 疗法为 T1D 成年患者提供每周胰岛素剂量建议。我们采用独特的虚拟平台生成超过 50000 个血糖观察值，以训练 k-最近邻决策支持系统（KNN-DSS），从而识别高血糖或低血糖的原因，并从 12 种潜在建议中确定必要的胰岛素调整。该 KNN-DSS 算法在对真实世界人类数据进行验证时，与经过董事会认证的内分泌学家的总体一致性为 67.9%，并且根据内分泌学家的审查提供安全的建议。与胰岛素泵治疗的医生间推荐调整的比较表明，内分泌学家之间的完全一致为 41.2%，这与以前的医生间一致性测量结果（41-45%）一致。使用美国食品和药物管理局评估人工胰腺技术的平台进行的模拟基准测试表明，在使用 KNN-DSS 12 周后，血糖控制结果有了实质性的改善。我们的数据表明，KNN-DSS 可以早期识别危险的胰岛素方案，可用于改善血糖控制结果并预防 T1D 患者的危及生命的并发症。
