Indirect Assessment of Hyperechogenicity of Substantia Nigra Utilizing Sleep-based Biomarkers

Abstract

Transcranial sonography of the substantia nigra (TCS-SN) may serve as a suitable test for screening groups at a high risk of developing Lewy body diseases (LBDs) such as Parkinson's disease or dementia with Lewy bodies. Although one of the most prominent early markers of these neurodegenerative disorders is the idiopathic rapid eye movement (REM) sleep behavior disorder, the relationship between TCS-SN and sleep alterations has not been fully explored. The aim of this study is to investigate whether sleep-based biomarkers could be used to stratify subjects into three groups with different echogenic areas of the substantia nigra. To achieve this goal, we enrolled 93 participants who underwent TCS-SN and 7-night actigraphy. Additionally, participants completed a sleep diary and the REM sleep behavior disorder screening questionnaire. To assess the severity of pathological echogenicity, we employed a machine learning algorithm utilizing the XGBoost algorithm. The results show that a multimodal assessment of sleep was able to predict the outcomes of TCS-SN with a balanced accuracy of 96 %. Overall, these findings underscore the potential of a comprehensive approach to model the results of TCS-SN and its implications for the prodromal diagnosis of LBDs.Transcranial sonography of the substantia nigra (TCS-SN) may serve as a suitable test for screening groups at a high risk of developing Lewy body diseases (LBDs) such as Parkinson's disease or dementia with Lewy bodies. Although one of the most prominent early markers of these neurodegenerative disorders is the idiopathic rapid eye movement (REM) sleep behavior disorder, the relationship between TCS-SN and sleep alterations has not been fully explored. The aim of this study is to investigate whether sleep-based biomarkers could be used to stratify subjects into three groups with different echogenic areas of the substantia nigra. To achieve this goal, we enrolled 93 participants who underwent TCS-SN and 7-night actigraphy. Additionally, participants completed a sleep diary and the REM sleep behavior disorder screening questionnaire. To assess the severity of pathological echogenicity, we employed a machine learning algorithm utilizing the XGBoost algorithm. The results show that a multimodal assessment of sleep was able to predict the outcomes of TCS-SN with a balanced accuracy of 96 %. Overall, these findings underscore the potential of a comprehensive approach to model the results of TCS-SN and its implications for the prodromal diagnosis of LBDs.