Scalability of cardiovascular intrinsic frequencies: Validations in preclinical models and non-invasive clinical studies

A machine learning approach for computation of cardiovascular intrinsic frequencies

Analysis of cardiovascular waveforms provides valuable clinical information about the state of health and disease. The intrinsic frequency (IF) method is a recently introduced framework that uses a single arterial pressure waveform to extract physiologically relevant information about the cardiovascular system. The clinical usefulness and physiological accuracy of the IF method have been well-established via several preclinical and clinical studies. However, the computational complexity of the current L2 optimization solver for IF calculations remains a bottleneck for practical deployment of the IF method in real-time settings. In this paper, we propose a machine learning (ML)-based methodology for determination of IF parameters from a single carotid waveform. We use a sequentially-reduced Feedforward Neural Network (FNN) model for mapping carotid waveforms to the output parameters of the IF method, thereby avoiding the non-convex L2 minimization problem arising from the conventional IF approach. Our methodology also includes procedures for data pre-processing, model training, and model evaluation. In our model development, we used both clinical and synthetic waveforms. Our clinical database is composed of carotid waveforms from two different sources: the Huntington Medical Research Institutes (HMRI) iPhone Heart Study and the Framingham Heart Study (FHS). In the HMRI and FHS clinical studies, various device platforms such as piezoelectric tonometry, optical tonometry (Vivio), and an iPhone camera were used to measure arterial waveforms. Our blind clinical test shows very strong correlations between IF parameters computed from the FNN-based method and those computed from the standard L2 optimization-based method (i.e., R≥0.93 and P-value ≤0.005 for each IF parameter). Our results also demonstrate that the performance of the FNN-based IF model introduced in this work is independent of measurement apparatus and of device sampling rate.

Assessment of Aortic Characteristic Impedance and Arterial Compliance from Non-invasive Carotid Pressure Waveform in The Framingham Heart Study

The primary goal of this study was to test the hypothesis that a hybrid intrinsic frequency-machine learning (IF-ML) approach can accurately evaluate total arterial compliance (TAC) and aortic characteristic impedance (Zao) from a single noninvasive carotid pressure waveform in both women and men with heart failure (HF). TAC and Zao are cardiovascular biomarkers with established clinical significance. TAC is lower and Zao is higher in women than in men, so women are more susceptible to the consequent deleterious effects of them. Although the principles of TAC and Zao are pertinent to a multitude of cardiovascular diseases, including HF, their routine clinical use is limited because of the requirement for simultaneous measurements of flow and pressure waveforms. For this study, the data were obtained from the Framingham Heart Study (n = 6,201, 53% women). The reference values of Zao and TAC were computed from carotid pressure and aortic flow waveforms. IF parameters of carotid pressure waveform were used in ML models. IF models were developed on n = 5,168 of randomly selected data and blindly tested the remaining data (n = 1,033). The final models were evaluated in patients with HF. Correlations between IF-ML and reference values in all HF and HF with preserved ejection fraction for TAC were 0.88 and 0.90, and for Zao were 0.82 and 0.80, respectively. The classification accuracy in all HF and HF with preserved ejection fraction for TAC were 0.9 and 0.93, and for Zao were 0.81 and 0.89, respectively. In conclusion, the IF-ML method provides an accurate estimation of TAC and Zao in all subjects with HF and in the general population.

Instantaneous detection of acute myocardial infarction and ischaemia from a single carotid pressure waveform in rats

Aims

Myocardial infarction (MI) is one of the leading causes of death worldwide. It is well accepted that early diagnosis followed by early reperfusion therapy significantly increases the MI survival. Diagnosis of acute MI is traditionally based on the presence of chest pain and electrocardiogram (ECG) criteria. However, around 50% of the MIs are without chest pain, and ECG is neither completely specific nor definitive. Therefore, there is an unmet need for methods that allow detection of acute MI or ischaemia without using ECG. Our hypothesis is that a hybrid physics-based machine learning (ML) method can detect the occurrence of acute MI or ischaemia from a single carotid pressure waveform.

Methods and results

We used a standard occlusion/reperfusion rat model. Physics-based ML classifiers were developed using intrinsic frequency parameters extracted from carotid pressure waveforms. ML models were trained, validated, and generalized using data from 32 rats. The final ML models were tested on an external stratified blind dataset from additional 13 rats. When tested on blind data, the best ML model showed specificity = 0.92 and sensitivity = 0.92 for detecting acute MI. The best model's specificity and sensitivity for ischaemia detection were 0.85 and 0.92, respectively.

Conclusion

We demonstrated that a hybrid physics-based ML approach can detect the occurrence of acute MI and ischaemia from carotid pressure waveform in rats. Since carotid pressure waveforms can be measured non-invasively, this proof-of-concept pre-clinical study can potentially be expanded in future studies for non-invasive detection of MI or myocardial ischaemia.

Screening left ventricular systolic dysfunction in children using intrinsic frequencies of carotid pressure waveforms measured by a novel smartphone-based device

Objective.Children with heart failure have higher rates of emergency department utilization, health care expenditure, and hospitalization. Therefore, a need exists for a simple, non-invasive, and inexpensive method of screening for left ventricular (LV) dysfunction. We recently demonstrated the practicality and reliability of a wireless smartphone-based handheld device in capturing carotid pressure waveforms and deriving cardiovascular intrinsic frequencies (IFs) in children with normal LV function. Our goal in this study was to demonstrate that an IF-based machine learning method (IF-ML) applied to noninvasive carotid pressure waveforms can distinguish between normal and abnormal LV ejection fraction (LVEF) in pediatric patients.Approach. Fifty patients ages 0 to 21 years underwent LVEF measurement by echocardiogram or cardiac magnetic resonance imaging. On the same day, patients had carotid waveforms recorded using Vivio. The exclusion criterion was known vascular disease that would interfere with obtaining a carotid artery pulse. We adopted a hybrid IF- Machine Learning (IF-ML) method by applying physiologically relevant IF parameters as inputs to Decision Tree classifiers. The threshold for low LVEF was chosen as <50%.Main results.The proposed IF-ML method was able to detect an abnormal LVEF with an accuracy of 92% (sensitivity = 100%, specificity = 89%, area under the curve (AUC) = 0.95). Consistent with previous clinical studies, the IF parameterω1was elevated among patients with reduced LVEF.Significance.A hybrid IF-ML method applied on a carotid waveform recorded by a hand-held smartphone-based device can differentiate between normal and abnormal LV systolic function in children with normal cardiac anatomy.

A coupled atrioventricular-aortic setup for in-vitro hemodynamic study of the systemic circulation: Design, fabrication, and physiological relevancy

In-vitro models of the systemic circulation have gained a lot of interest for fundamental understanding of cardiovascular dynamics and for applied hemodynamic research. In this study, we introduce a physiologically accurate in-vitro hydraulic setup that models the hemodynamics of the coupled atrioventricular-aortic system. This unique experimental simulator has three major components: 1) an arterial system consisting of a human-scale artificial aorta along with the main branches, 2) an artificial left ventricle (LV) sac connected to a programmable piston-in-cylinder pump for simulating cardiac contraction and relaxation, and 3) an artificial left atrium (LA). The setup is designed in such a way that the basal LV is directly connected to the aortic root via an aortic valve, and to the LA via an artificial mitral valve. As a result, two-way hemodynamic couplings can be achieved for studying the effects that the LV, aorta, and LA have on each other. The collected pressure and flow measurements from this setup demonstrate a remarkable correspondence to clinical hemodynamics. We also investigate the physiological relevancies of isolated effects on cardiovascular hemodynamics of various major global parameters found in the circulatory system, including LV contractility, LV preload, heart rate, aortic compliance, and peripheral resistance. Subsequent control over such parameters ultimately captures physiological hemodynamic effects of LV systolic dysfunction, preload (cardiac) diseases, and afterload (arterial) diseases. The detailed design and fabrication of the proposed setup is also provided.

Effects of vessel wall mechanics on non-invasive evaluation of cardiovascular intrinsic frequencies

Intrinsic Frequency (IF) is a systems-based approach that provides valuable information for hemodynamic monitoring of the left ventricle (LV), the arterial system, and their coupling. Recent clinical studies have demonstrated the clinical significance of this method for prognosis and diagnosis of cardiovascular diseases. In IF analysis, two dominant instantaneous frequencies (ω₁ and ω₂) are extracted from arterial pressure waveforms. The value of ω₁ is related to the dynamics of the LV and the value of ω₂ is related to the dynamics of vascular function. This work investigates the effects of vessel wall mechanics on the accuracy and applicability of IFs extracted from vessel wall displacement waveforms compared to IFs extracted from pressure waveforms. In this study, we used a computational approach employing a fluid-structure interaction finite element method for various wall mechanics governed by linearly elastic, hyperelastic, and viscoelastic models. Results show that for vessels with elastic wall behavior, the error between displacement-based and pressure-based IFs is negligible. In the presence of stenosis or aneurysm in elastic arteries, the maximum errors associated with displacement-based IFs is less than 2%. For non-linear elastic and viscoelastic arteries, errors are more pronounced (where the former reaches up to 11% and the latter up to 27%). Our results ultimately suggest that displacement-based computations of ω₁ and ω₂ are accurate in vessels that exhibit elastic behavior (such as carotid arteries) and are suitable surrogates for pressure-based IFs. This is clinically significant because displacement-based IFs can be measured non-invasively.