为第二阶段抑郁症治疗优化精准医学:一种机器学习方法.
Joshua Curtiss1, Jordan W Smoller2, Paola Pedrelli1
1Depression Clinical and Research Program, Department of Psychiatry, Massachusetts General Hospital, Boston, MA, USA.
Psychological medicine
|March 27, 2024
概括
机器学习可以预测抑郁症治疗的成功. 整体模型显示认知治疗的准确性高于药物治疗,为个性化抑郁症护理提供了希望.
科学领域:
- 精神病学和计算科学 精神病学和计算科学
- 医疗保健中的机器学习
背景情况:
- 标准抗抑郁药单一治疗在超过三分之二的抑郁症患者中未能达到缓解.
- 目前选择的第二阶段抑郁症治疗方法严重依赖于临床直觉,导致延迟和患者负担.
- 建议采用整体机器学习方法来提高第二阶段治疗缓解的预测准确性.
研究的目的:
- 开发和评估一个整体机器学习模型,用于预测接受第二阶段护理的抑郁症患者的治疗缓解.
- 评估机器学习在预测对各种第二阶段治疗策略的反应中的准确性.
主要方法:
- 利用了STAR*D第二级数据集中的1439名患者的数据,随机分配到七种第二步治疗.
- 采用集体机器学习模型,结合多个算法,通过嵌套交叉验证进行评估.
- 包括155个预测变量,包括临床和人口统计指标.
主要成果:
- 组合模型在不同的第二步治疗中显示出不同的预测性能,AUC值从0.51到0.82.8不等.
- 预测缓解是认知疗法最准确的 (AUC = 0.82).
- 其他药物和联合治疗选择的预测准确性较低 (AUC = 0.51-0.66).
结论:
- 整体机器学习显示出预测第二阶段抑郁症治疗有效性的前景.
- 预测准确性因治疗类型而异,与药物治疗相比,行为干预的预测准确性更高.
- 未来的研究应该探索额外的预测模式,以改善对治疗反应的预测.
相关概念视频
Antidepressant Drugs: MAOIs and Other Agents
228
Atypical antidepressants, including bupropion (Wellbutrin), mirtazapine (Remeron), nefazodone (Serzone), trazodone (Desyrel), and vilazodone (Viibryd), offer unique mechanisms of action. Bupropion weakly inhibits dopamine and norepinephrine reuptake, aiding depression treatment and smoking cessation, with a low risk of sexual dysfunction. Mirtazapine enhances serotonin and norepinephrine neurotransmission, leading to sedation, increased appetite, and weight gain. As a result, it helps treat...
228
Human Genetics
566
Human genetics provides a profound framework for understanding the interplay between genetic predispositions and human psychology. At the heart of this discipline lies the study of how genes influence physical traits, behaviors, and susceptibility to diseases. Each person carries a unique genetic code that subtly or significantly shapes their psychological and behavioral landscape.
The complex relationship between genetics and psychology is observable through common biological components such...
The complex relationship between genetics and psychology is observable through common biological components such...
566


