Artificial intelligence(AI)has impacted many areas of healthcare.AI in healthcare uses machine learning,deep learning,and natural language processing to analyze copious amounts of healthcare data and yield valuable ou...Artificial intelligence(AI)has impacted many areas of healthcare.AI in healthcare uses machine learning,deep learning,and natural language processing to analyze copious amounts of healthcare data and yield valuable outcomes.In the sleep medicine field,a large amount of physiological data is gathered compared to other branches of medicine.This field is primed for innovations with the help of AI.A good quality of sleep is crucial for optimal health.About one billion people are estimated to have obstructive sleep apnea worldwide,but it is difficult to diagnose and treat all the people with limited resources.Sleep apnea is one of the major contributors to poor health.Most of the sleep apnea patients remain undiagnosed.Those diagnosed with sleep apnea have difficulty getting it optimally treated due to several factors,and AI can help in this situation.AI can also help in the diagnosis and management of other sleep disorders such as insomnia,hypersomnia,parasomnia,narcolepsy,shift work sleep disorders,periodic leg movement disorders,etc.In this manuscript,we aim to address three critical issues about the use of AI in sleep medicine:(1)How can AI help in diagnosing and treating sleep disorders?(2)How can AI fill the gap in the care of sleep disorders?and(3)What are the ethical and legal considerations of using AI in sleep medicine?展开更多
Sleep posture monitoring is an essential assessment for obstructive sleep apnea(OSA)patients.The objective of this study is to develop a machine learning-based sleep posture recognition system using a dual ultra-wideb...Sleep posture monitoring is an essential assessment for obstructive sleep apnea(OSA)patients.The objective of this study is to develop a machine learning-based sleep posture recognition system using a dual ultra-wideband radar system.We collected radiofrequency data from two radars positioned over and at the side of the bed for 16 patients performing four sleep postures(supine,left and right lateral,and prone).We proposed and evaluated deep learning approaches that streamlined feature extraction and classification,and the traditional machine learning approaches that involved different combinations of feature extractors and classifiers.Our results showed that the dual radar system performed better than either single radar.Predetermined statistical features with random forest classifier yielded the best accuracy(0.887),which could be further improved via an ablation study(0.938).Deep learning approach using transformer yielded accuracy of 0.713.展开更多
针对睡眠多导图中各模态信息在睡眠各阶段存在差异性,而导致特征利用不充分的问题,本文提出了一种基于通道注意力机制和多模态门控机制的睡眠分期模型。首先利用残差收缩网络设计各模态特征提取网络用于提取各模态特征,并在通道维度上...针对睡眠多导图中各模态信息在睡眠各阶段存在差异性,而导致特征利用不充分的问题,本文提出了一种基于通道注意力机制和多模态门控机制的睡眠分期模型。首先利用残差收缩网络设计各模态特征提取网络用于提取各模态特征,并在通道维度上进行拼接融合,利用通道注意力机制进一步对融合特征进行重标定得到睡眠多导图的时不变特征;之后提出了一种基于自适应门控机制的多模态门控模块,对各模态特征及时不变特征按照重要程度进行加权融合,实现特征融合;最后利用双向长短时记忆网络提取睡眠多导图的时序特征。实验结果表明,本文提出的睡眠分期模型在欧洲数据格式睡眠数据集(sleep-European data format, sleep-EDF)上准确率为87.6%,M_(F1)为82.0%,取得了目前最好的分期效果。展开更多
AIM:To detect retinal microvascular variations in obstructive sleep apnea syndrome patients.METHODS:This prospective,observational case-control study included healthy controls and patients with mild,moderate,and sever...AIM:To detect retinal microvascular variations in obstructive sleep apnea syndrome patients.METHODS:This prospective,observational case-control study included healthy controls and patients with mild,moderate,and severe obstructive sleep apnea syndrome.Vascular parameters,foveal avascular area,and flow areas in macula-centered,6.00×6.00 mm2 scan size optical coherence tomography angiography images were compared.RESULTS:The control group had the highest whole image,parafoveal,and perifoveal vessel density among the groups in both superficial and the deep capillary plexus(all P<0.05).Rapid eye movement sleep apnoea-hypopnoea index was reversely correlated with whole(Rho=-0.195,P=0.034),parafoveal(Rho=-0.242,P=0.008),perifoveal(Rho=-0.187,P=0.045)vessel density in the superficial capillary plexus,and whole(Rho=-0.186,P=0.046),parafoveal(Rho=-0.260,P=0.004),perifoveal(Rho=-0.189,P=0.043)vessel density in the deep capillary plexus,though the mean and non-rapid eye movement sleep apnoeahypopnoea index related with only parafoveal vessel density in the superficial capillary plexus(Rho=-0.213,P=0.020;Rho=-0.191,P=0.038)and the deep capillary plexus(Rho=-0.254,P=0.005;Rho=-0.194,P=0.035).CONCLUSION:This study shows decreased vessel density and its reverse correlation with the apnoea-hypopnoea index in patients with obstructive sleep apnea syndrome.展开更多
BACKGROUND Sedation is commonly performed for the endoscopic submucosal dissection(ESD)of early gastric cancer.Severe hypoxemia occasionally occurs due to the respiratory depression during sedation.AIM To establish pr...BACKGROUND Sedation is commonly performed for the endoscopic submucosal dissection(ESD)of early gastric cancer.Severe hypoxemia occasionally occurs due to the respiratory depression during sedation.AIM To establish predictive models for respiratory depression during sedation for ESD.METHODS Thirty-five adult patients undergoing sedation using propofol and pentazocine for gastric ESDs participated in this prospective observational study.Preoperatively,a portable sleep monitor and STOP questionnaires,which are the established screening tools for sleep apnea syndrome,were utilized.Respiration during sedation was assessed by a standard polysomnography technique including the pulse oximeter,nasal pressure sensor,nasal thermistor sensor,and chest and abdominal respiratory motion sensors.The apnea-hypopnea index(AHI)was obtained using a preoperative portable sleep monitor and polysomnography during ESD.A predictive model for the AHI during sedation was developed using either the preoperative AHI or STOP questionnaire score.RESULTS All ESDs were completed successfully and without complications.Seventeen patients(49%)had a preoperative AHI greater than 5/h.The intraoperative AHI was significantly greater than the preoperative AHI(12.8±7.6 events/h vs 9.35±11.0 events/h,P=0.049).Among the potential predictive variables,age,body mass index,STOP questionnaire score,and preoperative AHI were significantly correlated with AHI during sedation.Multiple linear regression analysis determined either STOP questionnaire score or preoperative AHI as independent predictors for intraoperative AHI≥30/h(area under the curve[AUC]:0.707 and 0.833,respectively)and AHI between 15 and 30/h(AUC:0.761 and 0.778,respectively).CONCLUSION The cost-effective STOP questionnaire shows performance for predicting abnormal breathing during sedation for ESD that was equivalent to that of preoperative portable sleep monitoring.展开更多
文摘Artificial intelligence(AI)has impacted many areas of healthcare.AI in healthcare uses machine learning,deep learning,and natural language processing to analyze copious amounts of healthcare data and yield valuable outcomes.In the sleep medicine field,a large amount of physiological data is gathered compared to other branches of medicine.This field is primed for innovations with the help of AI.A good quality of sleep is crucial for optimal health.About one billion people are estimated to have obstructive sleep apnea worldwide,but it is difficult to diagnose and treat all the people with limited resources.Sleep apnea is one of the major contributors to poor health.Most of the sleep apnea patients remain undiagnosed.Those diagnosed with sleep apnea have difficulty getting it optimally treated due to several factors,and AI can help in this situation.AI can also help in the diagnosis and management of other sleep disorders such as insomnia,hypersomnia,parasomnia,narcolepsy,shift work sleep disorders,periodic leg movement disorders,etc.In this manuscript,we aim to address three critical issues about the use of AI in sleep medicine:(1)How can AI help in diagnosing and treating sleep disorders?(2)How can AI fill the gap in the care of sleep disorders?and(3)What are the ethical and legal considerations of using AI in sleep medicine?
基金supported by General Research Fund from the Research Grants Council of Hong Kong,China (Project No.PolyU15223822)Internal fund from the Research Institute for Smart Ageing (Project No.P0039001)Department of Biomedical Engineering (Project No.P0033913 and P0035896)from the Hong Kong Polytechnic University.
文摘Sleep posture monitoring is an essential assessment for obstructive sleep apnea(OSA)patients.The objective of this study is to develop a machine learning-based sleep posture recognition system using a dual ultra-wideband radar system.We collected radiofrequency data from two radars positioned over and at the side of the bed for 16 patients performing four sleep postures(supine,left and right lateral,and prone).We proposed and evaluated deep learning approaches that streamlined feature extraction and classification,and the traditional machine learning approaches that involved different combinations of feature extractors and classifiers.Our results showed that the dual radar system performed better than either single radar.Predetermined statistical features with random forest classifier yielded the best accuracy(0.887),which could be further improved via an ablation study(0.938).Deep learning approach using transformer yielded accuracy of 0.713.
文摘针对睡眠多导图中各模态信息在睡眠各阶段存在差异性,而导致特征利用不充分的问题,本文提出了一种基于通道注意力机制和多模态门控机制的睡眠分期模型。首先利用残差收缩网络设计各模态特征提取网络用于提取各模态特征,并在通道维度上进行拼接融合,利用通道注意力机制进一步对融合特征进行重标定得到睡眠多导图的时不变特征;之后提出了一种基于自适应门控机制的多模态门控模块,对各模态特征及时不变特征按照重要程度进行加权融合,实现特征融合;最后利用双向长短时记忆网络提取睡眠多导图的时序特征。实验结果表明,本文提出的睡眠分期模型在欧洲数据格式睡眠数据集(sleep-European data format, sleep-EDF)上准确率为87.6%,M_(F1)为82.0%,取得了目前最好的分期效果。
文摘AIM:To detect retinal microvascular variations in obstructive sleep apnea syndrome patients.METHODS:This prospective,observational case-control study included healthy controls and patients with mild,moderate,and severe obstructive sleep apnea syndrome.Vascular parameters,foveal avascular area,and flow areas in macula-centered,6.00×6.00 mm2 scan size optical coherence tomography angiography images were compared.RESULTS:The control group had the highest whole image,parafoveal,and perifoveal vessel density among the groups in both superficial and the deep capillary plexus(all P<0.05).Rapid eye movement sleep apnoea-hypopnoea index was reversely correlated with whole(Rho=-0.195,P=0.034),parafoveal(Rho=-0.242,P=0.008),perifoveal(Rho=-0.187,P=0.045)vessel density in the superficial capillary plexus,and whole(Rho=-0.186,P=0.046),parafoveal(Rho=-0.260,P=0.004),perifoveal(Rho=-0.189,P=0.043)vessel density in the deep capillary plexus,though the mean and non-rapid eye movement sleep apnoeahypopnoea index related with only parafoveal vessel density in the superficial capillary plexus(Rho=-0.213,P=0.020;Rho=-0.191,P=0.038)and the deep capillary plexus(Rho=-0.254,P=0.005;Rho=-0.194,P=0.035).CONCLUSION:This study shows decreased vessel density and its reverse correlation with the apnoea-hypopnoea index in patients with obstructive sleep apnea syndrome.
基金Supported by Japan Society for the Promotion of Science KAKENHI,No.15K09056.
文摘BACKGROUND Sedation is commonly performed for the endoscopic submucosal dissection(ESD)of early gastric cancer.Severe hypoxemia occasionally occurs due to the respiratory depression during sedation.AIM To establish predictive models for respiratory depression during sedation for ESD.METHODS Thirty-five adult patients undergoing sedation using propofol and pentazocine for gastric ESDs participated in this prospective observational study.Preoperatively,a portable sleep monitor and STOP questionnaires,which are the established screening tools for sleep apnea syndrome,were utilized.Respiration during sedation was assessed by a standard polysomnography technique including the pulse oximeter,nasal pressure sensor,nasal thermistor sensor,and chest and abdominal respiratory motion sensors.The apnea-hypopnea index(AHI)was obtained using a preoperative portable sleep monitor and polysomnography during ESD.A predictive model for the AHI during sedation was developed using either the preoperative AHI or STOP questionnaire score.RESULTS All ESDs were completed successfully and without complications.Seventeen patients(49%)had a preoperative AHI greater than 5/h.The intraoperative AHI was significantly greater than the preoperative AHI(12.8±7.6 events/h vs 9.35±11.0 events/h,P=0.049).Among the potential predictive variables,age,body mass index,STOP questionnaire score,and preoperative AHI were significantly correlated with AHI during sedation.Multiple linear regression analysis determined either STOP questionnaire score or preoperative AHI as independent predictors for intraoperative AHI≥30/h(area under the curve[AUC]:0.707 and 0.833,respectively)and AHI between 15 and 30/h(AUC:0.761 and 0.778,respectively).CONCLUSION The cost-effective STOP questionnaire shows performance for predicting abnormal breathing during sedation for ESD that was equivalent to that of preoperative portable sleep monitoring.
