背景血脂异常包括通过两种分类(Freickson-Levy [FL]或Sniderman)分类的各种脂肪蛋白疾病。然而,这两种分类都因依靠脂蛋白代谢的不完整知识而受到批评,尤其是在新型治疗方案的明确性和个体治疗反应中的变化中。聚集是一种无监督的机器学习(ML)算法,可以处理广泛的变量,有可能揭示具有独特的分子特征和独特的治疗靶标的患者群体,可以为心脏疾病(CVD)提供更有效的预防策略。We aimed to use unsupervised ML algo- rithms to discover intrinsic dyslipidaemia categories from lipoprotein measurements, to recognise the necessary compo- nents of lipid panels for classification, and to analyse the similarities between the newly formed clusters, FL and Sni- derman classifications.
主要关键词