Pub Date : 2024-09-06DOI: 10.1088/1361-6579/ad74d5
Buket Sonbas Cobb, Stephen J Kolb, Seward B Rutkove
Objective.To evaluate electrical impedance myography (EIM) in conjunction with machine learning (ML) to detect infantile spinal muscular atrophy (SMA) and disease progression.Approach. Twenty-six infants with SMA and twenty-seven healthy infants had been enrolled and assessed with EIM as part of the NeuroNEXT SMA biomarker study. We applied a variety of modern, supervised ML approaches to this data, first seeking to differentiate healthy from SMA muscle, and then, using the best method, to track SMA progression.Main Results.Several of the ML algorithms worked well, but linear discriminant analysis (LDA) achieved 88.6% accuracy on subject muscles studied. This contrasts with a maximum of 60% accuracy that could be achieved using the single or multifrequency assessment approaches available at the time. LDA scores were also able to track progression effectively, although a multifrequency reactance-based measure also performed very well in this context.Significance.EIM enhanced with ML promises to be effective for providing effective diagnosis and tracking children and adults with SMA treated with currently available therapies. The normative trends identified here may also inform future applications of the technology in very young children. The basic analyses applied here could also likely be applied to other neuromuscular disorders characterized by muscle atrophy.
目的:
评估电阻抗肌电图(EIM)与机器学习相结合检测小儿脊髓性肌萎缩症(SMA)和疾病进展的效果
方法:
作为 NeuroNEXT SMA 生物标记物研究的一部分,我们对 26 名 SMA 婴儿和 27 名健康婴儿进行了登记和 EIM 评估。我们对这些数据采用了多种现代、有监督的机器学习方法,首先寻求区分健康和 SMA 肌肉,然后使用最佳方法跟踪 SMA 的进展。这与当时使用单频或多频评估方法达到的最高 66% 的准确率形成了鲜明对比。尽管基于多频反应的测量方法在这方面也表现出色,但 LDA 分数也能有效跟踪病情进展。这里确定的标准值和趋势对该技术的其他儿科应用也很有价值。这里应用的基本分析方法也可能适用于其他以肌肉萎缩为特征的神经肌肉疾病。
{"title":"Machine learning-enhanced electrical impedance myography to diagnose and track spinal muscular atrophy progression.","authors":"Buket Sonbas Cobb, Stephen J Kolb, Seward B Rutkove","doi":"10.1088/1361-6579/ad74d5","DOIUrl":"10.1088/1361-6579/ad74d5","url":null,"abstract":"<p><p><i>Objective.</i>To evaluate electrical impedance myography (EIM) in conjunction with machine learning (ML) to detect infantile spinal muscular atrophy (SMA) and disease progression.<i>Approach</i>. Twenty-six infants with SMA and twenty-seven healthy infants had been enrolled and assessed with EIM as part of the NeuroNEXT SMA biomarker study. We applied a variety of modern, supervised ML approaches to this data, first seeking to differentiate healthy from SMA muscle, and then, using the best method, to track SMA progression.<i>Main Results.</i>Several of the ML algorithms worked well, but linear discriminant analysis (LDA) achieved 88.6% accuracy on subject muscles studied. This contrasts with a maximum of 60% accuracy that could be achieved using the single or multifrequency assessment approaches available at the time. LDA scores were also able to track progression effectively, although a multifrequency reactance-based measure also performed very well in this context.<i>Significance.</i>EIM enhanced with ML promises to be effective for providing effective diagnosis and tracking children and adults with SMA treated with currently available therapies. The normative trends identified here may also inform future applications of the technology in very young children. The basic analyses applied here could also likely be applied to other neuromuscular disorders characterized by muscle atrophy.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":null,"pages":null},"PeriodicalIF":2.3,"publicationDate":"2024-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11446569/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142093680","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-04DOI: 10.1088/1361-6579/ad7348
William B Hammert, Ryo Kataoka, Yujiro Yamada, Jun Seob Song, Anna Kang, Robert W Spitz, Jeremy P Loenneke
Progressive overload describes the gradual increase of stress placed on the body during exercise training, and is often quantified (i.e. in resistance training studies) through increases in total training volume (i.e. sets × repetitions × load) from the first to final week of the exercise training intervention. Within the literature, it has become increasingly common for authors to discuss skeletal muscle growth adaptations in the context of increases in total training volume (i.e. the magnitude progression in total training volume). The present manuscript discusses a physiological rationale for progressive overload and then explains why, in our opinion, quantifying the progression of total training volume within research investigations tells very little about muscle growth adaptations to resistance training. Our opinion is based on the following research findings: (1) a noncausal connection between increases in total training volume (i.e. progressively overloading the resistance exercise stimulus) and increases in skeletal muscle size; (2) similar changes in total training volume may not always produce similar increases in muscle size; and (3) the ability to exercise more and consequently amass larger increases in total training volume may not inherently produce more skeletal muscle growth. The methodology of quantifying changes in total training volume may therefore provide a means through which researchers can mathematically determine the total amount of external 'work' performed within a resistance training study. It may not, however, always explain muscle growth adaptations.
渐进性超负荷是指在运动训练过程中身体所承受的压力逐渐增加,通常通过从运动训练干预的第一周到最后一周总训练量(即组数 x 重复次数 x 负荷)的增加来量化(例如,在阻力训练研究中)。在文献中,越来越多的作者结合总训练量的增加(即总训练量的幅度递增)来讨论骨骼肌的生长适应性。本手稿讨论了渐进超负荷的生理学原理,然后解释了为什么我们认为在研究调查中量化总训练量的增长对阻力训练的肌肉生长适应性意义不大。我们的观点基于以下研究结果:(1)总训练量的增加(即逐渐超负荷的阻力训练刺激)与骨骼肌体积的增加之间存在非因果关系;(2)类似的总训练量变化不一定总是会产生类似的肌肉体积增加;以及(3)能够进行更多的锻炼并因此积累更多的总训练量增加可能并不会在本质上产生更多的骨骼肌生长。因此,量化总训练量变化的方法可以为研究人员提供一种手段,通过这种方法,研究人员可以用数学方法确定阻力训练研究中进行的外部 "工作 "总量。但是,这种方法并不能解释肌肉生长的适应性。
{"title":"Progression of total training volume in resistance training studies and its application to skeletal muscle growth.","authors":"William B Hammert, Ryo Kataoka, Yujiro Yamada, Jun Seob Song, Anna Kang, Robert W Spitz, Jeremy P Loenneke","doi":"10.1088/1361-6579/ad7348","DOIUrl":"10.1088/1361-6579/ad7348","url":null,"abstract":"<p><p>Progressive overload describes the gradual increase of stress placed on the body during exercise training, and is often quantified (i.e. in resistance training studies) through increases in total training volume (i.e. sets × repetitions × load) from the first to final week of the exercise training intervention. Within the literature, it has become increasingly common for authors to discuss skeletal muscle growth adaptations in the context of increases in total training volume (i.e. the magnitude progression in total training volume). The present manuscript discusses a physiological rationale for progressive overload and then explains why, in our opinion, quantifying the progression of total training volume within research investigations tells very little about muscle growth adaptations to resistance training. Our opinion is based on the following research findings: (1) a noncausal connection between increases in total training volume (i.e. progressively overloading the resistance exercise stimulus) and increases in skeletal muscle size; (2) similar changes in total training volume may not always produce similar increases in muscle size; and (3) the ability to exercise more and consequently amass larger increases in total training volume may not inherently produce more skeletal muscle growth. The methodology of quantifying changes in total training volume may therefore provide a means through which researchers can mathematically determine the total amount of external 'work' performed within a resistance training study. It may not, however, always explain muscle growth adaptations.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":null,"pages":null},"PeriodicalIF":2.3,"publicationDate":"2024-09-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142046981","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-27DOI: 10.1088/1361-6579/ad6be4
Kai Li, Jiuai Sun
Objective. The widespread adoption of Photoplethysmography (PPG) as a non-invasive method for detecting blood volume variations and deriving vital physiological parameters reflecting health status has surged, primarily due to its accessibility, cost-effectiveness, and non-intrusive nature. This has led to extensive research around this technique in both daily life and clinical applications. Interestingly, despite the existence of contradictory explanations of the underlying mechanism of PPG signals across various applications, a systematic investigation into this crucial matter has not been conducted thus far. This gap in understanding hinders the full exploitation of PPG technology and undermines its accuracy and reliability in numerous applications.Approach. Building upon a comprehensive review of the fundamental principles and technological advancements in PPG, this paper initially attributes the origin of PPG signals to a combination of physical and physiological transmission processes. Furthermore, three distinct models outlining the concerned physiological transmission processes are synthesized, with each model undergoing critical examination based on theoretical underpinnings, empirical evidence, and constraints.Significance. The ultimate objective is to form a fundamental framework for a better understanding of physiological transmission processes in PPG signal generation and to facilitate the development of more reliable technologies for detecting physiological signals.
{"title":"Understanding the physiological transmission mechanisms of photoplethysmography signals: a comprehensive review.","authors":"Kai Li, Jiuai Sun","doi":"10.1088/1361-6579/ad6be4","DOIUrl":"10.1088/1361-6579/ad6be4","url":null,"abstract":"<p><p><i>Objective</i>. The widespread adoption of Photoplethysmography (PPG) as a non-invasive method for detecting blood volume variations and deriving vital physiological parameters reflecting health status has surged, primarily due to its accessibility, cost-effectiveness, and non-intrusive nature. This has led to extensive research around this technique in both daily life and clinical applications. Interestingly, despite the existence of contradictory explanations of the underlying mechanism of PPG signals across various applications, a systematic investigation into this crucial matter has not been conducted thus far. This gap in understanding hinders the full exploitation of PPG technology and undermines its accuracy and reliability in numerous applications.<i>Approach</i>. Building upon a comprehensive review of the fundamental principles and technological advancements in PPG, this paper initially attributes the origin of PPG signals to a combination of physical and physiological transmission processes. Furthermore, three distinct models outlining the concerned physiological transmission processes are synthesized, with each model undergoing critical examination based on theoretical underpinnings, empirical evidence, and constraints.<i>Significance</i>. The ultimate objective is to form a fundamental framework for a better understanding of physiological transmission processes in PPG signal generation and to facilitate the development of more reliable technologies for detecting physiological signals.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":null,"pages":null},"PeriodicalIF":2.3,"publicationDate":"2024-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141898021","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-22DOI: 10.1088/1361-6579/ad6acf
Marta Carrara, Riccardo Campitelli, Diletta Guberti, M Ignacio Monge Garcia, Manuela Ferrario
Objective.Arterial pulse wave analysis (PWA) is now established as a powerful tool to investigate the cardiovascular system, and several clinical studies have shown how PWA can provide valuable prognostic information over and beyond traditional cardiovascular risk factors. Typically these techniques are applied to chronic conditions, such as hypertension or aging, to monitor the slow structural changes of the vascular system which lead to important alterations of the arterial PW. However, their application to acute critical illness is not currently widespread, probably because of the high hemodynamic instability and acute dynamic alterations affecting the cardiovascular system of these patients.Approach.In this work we propose a review of the physiological and methodological basis of PWA, describing how it can be used to provide insights into arterial structure and function, cardiovascular biomechanical properties, and to derive information on wave propagation and reflection.Main results.The applicability of these techniques to acute critical illness, especially septic shock, is extensively discussed, highlighting the feasibility of their use in acute critical patients and their role in optimizing therapy administration and hemodynamic monitoring.Significance.The potential for the clinical use of these techniques lies in the ease of computation and availability of arterial blood pressure signals, as invasive arterial lines are commonly used in these patients. We hope that the concepts illustrated in the present review will soon be translated into clinical practice.
{"title":"The role of pulse wave analysis indexes for critically ill patients: a narrative review.","authors":"Marta Carrara, Riccardo Campitelli, Diletta Guberti, M Ignacio Monge Garcia, Manuela Ferrario","doi":"10.1088/1361-6579/ad6acf","DOIUrl":"10.1088/1361-6579/ad6acf","url":null,"abstract":"<p><p><i>Objective.</i>Arterial pulse wave analysis (PWA) is now established as a powerful tool to investigate the cardiovascular system, and several clinical studies have shown how PWA can provide valuable prognostic information over and beyond traditional cardiovascular risk factors. Typically these techniques are applied to chronic conditions, such as hypertension or aging, to monitor the slow structural changes of the vascular system which lead to important alterations of the arterial PW. However, their application to acute critical illness is not currently widespread, probably because of the high hemodynamic instability and acute dynamic alterations affecting the cardiovascular system of these patients.<i>Approach.</i>In this work we propose a review of the physiological and methodological basis of PWA, describing how it can be used to provide insights into arterial structure and function, cardiovascular biomechanical properties, and to derive information on wave propagation and reflection.<i>Main results.</i>The applicability of these techniques to acute critical illness, especially septic shock, is extensively discussed, highlighting the feasibility of their use in acute critical patients and their role in optimizing therapy administration and hemodynamic monitoring.<i>Significance.</i>The potential for the clinical use of these techniques lies in the ease of computation and availability of arterial blood pressure signals, as invasive arterial lines are commonly used in these patients. We hope that the concepts illustrated in the present review will soon be translated into clinical practice.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":null,"pages":null},"PeriodicalIF":2.3,"publicationDate":"2024-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141879232","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-22DOI: 10.1088/1361-6579/ad6529
Chunmiao Liang, Qinghua Sun, Jiali Li, Bing Ji, Weiming Wu, Fukai Zhang, Yuguo Chen, Cong Wang
Objective.In recent years, artificial intelligence-based electrocardiogram (ECG) methods have been massively applied to myocardial infarction (MI). However, the joint analysis of static and dynamic features to achieve accurate and interpretable MI detection has not been comprehensively addressed.Approach.This paper proposes a simplified ensemble tree method with a joint analysis of static and dynamic features to solve this issue for MI detection. Initially, the dynamic features are extracted by modeling the intrinsic dynamics of ECG via dynamic learning in addition to extracting classical static features. Secondly, a two-stage feature selection strategy is designed to identify a few significant features, which substitute the original variables that are employed in constructing the ensemble tree. This approach enhances the discriminative ability by selecting significant static and dynamic features. Subsequently, this paper presents an interpretable classification method named StackTree by introducing a stacked ensemble scheme to modify the ensemble tree simplification algorithm. The representative rules of the raw ensemble trees are selected as the intermediate training data that is used to retrain a decision tree with performance close to that of the source ensemble model. Using this scheme, the significant precision and interpretability of MI detection are thus comprehensively addressed.Main results.The effectiveness of our method in detecting MI is evaluated using the Physikalisch-Technische Bundesanstalt (PTB) and clinical database. The findings suggest that our algorithm outperforms the traditional methods based on a single type of feature. Additionally, it is comparable to the conventional random forest, achieving 97.1% accuracy under the inter-patient framework on the PTB database. Furthermore, feature subsets trained on PTB are validated using the clinical database, resulting in an accuracy of 84.5%. The chosen important features demonstrate that both static and dynamic information have crucial roles in MI detection. Crucially, the proposed method provides clear internal workings in an easy-to-understand visual manner.
目的:近年来,基于人工智能的心电图(ECG)方法被大量应用于心肌梗死(MI)。然而,如何通过对静态和动态特征的联合分析来实现准确、可解释的心肌梗死检测,还没有得到全面解决。本文提出了一种联合分析静态和动态特征的简化集合树方法,以解决 MI 检测中的这一问题。首先,在提取经典静态特征的基础上,通过动态学习对心电图的内在动态进行建模,从而提取动态特征。其次,设计了一种两阶段特征选择策略,以识别少数重要特征,这些特征可替代用于构建集合树的原始变量。这种方法通过选择重要的静态和动态特征来增强判别能力。随后,本文通过引入堆叠集合方案来修改集合树简化算法,提出了一种名为 StackTree 的可解释分类方法。原始集合树中具有代表性的规则被选为中间训练数据,用于重新训练一棵性能接近源集合模型的决策树。采用这种方案,可以全面解决 MI 检测的高精度和可解释性问题。我们使用 PTB 和临床数据库评估了我们的方法在检测 MI 方面的有效性。结果表明,我们的算法优于基于单一类型特征的传统方法。此外,在 PTB 数据库的患者间框架下,该算法的准确率达到了 97.1%,与传统的随机森林算法不相上下。此外,使用临床数据库验证了在 PTB 上训练的特征子集,结果准确率达到 84.5%。所选的重要特征表明,静态和动态信息在 MI 检测中都起着至关重要的作用。最重要的是,所提出的方法以易于理解的可视化方式提供了清晰的内部工作原理。
{"title":"An interpretable ensemble trees method with joint analysis of static and dynamic features for myocardial infarction detection.","authors":"Chunmiao Liang, Qinghua Sun, Jiali Li, Bing Ji, Weiming Wu, Fukai Zhang, Yuguo Chen, Cong Wang","doi":"10.1088/1361-6579/ad6529","DOIUrl":"10.1088/1361-6579/ad6529","url":null,"abstract":"<p><p><i>Objective.</i>In recent years, artificial intelligence-based electrocardiogram (ECG) methods have been massively applied to myocardial infarction (MI). However, the joint analysis of static and dynamic features to achieve accurate and interpretable MI detection has not been comprehensively addressed.<i>Approach.</i>This paper proposes a simplified ensemble tree method with a joint analysis of static and dynamic features to solve this issue for MI detection. Initially, the dynamic features are extracted by modeling the intrinsic dynamics of ECG via dynamic learning in addition to extracting classical static features. Secondly, a two-stage feature selection strategy is designed to identify a few significant features, which substitute the original variables that are employed in constructing the ensemble tree. This approach enhances the discriminative ability by selecting significant static and dynamic features. Subsequently, this paper presents an interpretable classification method named StackTree by introducing a stacked ensemble scheme to modify the ensemble tree simplification algorithm. The representative rules of the raw ensemble trees are selected as the intermediate training data that is used to retrain a decision tree with performance close to that of the source ensemble model. Using this scheme, the significant precision and interpretability of MI detection are thus comprehensively addressed.<i>Main results.</i>The effectiveness of our method in detecting MI is evaluated using the Physikalisch-Technische Bundesanstalt (PTB) and clinical database. The findings suggest that our algorithm outperforms the traditional methods based on a single type of feature. Additionally, it is comparable to the conventional random forest, achieving 97.1% accuracy under the inter-patient framework on the PTB database. Furthermore, feature subsets trained on PTB are validated using the clinical database, resulting in an accuracy of 84.5%. The chosen important features demonstrate that both static and dynamic information have crucial roles in MI detection. Crucially, the proposed method provides clear internal workings in an easy-to-understand visual manner.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":null,"pages":null},"PeriodicalIF":2.3,"publicationDate":"2024-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141724158","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-19DOI: 10.1088/1361-6579/ad68c1
Richard Bayford, Rosalind Sadleir, Inéz Frerichs, Tong In Oh, Steffen Leonhardt
Scope. This focus collection aims at presenting recent advances in electrical impedance tomography (EIT), including algorithms, hardware, and clinical applications.Editorial. This focus collection of articles published by the journalPhysiological Measurementintroduces the Progress in EIT and Bioimpedance. It follows conferences in South Korea and Germany, that provided a platform for new research ideas.
{"title":"Progress in electrical impedance tomography and bioimpedance.","authors":"Richard Bayford, Rosalind Sadleir, Inéz Frerichs, Tong In Oh, Steffen Leonhardt","doi":"10.1088/1361-6579/ad68c1","DOIUrl":"10.1088/1361-6579/ad68c1","url":null,"abstract":"<p><p><i>Scope</i>. This focus collection aims at presenting recent advances in electrical impedance tomography (EIT), including algorithms, hardware, and clinical applications.<i>Editorial</i>. This focus collection of articles published by the journal<i>Physiological Measurement</i>introduces the Progress in EIT and Bioimpedance. It follows conferences in South Korea and Germany, that provided a platform for new research ideas.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":null,"pages":null},"PeriodicalIF":2.3,"publicationDate":"2024-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141793064","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-14DOI: 10.1088/1361-6579/ad69fd
Niema M Pahlevan, Rashid Alavi, Jing Liu, Melissa Ramos, Antreas Hindoyan, Ray V Matthews
Objective.Instantaneous, non-invasive evaluation of left ventricular end-diastolic pressure (LVEDP) would have significant value in the diagnosis and treatment of heart failure. A new approach called cardiac triangle mapping (CTM) has been recently proposed, which can provide a non-invasive estimate of LVEDP. We hypothesized that a hybrid machine-learning (ML) method based on CTM can instantaneously identify an elevated LVEDP using simultaneously measured femoral pressure waveform and electrocardiogram (ECG).Approach.We studied 46 patients (Age: 39-90 (66.4 ± 9.9), BMI: 20.2-36.8 (27.6 ± 4.1), 12 females) scheduled for clinical left heart catheterizations or coronary angiograms at University of Southern California Keck Medical Center. Exclusion criteria included severe mitral/aortic valve disease; severe carotid stenosis; aortic abnormalities; ventricular paced rhythm; left bundle branch and anterior fascicular blocks; interventricular conduction delay; and atrial fibrillation. Invasive LVEDP and pressure waveforms at the iliac bifurcation were measured using transducer-tipped Millar catheters with simultaneous ECG. LVEDP range was 9.3-40.5 mmHg. LVEDP = 18 mmHg was used as cutoff. Random forest (RF) classifiers were trained using data from 36 patients and blindly tested on 10 patients.Main results.Our proposed ML classifier models accurately predict true LVEDP classes using appropriate physics-based features, where the most accurate demonstrates 100.0% (elevated) and 80.0% (normal) success in predicting true LVEDP classes on blind data.Significance.We demonstrated that physics-based ML models can instantaneously classify LVEDP using information from femoral waveforms and ECGs. Although an invasive validation, the required ML inputs can be potentially obtained non-invasively.
{"title":"Detecting elevated left ventricular end diastolic pressure from simultaneously measured femoral pressure waveform and electrocardiogram.","authors":"Niema M Pahlevan, Rashid Alavi, Jing Liu, Melissa Ramos, Antreas Hindoyan, Ray V Matthews","doi":"10.1088/1361-6579/ad69fd","DOIUrl":"10.1088/1361-6579/ad69fd","url":null,"abstract":"<p><p><i>Objective.</i>Instantaneous, non-invasive evaluation of left ventricular end-diastolic pressure (LVEDP) would have significant value in the diagnosis and treatment of heart failure. A new approach called cardiac triangle mapping (CTM) has been recently proposed, which can provide a non-invasive estimate of LVEDP. We hypothesized that a hybrid machine-learning (ML) method based on CTM can instantaneously identify an elevated LVEDP using simultaneously measured femoral pressure waveform and electrocardiogram (ECG).<i>Approach.</i>We studied 46 patients (Age: 39-90 (66.4 ± 9.9), BMI: 20.2-36.8 (27.6 ± 4.1), 12 females) scheduled for clinical left heart catheterizations or coronary angiograms at University of Southern California Keck Medical Center. Exclusion criteria included severe mitral/aortic valve disease; severe carotid stenosis; aortic abnormalities; ventricular paced rhythm; left bundle branch and anterior fascicular blocks; interventricular conduction delay; and atrial fibrillation. Invasive LVEDP and pressure waveforms at the iliac bifurcation were measured using transducer-tipped Millar catheters with simultaneous ECG. LVEDP range was 9.3-40.5 mmHg. LVEDP = 18 mmHg was used as cutoff. Random forest (RF) classifiers were trained using data from 36 patients and blindly tested on 10 patients.<i>Main results.</i>Our proposed ML classifier models accurately predict true LVEDP classes using appropriate physics-based features, where the most accurate demonstrates 100.0% (elevated) and 80.0% (normal) success in predicting true LVEDP classes on blind data.<i>Significance.</i>We demonstrated that physics-based ML models can instantaneously classify LVEDP using information from femoral waveforms and ECGs. Although an invasive validation, the required ML inputs can be potentially obtained non-invasively.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":null,"pages":null},"PeriodicalIF":2.3,"publicationDate":"2024-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141860590","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-12DOI: 10.1088/1361-6579/ad6746
Andrew Barros, Ian German Mesner, N Rich Nguyen, J Randall Moorman
Objective.The 12-lead electrocardiogram (ECG) is routine in clinical use and deep learning approaches have been shown to have the identify features not immediately apparent to human interpreters including age and sex. Several models have been published but no direct comparisons exist.Approach.We implemented three previously published models and one unpublished model to predict age and sex from a 12-lead ECG and then compared their performance on an open-access data set.Main results.All models converged and were evaluated on the holdout set. The best preforming age prediction model had a hold-out set mean absolute error of 8.06 years. The best preforming sex prediction model had a hold-out set area under the receiver operating curve of 0.92.Significance.We compared performance of four models on an open-access dataset.
{"title":"Age prediction from 12-lead electrocardiograms using deep learning: a comparison of four models on a contemporary, freely available dataset.","authors":"Andrew Barros, Ian German Mesner, N Rich Nguyen, J Randall Moorman","doi":"10.1088/1361-6579/ad6746","DOIUrl":"10.1088/1361-6579/ad6746","url":null,"abstract":"<p><p><i>Objective.</i>The 12-lead electrocardiogram (ECG) is routine in clinical use and deep learning approaches have been shown to have the identify features not immediately apparent to human interpreters including age and sex. Several models have been published but no direct comparisons exist.<i>Approach.</i>We implemented three previously published models and one unpublished model to predict age and sex from a 12-lead ECG and then compared their performance on an open-access data set.<i>Main results.</i>All models converged and were evaluated on the holdout set. The best preforming age prediction model had a hold-out set mean absolute error of 8.06 years. The best preforming sex prediction model had a hold-out set area under the receiver operating curve of 0.92.<i>Significance.</i>We compared performance of four models on an open-access dataset.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":null,"pages":null},"PeriodicalIF":2.3,"publicationDate":"2024-08-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11334242/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141760279","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Objective. Physiological data are often low quality and thereby compromises the effectiveness of related health monitoring. The primary goal of this study is to develop a robust foundation model that can effectively handle low-quality issue in physiological data.Approach. We introduce SiamQuality, a self-supervised learning approach using convolutional neural networks (CNNs) as the backbone. SiamQuality learns to generate similar representations for both high and low quality photoplethysmography (PPG) signals that originate from similar physiological states. We leveraged a substantial dataset of PPG signals from hospitalized intensive care patients, comprised of over 36 million 30 s PPG pairs.Main results. After pre-training the SiamQuality model, it was fine-tuned and tested on six PPG downstream tasks focusing on cardiovascular monitoring. Notably, in tasks such as respiratory rate estimation and atrial fibrillation detection, the model's performance exceeded the state-of-the-art by 75% and 5%, respectively. The results highlight the effectiveness of our model across all evaluated tasks, demonstrating significant improvements, especially in applications for heart monitoring on wearable devices.Significance. This study underscores the potential of CNNs as a robust backbone for foundation models tailored to physiological data, emphasizing their capability to maintain performance despite variations in data quality. The success of the SiamQuality model in handling real-world, variable-quality data opens new avenues for the development of more reliable and efficient healthcare monitoring technologies.
{"title":"SiamQuality: a ConvNet-based foundation model for photoplethysmography signals.","authors":"Cheng Ding, Zhicheng Guo, Zhaoliang Chen, Randall J Lee, Cynthia Rudin, Xiao Hu","doi":"10.1088/1361-6579/ad6747","DOIUrl":"10.1088/1361-6579/ad6747","url":null,"abstract":"<p><p><i>Objective</i>. Physiological data are often low quality and thereby compromises the effectiveness of related health monitoring. The primary goal of this study is to develop a robust foundation model that can effectively handle low-quality issue in physiological data.<i>Approach</i>. We introduce SiamQuality, a self-supervised learning approach using convolutional neural networks (CNNs) as the backbone. SiamQuality learns to generate similar representations for both high and low quality photoplethysmography (PPG) signals that originate from similar physiological states. We leveraged a substantial dataset of PPG signals from hospitalized intensive care patients, comprised of over 36 million 30 s PPG pairs.<i>Main results</i>. After pre-training the SiamQuality model, it was fine-tuned and tested on six PPG downstream tasks focusing on cardiovascular monitoring. Notably, in tasks such as respiratory rate estimation and atrial fibrillation detection, the model's performance exceeded the state-of-the-art by 75% and 5%, respectively. The results highlight the effectiveness of our model across all evaluated tasks, demonstrating significant improvements, especially in applications for heart monitoring on wearable devices.<i>Significance</i>. This study underscores the potential of CNNs as a robust backbone for foundation models tailored to physiological data, emphasizing their capability to maintain performance despite variations in data quality. The success of the SiamQuality model in handling real-world, variable-quality data opens new avenues for the development of more reliable and efficient healthcare monitoring technologies.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":null,"pages":null},"PeriodicalIF":2.3,"publicationDate":"2024-08-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11334241/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141760233","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-06DOI: 10.1088/1361-6579/ad66aa
Riki Shimizu, Hau-Tieng Wu
Objective.Sleep spindles contain crucial brain dynamics information. We introduce the novel non-linear time-frequency (TF) analysis tool 'Concentration of Frequency and Time' (ConceFT) to create an interpretable automated algorithm for sleep spindle annotation in EEG data and to measure spindle instantaneous frequencies (IFs).Approach.ConceFT effectively reduces stochastic EEG influence, enhancing spindle visibility in the TF representation. Our automated spindle detection algorithm, ConceFT-Spindle (ConceFT-S), is compared to A7 (non-deep learning) and SUMO (deep learning) using Dream and Montreal Archive of Sleep Studies (MASS) benchmark databases. We also quantify spindle IF dynamics.Main results.ConceFT-S achieves F1 scores of 0.765 in Dream and 0.791 in MASS, which surpass A7 and SUMO. We reveal that spindle IF is generally nonlinear.Significance.ConceFT offers an accurate, interpretable EEG-based sleep spindle detection algorithm and enables spindle IF quantification.
目的睡眠棘波包含重要的大脑动力学信息。我们介绍了新颖的非线性时频分析工具 "频率和时间的集中"(ConceFT),以创建一种可解释的自动算法,用于在脑电图数据中标注睡眠纺锤体,并测量纺锤体的瞬时频率(IFs):方法:ConceFT 可有效降低随机脑电图的流变性,提高主轴在时频表征中的可见度。我们的自动纺锤体检测算法 ConceFT-Spindle(ConceFT-S)使用 Dream 和 MASS 基准数据库与 A7(非深度学习)和 SUMO(深度学习)进行了比较。我们还量化了主轴中频动态。主要结果:ConceFT-S 在 Dream 和 MASS 中的 F1 分数分别为 0.765 和 0.791,超过了 A7 和 SUMO。我们发现纺锤体中频一般是非线性的:ConceFT提供了一种准确、可解释的基于脑电图的睡眠纺锤体检测算法,并能对纺锤体中频进行量化。
{"title":"Unveil sleep spindles with concentration of frequency and time (ConceFT).","authors":"Riki Shimizu, Hau-Tieng Wu","doi":"10.1088/1361-6579/ad66aa","DOIUrl":"10.1088/1361-6579/ad66aa","url":null,"abstract":"<p><p><i>Objective.</i>Sleep spindles contain crucial brain dynamics information. We introduce the novel non-linear time-frequency (TF) analysis tool 'Concentration of Frequency and Time' (ConceFT) to create an interpretable automated algorithm for sleep spindle annotation in EEG data and to measure spindle instantaneous frequencies (IFs).<i>Approach.</i>ConceFT effectively reduces stochastic EEG influence, enhancing spindle visibility in the TF representation. Our automated spindle detection algorithm, ConceFT-Spindle (ConceFT-S), is compared to A7 (non-deep learning) and SUMO (deep learning) using Dream and Montreal Archive of Sleep Studies (MASS) benchmark databases. We also quantify spindle IF dynamics.<i>Main results.</i>ConceFT-S achieves F1 scores of 0.765 in Dream and 0.791 in MASS, which surpass A7 and SUMO. We reveal that spindle IF is generally nonlinear.<i>Significance.</i>ConceFT offers an accurate, interpretable EEG-based sleep spindle detection algorithm and enables spindle IF quantification.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":null,"pages":null},"PeriodicalIF":2.3,"publicationDate":"2024-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141748779","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}