使用睡眠数据预测130多种疾病, 表明帕金森氏病、痴呆症、心脏病、癌症、心理健康等问题的准确性很强。 A new AI model, SleepFM, predicts over 130 diseases using sleep data, showing strong accuracy for Parkinson’s, dementia, heart attacks, cancers, and mental health issues.

新的人工智能模型“睡眠FM”精确地预测了130多种疾病的风险,该模型使用来自多功能摄影的睡眠数据,显示了帕金森氏病、痴呆症、心脏病、某些癌症和心理健康状况的强效表现。 A new AI model, SleepFM, accurately predicts the risk of over 130 diseases using sleep data from polysomnography, showing strong performance for Parkinson’s, dementia, heart attacks, certain cancers, and mental health conditions. 该模型对1 000多名患者的数据进行了培训,并进行了长达25年的健康后续跟踪,该模型利用来自大脑、心脏和呼吸活动的同步信号来识别早期疾病标记。 Trained on data from over 1,000 patients with up to 25 years of health follow-up, the model uses synchronized signals from brain, heart, and respiratory activity to identify early disease markers. 虽然目前它缺乏简单的语言解释,但研究人员正在开发工具来解释其调查结果。 While it currently lacks plain-language explanations, researchers are developing tools to interpret its findings. 该研究由斯坦福牵头,并由国家卫生研究所资助,表明睡眠研究可以成为早期疾病检测的有力工具,未来可以通过可磨损的设备数据加以改进。 The study, led by Stanford and funded by the NIH, suggests sleep studies could become a powerful tool for early disease detection, with future enhancements possible through wearable device data.