TY - JOUR
T1 - Artificial intelligence-enabled detection and assessment of Parkinson’s disease using nocturnal breathing signals
AU - Yang, Yuzhe
AU - Yuan, Yuan
AU - Zhang, Guo
AU - Wang, Hao
AU - Chen, Ying Cong
AU - Liu, Yingcheng
AU - Tarolli, Christopher G.
AU - Crepeau, Daniel
AU - Bukartyk, Jan
AU - Junna, Mithri R.
AU - Videnovic, Aleksandar
AU - Ellis, Terry D.
AU - Lipford, Melissa C.
AU - Dorsey, Ray
AU - Katabi, Dina
N1 - Publisher Copyright:
© 2022, The Author(s).
PY - 2022/10
Y1 - 2022/10
N2 - There are currently no effective biomarkers for diagnosing Parkinson’s disease (PD) or tracking its progression. Here, we developed an artificial intelligence (AI) model to detect PD and track its progression from nocturnal breathing signals. The model was evaluated on a large dataset comprising 7,671 individuals, using data from several hospitals in the United States, as well as multiple public datasets. The AI model can detect PD with an area-under-the-curve of 0.90 and 0.85 on held-out and external test sets, respectively. The AI model can also estimate PD severity and progression in accordance with the Movement Disorder Society Unified Parkinson’s Disease Rating Scale (R = 0.94, P = 3.6 × 10–25). The AI model uses an attention layer that allows for interpreting its predictions with respect to sleep and electroencephalogram. Moreover, the model can assess PD in the home setting in a touchless manner, by extracting breathing from radio waves that bounce off a person’s body during sleep. Our study demonstrates the feasibility of objective, noninvasive, at-home assessment of PD, and also provides initial evidence that this AI model may be useful for risk assessment before clinical diagnosis.
AB - There are currently no effective biomarkers for diagnosing Parkinson’s disease (PD) or tracking its progression. Here, we developed an artificial intelligence (AI) model to detect PD and track its progression from nocturnal breathing signals. The model was evaluated on a large dataset comprising 7,671 individuals, using data from several hospitals in the United States, as well as multiple public datasets. The AI model can detect PD with an area-under-the-curve of 0.90 and 0.85 on held-out and external test sets, respectively. The AI model can also estimate PD severity and progression in accordance with the Movement Disorder Society Unified Parkinson’s Disease Rating Scale (R = 0.94, P = 3.6 × 10–25). The AI model uses an attention layer that allows for interpreting its predictions with respect to sleep and electroencephalogram. Moreover, the model can assess PD in the home setting in a touchless manner, by extracting breathing from radio waves that bounce off a person’s body during sleep. Our study demonstrates the feasibility of objective, noninvasive, at-home assessment of PD, and also provides initial evidence that this AI model may be useful for risk assessment before clinical diagnosis.
UR - http://www.scopus.com/inward/record.url?scp=85136884360&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85136884360&partnerID=8YFLogxK
U2 - 10.1038/s41591-022-01932-x
DO - 10.1038/s41591-022-01932-x
M3 - Article
C2 - 35995955
AN - SCOPUS:85136884360
SN - 1078-8956
VL - 28
SP - 2207
EP - 2215
JO - Nature Medicine
JF - Nature Medicine
IS - 10
ER -