Empirical investigation of multi-source cross-validation in clinical ECG classification.

Traditionally, machine learning-based clinical prediction models have been trained and evaluated on patient data from a single source, such as a hospital. Cross-validation methods can be used to estimate the accuracy of such models on new patients originating from the same source, by repeated random splitting of the data. However, such estimates tend to be highly overoptimistic when compared to accuracy obtained from deploying models to sources not represented in the dataset, such as a new hospital. The increasing availability of multi-source medical datasets provides new opportunities for obtaining more comprehensive and realistic evaluations of expected accuracy through source-level cross-validation designs. In this study, we present a systematic empirical evaluation of standard K-fold cross-validation and leave-source-out cross-validation methods in a multi-source setting. We consider the task of electrocardiogram based cardiovascular disease classification, combining and harmonizing the openly available PhysioNet/CinC Challenge 2021 and the Shandong Provincial Hospital datasets for our study. Our results show that K-fold cross-validation, both on single-source and multi-source data, systemically overestimates prediction performance when the end goal is to generalize to new sources. Leave-source-out cross-validation provides more reliable performance estimates, having close to zero bias though larger variability. The evaluation highlights the dangers of obtaining misleading cross-validation results on medical data and demonstrates how these issues can be mitigated when having access to multi-source data.

Hemodynamic Bedside Monitoring Instrument with Pressure and Optical Sensors: Validation and Modality Comparison.

Results from two independent clinical validation studies for measuring hemodynamics at the patient's bedside using a compact finger probe are reported. Technology comprises a barometric pressure sensor, and in one implementation, additionally, an optical sensor for photoplethysmography (PPG) is developed, which can be used to measure blood pressure and analyze rhythm, including the continuous detection of atrial fibrillation. The capabilities of the technology are shown in several form factors, including a miniaturized version resembling a common pulse oximeter to which the technology could be integrated in. Several main results are presented: i) the miniature finger probe meets the accuracy requirements of non-invasive blood pressure instrument validation standard, ii) atrial fibrillation can be detected during the blood pressure measurement and in a continuous recording, iii) a unique comparison between optical and pressure sensing mechanisms is provided, which shows that the origin of both modalities can be explained using a pressure-volume model and that recordings are close to identical between the sensors. The benefits and limitations of both modalities in hemodynamic monitoring are further discussed.

Non-Invasive Hemodynamic Monitoring System Integrating Spectrometry, Photoplethysmography, and Arterial Pressure Measurement Capabilities.

Minimally invasive and non-invasive hemodynamic monitoring technologies have recently gained more attention, driven by technological advances and the inherent risk of complications in invasive techniques. In this article, an experimental non-invasive system is presented that effectively combines the capabilities of spectrometry, photoplethysmography (PPG), and arterial pressure measurement. Both time- and wavelength-resolved optical signals from the fingertip are measured under external pressure, which gradually increased above the level of systolic blood pressure. The optical channels measured at 434-731 nm divided into three groups separated by a group of channels with wavelengths approximately between 590 and 630 nm. This group of channels, labeled transition band, is characterized by abrupt changes resulting from a decrease in the absorption coefficient of whole blood. External pressure levels of maximum pulsation showed that shorter wavelengths (<590 nm) probe superficial low-pressure blood vessels, whereas longer wavelengths (>630 nm) probe high-pressure arteries. The results on perfusion indices and DC component level changes showed clear differences between the optical channels, further highlighting the importance of wavelength selection in optical hemodynamic monitoring systems. Altogether, the results demonstrated that the integrated system presented has the potential to extract new hemodynamic information simultaneously from macrocirculation to microcirculation.

Smartphone-Based Recognition of Heart Failure by Means of Microelectromechanical Sensors.

BACKGROUND: Heart failure (HF) is the leading cause of hospitalization in individuals over 65 years of age. Identifying noninvasive methods to detect HF may address the epidemic of HF. Seismocardiography which measures cardiac vibrations transmitted to the chest wall has recently emerged as a promising technology to detect HF. OBJECTIVES: In this multicenter study, the authors examined whether seismocardiography using commercially available smartphones can differentiate control subjects from patients with stage C HF. METHODS: Both inpatients and outpatients with HF were enrolled from Finland and the United States. Inpatients with HF were assessed within 2 days of admission, and outpatients were assessed in the ambulatory setting. In a prespecified pooled data analysis, algorithms were derived using logistic regression and then validated using a bootstrap aggregation method. RESULTS: A total of 217 participants with HF (174 inpatients and 172 outpatients) and 786 control subjects from cardiovascular clinics were enrolled. The mean age of participants with acute HF was 64 ± 13 years, 64.9% were male, left ventricular ejection fraction was 39% ± 15%, and median N-terminal pro-B-type natriuretic peptide was 5,778 ng/L (Q1-Q3: 1,933-6,703). The majority (74%) of participants with HF had reduced EF, and 38% had atrial fibrillation. Across both HF cohorts, the algorithms had an area under the receiver operating characteristic curve of 0.95 with a sensitivity of 85%, specificity of 90%, and accuracy of 89% for the detection of HF, with a decision threshold of 0.5. The positive and negative likelihood ratios were 8.50 and 0.17, respectively. The accuracy of the algorithms was not significantly different in subgroups based on age, sex, body mass index, and atrial fibrillation. CONCLUSIONS: Smartphone-based assessment of cardiac function using seismocardiography is feasible and differentiates patients with HF from control subjects with high diagnostic accuracy. (Recognition of Heart Failure With Micro Electro-mechanical Sensors FI; NCT04444583; Recognition of Heart Failure With Micro Electro-mechanical Sensors [NCT04378179]; Detection of Coronary Artery Disease With Micro Electro-mechanical Sensors; NCT04290091).

Parallel, Continuous Monitoring and Quantification of Programmed Cell Death in Plant Tissue.

Accurate quantification of hypersensitive response (HR) programmed cell death is imperative for understanding plant defense mechanisms and developing disease-resistant crop varieties. Here, a phenotyping platform for rapid, continuous-time, and quantitative assessment of HR is demonstrated: Parallel Automated Spectroscopy Tool for Electrolyte Leakage (PASTEL). Compared to traditional HR assays, PASTEL significantly improves temporal resolution and has high sensitivity, facilitating detection of microscopic levels of cell death. Validation is performed by transiently expressing the effector protein AVRblb2 in transgenic Nicotiana benthamiana (expressing the corresponding resistance protein Rpi-blb2) to reliably induce HR. Detection of cell death is achieved at microscopic intensities, where leaf tissue appears healthy to the naked eye one week after infiltration. PASTEL produces large amounts of frequency domain impedance data captured continuously. This data is used to develop supervised machine-learning (ML) models for classification of HR. Input data (inclusive of the entire tested concentration range) is classified as HR-positive or negative with 84.1% mean accuracy (F1 score = 0.75) at 1 h and with 87.8% mean accuracy (F1 score = 0.81) at 22 h. With PASTEL and the ML models produced in this work, it is possible to phenotype disease resistance in plants in hours instead of days to weeks.

Method for measuring jugular venous pulse with a miniature gyroscope sensor patch.

The right internal jugular vein is connected to the right atrium of the heart via the superior vena cava, and consequently its pressure, known as the jugular venous pressure or the jugular venous pulse (JVP), is an important indicator of cardiac function. The JVP can be estimated visually from the neck but it is rather difficult and imprecise. In this article we propose a method to measure the JVP using a motion sensor patch attached to the neck. The JVP signal was extracted from the sensor's 3-axes gyroscope signal and aligned with simultaneously measured ECG and seismocardiogram signals.The method was tested on 20 healthy subjects. The timings of the characteristic JVP waves were compared with the ECG R peaks and seismocardiogram heart sounds S1 and S2. The JVP was reliably measured from 18 subjects with all three waves identified. The timings of the waves were also physiologically plausible when compared to the ECG R peak and the heart sounds. Importantly, the JVP was also found to modulate with respiration, further indicating that the measured signal was indeed the JVP and not the carotid pulse.The results show that the JVP can be measured with a wearable patch-like device registering the delicate motions of the right internal jugular vein. The method has potential to be developed into a clinical tool to measure cardiac health in diseases such as heart failure and chronic obstructive pulmonary disease (COPD).Clinical relevance-The developed method could enable an affordable measurement of clinically important cardiac parameter, jugular venous pulse, as a part of a routine examination.

Development and clinical validation of a miniaturized finger probe for bedside hemodynamic monitoring.

Our aim is to develop a blood pressure (BP) measurement technology that could be integrated into a finger-worn pulse oximeter, eliminating the need for a brachial cuff. We present a miniature cuffless tonometric finger probe system that uses the oscillometric method to measure BP. Our approach uses a motorized press that is used to apply pressure to the fingertip to measure BP. We verified the functionality of the device in a clinical trial (n = 43) resulting in systolic and diastolic pressures ((mean ± SD) mmHg) of (-3.5 ± 8.4) mmHg and (-4.0 ± 4.4) mmHg, respectively. Comparison was made with manual auscultation (n = 26) and automated cuff oscillometry (n = 18). In addition to BP, we demonstrated the ability of the device to assess arterial stiffness (n = 18) and detect atrial fibrillation (n = 6). We were able to introduce a sufficiently small device that could be used for convenient ambulatory measurements with minimal discomfort.

Domain randomization using synthetic electrocardiograms for training neural networks.

We present a method for training neural networks with synthetic electrocardiograms that mimic signals produced by a wearable single lead electrocardiogram monitor. We use domain randomization where the synthetic signal properties such as the waveform shape, RR-intervals and noise are varied for every training example. Models trained with synthetic data are compared to their counterparts trained with real data. Detection of r-waves in electrocardiograms recorded during different physical activities and in atrial fibrillation is used to assess the performance. By allowing the randomization of the synthetic signals to increase beyond what is typically observed in the real-world data the performance is on par or superseding the performance of networks trained with real data. Experiments show robust model performance using different seeds and on different unseen test sets that were fully separated from the training phase. The ability of the model to generalize well to hidden test sets without any specific tuning provides a simple and explainable alternative to more complex adversarial domain adaptation methods for model generalization. This method opens up the possibility of extending the use of synthetic data towards domain insensitive cardiac disease classification when disease specific a priori information is used in the electrocardiogram generation. Additionally, the method provides training with free-to-collect data with accurate labels, control of the data distribution eliminating class imbalances that are typically observed in health-related data, and the generated data is inherently private.

Advances in Non-Invasive Blood Pressure Measurement Techniques.

Hypertension, or elevated blood pressure (BP), is a marker for many cardiovascular diseases and can lead to life threatening conditions such as heart failure, coronary artery disease and stroke. Several techniques have recently been proposed and investigated for non-invasive BP monitoring. The increasing desire for telemonitoring solutions that allow patients to manage their own conditions from home has accelerated the development of new BP monitoring techniques. In this review, we present the recent progress in non-invasive blood pressure monitoring solutions emphasizing clinical validation and trade-offs between available techniques. We introduce the current BP measurement techniques with their underlying operating principles. New promising proof-of-concept studies are presented and recent modeling and machine learning approaches for improved BP estimation are summarized. This aids discussions on how new BP monitors should evaluated in order to bring forth new home monitoring solutions in wearable form factor. Finally, we discuss on unresolved challenges in making convenient, reliable and validated BP monitoring solutions.

Miniaturization of a Finger-Worn Blood Pressure Instrument.

Blood pressure monitoring using a traditional arm cuff device is often inconvenient and possibly painful. We present a miniature cuffless tonometric finger probe system, that uses the oscillometric method to measure blood pressure (BP). A small enough device could be used for convenient ambulatory measurement and be worn during sleep with minimal discomfort. In addition to BP, the device is able to collect arterial pulse wave data that can further be used to derive other cardiovascular parameters, such as heart rate (HR), heart rate variability (HRV) and central aortic systolic pressure (CASP). The device uses a motor controlled press that is used to apply pressure to the finger tip to measure the oscillometric response. We verified the functionality of the device by proof-of-concept measurements. Lastly we evaluate methods for further developing the concept and discuss the future directions.

Sleep during menopausal transition: a 10-year follow-up.

STUDY OBJECTIVES: A 10-year observational follow-up study to evaluate the changes in sleep architecture during the menopausal transition. METHODS: Fifty-seven premenopausal women (mean age 46 years, SD 0.9) were studied at baseline and after a 10-year follow-up. At both time points, polysomnography (PSG) was performed, and the serum follicle-stimulating hormone (S-FSH) concentration was measured. Linear regression models were used to study the effects of aging and menopause (assessed as change in S-FSH) on sleep. RESULTS: After controlling for body mass index, vasomotor, and depressive symptoms, higher S-FSH level was associated with longer sleep latency (B 0.45, 95% confidence interval [CI]: 0.07 to 0.83). Aging of 10 years was associated with shorter sleep latency (B -46.8, 95% CI: -77.2 to -16.4), shorter latency to stage 2 sleep (B -50.6, 95% CI: -85.3 to -15.9), decreased stage 2 sleep (B -12.4, 95% CI: -21.4 to -3.4), and increased slow-wave sleep (B 12.8, 95% CI: 2.32 to 23.3) after controlling for confounding factors. CONCLUSIONS: This study suggests that PSG measured sleep of middle-aged women does not worsen over a 10-year time span due to the menopausal transition. The observed changes seem to be rather age- than menopause-dependent.

Disposable silicon-based all-in-one micro-qPCR for rapid on-site detection of pathogens.

Rapid screening and low-cost diagnosis play a crucial role in choosing the correct course of intervention when dealing with highly infectious pathogens. This is especially important if the disease-causing agent has no effective treatment, such as the novel coronavirus SARS-CoV-2, and shows no or similar symptoms to other common infections. Here, we report a disposable silicon-based integrated Point-of-Need transducer (TriSilix) for real-time quantitative detection of pathogen-specific sequences of nucleic acids. TriSilix can be produced at wafer-scale in a standard laboratory (37 chips of 10 × 10 × 0.65 mm in size can be produced in 7 h, costing ~0.35 USD per device). We are able to quantitatively detect a 563 bp fragment of genomic DNA of Mycobacterium avium subspecies paratuberculosis through real-time PCR with a limit-of-detection of 20 fg, equivalent to a single bacterium, at the 35th cycle. Using TriSilix, we also detect the cDNA from SARS-CoV-2 (1 pg) with high specificity against SARS-CoV (2003).

Stretchable Composite Acoustic Transducer for Wearable Monitoring of Vital Signs.

A highly flexible, stretchable, and mechanically robust low-cost soft composite consisting of silicone polymers and water (or hydrogels) is reported. When combined with conventional acoustic transducers, the materials reported enable high performance real-time monitoring of heart and respiratory patterns over layers of clothing (or furry skin of animals) without the need for direct contact with the skin. The approach enables an entirely new method of fabrication that involves encapsulation of water and hydrogels with silicones and exploits the ability of sound waves to travel through the body. The system proposed outperforms commercial, metal-based stethoscopes for the auscultation of the heart when worn over clothing and is less susceptible to motion artefacts. The system both with human and furry animal subjects (i.e., dogs), primarily focusing on monitoring the heart, is tested; however, initial results on monitoring breathing are also presented. This work is especially important because it is the first demonstration of a stretchable sensor that is suitable for use with furry animals and does not require shaving of the animal for data acquisition.

An instrument for measuring blood pressure and assessing cardiovascular health from the fingertip.

Despite blood pressure being one the leading modifiable risk factors for cardiovascular disease and death, it is severely under-monitored. For this challenge we propose a finger artery non-invasive tono-oscillometric monitor (FANTOM) which is an automated low-cost instrument for measuring blood pressure and hemodynamic parameters from the fingertip. The sensing technology is highly scalable and could be integrated to a pulse oximeter probe for increased patient comfort. A tonometric cuff-less mechatronic system is used to apply pressure on the fingertip for (i) measuring oscillometric blood pressure, (ii) recording arterial waveform and for (iii) constructing central blood pressure (CBP) waveform. Clinical study on volunteers (n = 33) was performed against a commercially available arm cuff device yielding systolic and diastolic readings ((mean±SD) mmHg) of (-0.9 ± 7.3) mmHg and (-3.3 ± 6.6) mmHg respectively. The results comply with the Association for the Advancement of Medical Instrumentation (AAMI) standard for non-invasive blood pressure monitors. The arterial pulse recording morphology was compared against a volume clamp device (CNSystems CNAP 500) (n = 3) resulting in similar performance. Comparison of CBP against a pulse wave analysis (PWA) device (Atcor Medical Sphygmocor XCEL) (n = 5) revealed central aortic systolic pulse (CASP) and central augmentation index (cAIx) estimates with precision and accuracy of (2.0 ± 3.7) mmHg and (1.4 ± 6.2)% respectively. In conclusion, the results indicate that the proposed technology could be useful in the development of new portable or wearable blood pressure monitors. The sensing technology is highly scalable and could be integrated to a pulse oximeter probe for increased patient comfort.

Clinical assessment of a non-invasive wearable MEMS pressure sensor array for monitoring of arterial pulse waveform, heart rate and detection of atrial fibrillation.

There is an unmet clinical need for a low cost and easy to use wearable devices for continuous cardiovascular health monitoring. A flexible and wearable wristband, based on microelectromechanical sensor (MEMS) elements array was developed to support this need. The performance of the device in cardiovascular monitoring was investigated by (i) comparing the arterial pressure waveform recordings to the gold standard, invasive catheter recording (n = 18), (ii) analyzing the ability to detect irregularities of the rhythm (n = 7), and (iii) measuring the heartrate monitoring accuracy (n = 31). Arterial waveforms carry important physiological information and the comparison study revealed that the recordings made with the wearable device and with the gold standard device resulted in almost identical (r = 0.9-0.99) pulse waveforms. The device can measure the heart rhythm and possible irregularities in it. A clustering analysis demonstrates a perfect classification accuracy between atrial fibrillation (AF) and sinus rhythm. The heartrate monitoring study showed near perfect beat-to-beat accuracy (sensitivity = 99.1%, precision = 100%) on healthy subjects. In contrast, beat-to-beat detection from coronary artery disease patients was challenging, but the averaged heartrate was extracted successfully (95% CI: -1.2 to 1.1 bpm). In conclusion, the results indicate that the device could be useful in remote monitoring of cardiovascular diseases and personalized medicine.

Cardiac monitoring of dogs via smartphone mechanocardiography: a feasibility study.

BACKGROUND: In the context of monitoring dogs, usually, accelerometers have been used to measure the dog's movement activity. Here, we study another application of the accelerometers (and gyroscopes)-seismocardiography (SCG) and gyrocardiography (GCG)-to monitor the dog's heart. Together, 3-axis SCG and 3-axis GCG constitute of 6-axis mechanocardiography (MCG), which is inbuilt to most modern smartphones. Thus, the objective of this study is to assess the feasibility of using a smartphone-only solution to studying dog's heart. METHODS: A clinical trial (CT) was conducted at the University Small Animal Hospital, University of Helsinki, Finland. 14 dogs (3 breeds) including 18 measurements (about one half of all) where the dog's status was such that it was still and not panting were further selected for the heart rate (HR) analysis (each signal with a duration of 1 min). The measurement device in the CT was a custom Holter monitor including synchronized 6-axis MCG and ECG. In addition, 16 dogs (9 breeds, one mixed-breed) were measured at home settings by the dog owners themselves using Sony Xperia Android smartphone sensor to further validate the applicability of the method. RESULTS: The developed algorithm was able to select 10 good-quality signals from the 18 CT measurements, and for 7 of these, the automated algorithm was able to detect HR with deviation below or equal to 5 bpm (compared to ECG). Further visual analysis verified that, for approximately half of the dogs, the signal quality at home environment was sufficient for HR extraction at least in some signal locations, while the motion artifacts due to dog's movements are the main challenges of the method. CONCLUSION: With improved data analysis techniques for managing noisy measurements, the proposed approach could be useful in home use. The advantage of the method is that it can operate as a stand-alone application without requiring any extra equipment (such as smart collar or ECG patch).

Real-time wash-free detection of unlabeled PNA-DNA hybridization using discrete FET sensor.

We demonstrate an electrochemical sensor for detection of unlabeled single-stranded DNA using peptide nucleic acid (PNA) probes coupled to the field-effect transistor (FET) gate. The label-free detection relies on the intrinsic charge of the DNA backbone. Similar detection schemes have mainly concentrated on sensitivity improvement with an emphasis on new sensor structures. Our approach focuses on using an extended-gate that separates the FET and the sensing electrode yielding a simple and mass fabricable device. We used PNA probes for efficient hybridization in low salt conditions that is required to avoid the counter ion screening. As a result, significant part of the target DNA lies within the screening length of the sensor. With this, we achieved a wash-free detection where  typical gate potential shifts are more than 70 mV with 1 µM target DNA. We routinely obtained a real-time, label- and wash-free specific detection of target DNA in nanomolar concentration with low-cost electronics and the responses were achieved within minutes after introducing targets to the solution. Furthermore, the results suggest that the sensor performance is limited by specificity rather than by sensitivity and using low-cost electronics does not limit the sensor performance in the presented sensor configuration.

Detection principles of biological and chemical FET sensors.

The seminal importance of detecting ions and molecules for point-of-care tests has driven the search for more sensitive, specific, and robust sensors. Electronic detection holds promise for future miniaturized in-situ applications and can be integrated into existing electronic manufacturing processes and technology. The resulting small devices will be inherently well suited for multiplexed and parallel detection. In this review, different field-effect transistor (FET) structures and detection principles are discussed, including label-free and indirect detection mechanisms. The fundamental detection principle governing every potentiometric sensor is introduced, and different state-of-the-art FET sensor structures are reviewed. This is followed by an analysis of electrolyte interfaces and their influence on sensor operation. Finally, the fundamentals of different detection mechanisms are reviewed and some detection schemes are discussed. In the conclusion, current commercial efforts are briefly considered.