Arterial Pressure Waveform Research Articles

Robust modelling of arterial blood pressure reconstruction from photoplethysmography

Blood pressure is a crucial indicator of cardiovascular disease, and arterial blood pressure (ABP) waveforms contain information that reflects the cardiovascular status. We propose a novel deep-learning method that converts photoplethysmogram (PPG) signals into ABP waveforms. We used -Net as a feature extractor and designed a Bi-block to capture individualised time information in encoder feature extraction. We further enhanced the prediction accuracy of the ABP waveforms by applying a combined loss function to each layer of deep supervision. We also propose a total error index (TEI) to measure overall performance. Furthermore, we extended our method from the UCI dataset to the VitalDB dataset, achieving mean absolute error ± standard deviation (MAE ± STD) values of 2.48 ± 1.95, 1.42 ± 1.42, and 1.48 ± 1.36 mmHg for systolic blood pressure (SBP), diastolic blood pressure (DBP), and mean arterial pressure (MAP) in UCI dataset, and 2.16 ± 1.53, 1.12 ± 0.59, and 1.35 ± 0.84 mmHg in VitalDB dataset, respectively. The mean ± STD values of the TEI index are 0.29 ± 0.10 in UCI dataset and 0.29 ± 0.15 in VitalDB dataset. These results demonstrate the superiority of the proposed method over existing methods and its robustness to different sampling frequencies and devices.

Evaluation of central hemodynamics using the radial arterial blood pressure waveform in children, adolescents, and young adults

Evaluation of central hemodynamics using the radial arterial blood pressure waveform in children, adolescents, and young adults

Deciphering postoperative hemodynamics in pediatric intensive care: insights from patient-specific computational fluid dynamics models for congenital heart disease

Abstract Purpose Precise hemodynamic control is imperative for optimizing outcomes in children undergoing cardiac surgery, particularly in the pediatric intensive care unit (PICU). Computational fluid dynamic (CFD) models tailored to individual pediatric patients following definitive repairs for congenital heart disease hold the potential to discern and analyze the intricacies of their hemodynamic systems. This study aims to explore the applicability of patient-specific CFD models in determining optimal treatment strategies to enhance patient hemodynamics. Methods Ten postoperative pediatric patients (6 females and 4 males, median age 4.5±3.6 months, median height 62.5±5.3 cm, and median weight 6.2±1.3 kg) diagnosed with ventricular septal defect (VSD, n=5), atrial septal defect (ASD, n=2), and atrioventricular septal defect (AVSD, n=3) were enrolled. Continuous radial arterial waveforms and sporadic arterial blood draws, including lactate levels, were recorded in the PICU postoperatively. Fluid dynamics equations were employed to model major and peripheral blood vessels, utilizing a 0–1-dimensional hemodynamic model validated in previous studies and refined for this investigation. Sensitivity analysis based on correlation coefficients identified significant parameters, adjusted for replicating pressure waveforms. Comparison and re-approximation was made with actual radial artery pressure waveforms from 78 arterial waveforms and blood draws. Updated parameters were then correlated with postoperative time and lactate levels in the ICU. Results Replicated waveforms demonstrated high agreement with original arterial waveforms, with mean, maximal, and minimal blood pressure percentages of 96.7±1.8%, 96.7±2.4%, and 95.8±2.4%, respectively. Notably, systemic vascular resistance and compliance exhibited a substantial relationship with time elapsed after admission to the PICU, with correlation coefficients (R) of 0.26 and 0.43, respectively. Lactate levels demonstrated a decrease with reduced peripheral vascular resistance and an increase in vascular compliance, with R of 0.23 and 0.32, respectively, aligning with expected clinical improvement and stability. Conclusion This study successfully attained precise replication of radial artery blood pressures and intricate pressure waveforms using patient-specific CFD models in pediatric patients following congenital heart disease repairs. The insights gained shed light on the pivotal hemodynamic parameters that impact both stability and instability. This research sets the stage for theoretically informed treatments by amalgamating real-time biological information with patient-specific CFD models.Re-approximation of arterial waveformCorrelation of hemodynamic parameters

Invasive validation of central blood pressure measurements using a suprasystolic monitor

Abstract Interest in the assessment of Central Blood Pressure (cBP) is steadily growing since it is a better predictor of myocardial infarction, stroke, or death than Peripheral Blood Pressure (pBP). However, pBP remains the first line for the diagnosis of hypertension (HTN) due to the ease and speed of the measurement. We sought to validate a new method to estimate cBP from brachial sphygmomanometer measurements. We estimated cBP using a sphygmomanometer that can record, after conventional pBP measurements, the pressure in the cuff maintained at a constant suprasystolic pressure. cBP is calculated using a method based on applying a physics-based model of the left subclavian artery to the brachial arterial branch to the suprasystolic brachial artery pressure waveforms. We compared this noninvasive estimated cBP to an invasive gold-standard measurement during cardiac catheterization. After clinically indicated cardiac catheterization requiring a haemodynamic assessment of a moderate coronary artery stenosis by Fractional Flow Reserve (FFR), the wire that incorporates a high fidelity electronic pressure transducer was positioned into the guiding catheter left in the ascending aorta. Simultaneous measurements were performed with a cuff placed on the patient's left arm. The output signals were recorded in a custom-based acquisition system. The comparison was obtained on the same 10 seconds recorded by the cuff and the pressure wire. The goal will be to reach 85 patients as recommended by the ARTERY Society Task Force. It recommends as an accuracy criterion to achieve a mean difference ≤ 5 mmHg with a standard deviation (SD) ≤ 8 mmHg, which are the standards adopted for this study. Patients with aortic stenosis, atrial fibrillation, or frequent (supra)ventricular extrasystoles were excluded. A total of 197 pairs of measured-estimated cBP were analysed in 78 patients with a mean age of 63 ± 8 yo, 75% of whom were men, 70% had hypertension, 26% fdiabetes, 93% dyslipidemia, 41% were tobacco users and 14% suffered from clinical peripheral artery diseases without history of revascularization. The median (InterQuartiles 25-75%) pBP was 123 [112 134] / 70 [64 77] mmHg. Measured cBP was 102 [92 11] / 78 [70 84] mmHg with a central augmentation index of 34 [23 44] %. The mean difference between estimated and measured systolic cBP was 4.9 ± 6.5 mmHg, while it was -3.9 ± 5.8 mmHg for diastolic cBP, meeting the prespecified accuracy criteria. In conclusion, our approach based on an exact matching and comparison of non-invasive suprasystolic arm cuff measurements and high-fidelity aortic BP validates that an estimation of central aortic blood pressure that can be easily carried out in daily practice and might improve HTN management. It remains to be demonstrated that more cardiovascular complications will be avoided by guiding the antihypertensive therapy with cBP instead of pBP measurements.Bland-Altman of estimated - measured cBP

Occurrence of Low Cardiac Index During Normotensive Periods in Cardiac Surgery: A Prospective Cohort Study Using Continuous Noninvasive Cardiac Output Monitoring.

Continuous cardiac output monitoring is not standard practice during cardiac surgery, even though patients are at substantial risk for systemic hypoperfusion. Thus, the frequency of low cardiac output during cardiac surgery is unknown. We conducted a prospective cohort study at a tertiary medical center from July 2021 to November 2023. Eligible patients were ≥18 undergoing isolated coronary bypass (CAB) surgery with the use of cardiopulmonary bypass (CPB). Cardiac output indexed to body surface area (CI) was continuously recorded at 5-second intervals throughout surgery using a US Food and Drug Administration (FDA)-approved noninvasive monitor from the arterial blood pressure waveform. Mean arterial blood pressure (MAP) and central venous pressure (CVP) were also analyzed. Low CI was defined as <2 L/min/m2 and low MAP as <65 mm Hg. We calculated time with low CI for each patient for the entire surgery, pre-CPB and post-CPB periods, and the proportion of time with low CI and normal MAP. We used Pearson correlation to evaluate the relationship between CI and MAP and paired Wilcoxon rank sum tests to assess the difference in correlations of CI with MAP before and after CPB. In total, 101 patients were analyzed (age [standard deviation, SD] 64.8 [9.8] years, 25% female). Total intraoperative time (mean [SD]) with low CI was 86.4 (62) minutes, with 61.2 (42) minutes of low CI pre-CPB and 25.2 (31) minutes post-CPB. Total intraoperative time with low CI and normal MAP was 66.5 (56) minutes, representing mean (SD) 69% (23%) of the total time with low CI; 45.8 (38) minutes occurred pre-CPB and 20.6 (27) minutes occurred post-CPB. Overall, the correlation (mean [SD]) between CI and MAP was 0.33 (0.31), and the correlation was significantly higher pre-CPB (0.53 [0.32]) than post-CPB (0.29 [0.28], 95% confidence interval [CI] for difference [0.18-0.34], P < .001); however, there was substantial heterogeneity among participants in correlations of CI with MAP before and after CPB. Secondary analyses that accounted for CVP did not alter the correlation between CI and MAP. Exploratory analyses suggested duration of low CI (C <2 L/min/m2) was associated with increased risk of postoperative acute kidney injury (odds ratios [ORs] = 1.09; 95% CI; 1.01-1.13; P = .018). In a prospective cohort of patients undergoing CAB surgery, low CI was common even when blood pressure was normal. CI and MAP were correlated modestly. Correlation was higher before than after CPB with substantial heterogeneity among individuals. Future studies are needed to examine the independent relation of low CI to postoperative kidney injury and other adverse outcomes related to hypoperfusion.

Ultrasonic Doppler as a guide for feline peripheral arterial catheterization.

The study aimed to determine if an ultrasonic Doppler-guided technique (UDGT) leads to improved placement efficacy (time, success) of feline dorsal pedal arterial catheters vs the traditional palpation-guided technique (TPT). A total of 26 adult, client-owned cats requiring sedation or general anesthesia for any reason, aged >12 months and weighing >3.0 kg, and with Doppler blood pressure measurements of at least 80 mmHg were enrolled. Each hindlimb was randomly assigned for dorsal pedal arterial catheterization using either the UDGT or TPT. With the UDGT, the location of the artery was identified by an audible sound using the Doppler. Successful catheter placement was confirmed by visualization of an arterial pressure waveform using a transducer and monitor system attached to the catheter. The Kaplan-Meier method and log-rank test were used to compare the two techniques. The overall proportion of successful arterial catheterization was 17% (9/52): 19% (5/26) via UDGT and 15% (4/26) via TPT. Among successful arterial catheterizations (n = 9), the mean time to catheterization was 339 ± 198 s: 328 ± 237 s (n = 5) with UDGT and 353 ± 171 s (n = 4) with TPT. The log-rank test showed the two techniques were not significantly different in likelihood of successful arterial catheter placement or time to successful catheterization (P = 0.698). An arterial flash occurred in 62% (32/52) of the limbs, 58% (15/26) with the UDGT and 65% (17/26) with the TPT. Complications (self-limiting bruising, hematoma formation) were observed equally between UDGT (3/26 limbs) and TPT (3/26 limbs) in six cats. The UDGT did not improve the efficacy of catheter placement compared with the TPT. Few complications were associated with arterial catheterization.

A retrospective study of the effects of a vasopressor bolus on systolic slope (dP/dt) and dynamic arterial elastance (Eadyn)

BackgroundTo enhance the utility of functional hemodynamic monitoring, the variables systolic slope (dP/dt) and dynamic arterial elastance (Eadyn) are calculated by the Hypotension Prediction Index (HPI) Acumen® Software. This study was designed to characterize the effects of phenylephrine and ephedrine on dP/dt and Eadyn.MethodsThis was a retrospective, non-randomized analysis of data collected during two clinical studies. All patients required intra-operative controlled mechanical ventilation and had an indwelling radial artery catheter connected to an Acumen IQ sensor. Raw arterial pressure waveform data was downloaded from the patient monitor and all hemodynamic measurements were calculated off-line. The anesthetic record was reviewed for bolus administrations of either phenylephrine or ephedrine. Cardiovascular variables prior to drug administration were compared to those following vasopressor administrations. The primary outcome was the difference for dP/dt and Eadyn at baseline compared with the average after the bolus administration. All data sets demonstrated non-normal distributions so statistical analysis of paired and unpaired data followed the Wilcoxon matched pairs signed-rank test or Mann-Whitney U test, respectively.Results201 doses of phenylephrine and 100 doses of ephedrine were analyzed. All data sets are reported as median [95% CI]. Mean arterial pressure (MAP) increased from 62 [54,68] to 78 [76,80] mmHg following phenylephrine and from 59 [55,62] to 80 [77,83] mmHg following ephedrine. Stroke volume and cardiac output both increased. Stroke volume variation and pulse pressure variation decreased. Both drugs produced significant increases in dP/dt, from 571 [531, 645] to 767 [733, 811] mmHg/sec for phenylephrine and from 537 [509, 596] to 848 [779, 930] mmHg/sec for ephedrine. No significant changes in Eadyn were observed.ConclusionBolus administration of phenylephrine or ephedrine increases dP/dt but does not change Eadyn. dP/dt demonstrates potential for predicting the inotropic response to phenylephrine or ephedrine, providing guidance for the most efficacious vasopressor when treating hypotension.Trial registrationData was collected from two protocols. The first was deemed to not require written, informed consent by the Institutional Review Board (IRB). The second was IRB-approved (Effect of Diastolic Dysfunction on Dynamic Cardiac Monitors) and registered on ClinicalTrials.gov (NCT04177225).

Enhancing Hypotension Prediction in Real-time Patient Monitoring Through Deep Learning: A Novel Application of XResNet with Contrastive Learning and Value Attention Mechanisms.

The precise prediction of hypotension is vital for advancing preemptive patient care strategies. Traditional machine learning approaches, while instrumental in this field, are hampered by their dependence on structured historical data and manual feature extraction techniques. These methods often fall short of recognizing the intricate patterns present in physiological signals. Addressing this limitation, our study introduces an innovative application of deep learning technologies, utilizing a sophisticated end-to-end architecture grounded in XResNet. This architecture is further enhanced by the integration of contrastive learning and a value attention mechanism, specifically tailored to analyze arterial blood pressure (ABP) waveform signals. Our approach improves the performance of hypotension prediction over the existing state-of-theart ABP model [7]. This research represents a step towards optimizing patient care, embodying the next generation of AI-driven healthcare solutions. Through our findings, we demonstrate the promise of deep learning in overcoming the limitations of conventional prediction models, thereby offering an avenue for enhancing patient outcomes in clinical settings.

Characterization of intraoperative hemodynamic instability in patients undergoing general anesthesia

BackgroundIntraoperative hypotension (IOH) has been associated with increased postoperative morbidity and mortality, emphasizing the importance of maintaining hemodynamic stability during general anesthesia. The Hypotension Prediction Index (HPI) is a recently developed tool obtained from the arterial pressure waveform for predicting the likelihood of hypotension and reflecting the level of hemodynamic instability (HI).MethodsIn this retrospective observational cohort study, we analyzed anonymized data that were prospectively collected in patients having major abdominal surgery in six Spanish centers during 2022. We analyzed data from patients who underwent general anesthesia and required intraoperative monitoring with the Hemosphere monitoring system and HPI software. We defined five patterns of HI using a decision-making algorithm based on cardiovascular pathophysiology: absolute and relative hypovolemia, vasoplegia, and myocardial depression with and without vasoplegia.ResultsAfter analyzing 2,038 hours of hemodynamic monitoring from 393 patients, 1,946 episodes of HI were identified. Vasoplegia was the most common type of HI, accounting for 50.1% HI episodes. Absolute and relative hypovolemia were the next most frequent types, contributing 30.7% and 15.5% of the HI episodes.ConclusionsWe identified five distinct patterns of hemodynamic instability based on the evaluation of hemodynamic parameters underlying arterial hypotension during general anesthesia for patients undergoing major abdominal surgery. Vasoplegia was found to be the most frequent type. These novel classifications can help guide targeted treatment strategies for managing hemodynamic instability and reducing the risk of intraoperative hypotension.

An algorithm to detect dicrotic notch in arterial blood pressure and photoplethysmography waveforms using the iterative envelope mean method

Background and objectiveDetection of the dicrotic notch (DN) within a cardiac cycle is essential for assessment of cardiac output, calculation of pulse wave velocity, estimation of left ventricular ejection time, and supporting feature-based machine learning models for noninvasive blood pressure estimation, and hypotension, or hypertension prediction. In this study, we present a new algorithm based on the iterative envelope mean (IEM) method to detect automatically the DN in arterial blood pressure (ABP) and photoplethysmography (PPG) waveforms. MethodsThe algorithm was evaluated on both ABP and PPG waveforms from a large perioperative dataset (MLORD dataset) comprising 17,327 patients. The analysis involved a total of 1,171,288 cardiac cycles for ABP waveforms and 3,424,975 cardiac cycles for PPG waveforms. To evaluate the algorithm's performance, the systolic phase duration (SPD) was employed, which represents the duration from the onset of the systolic phase to the DN in the cardiac cycle. Correlation plots and regression analysis were used to compare the algorithm against marked DN detection, while box plots and Bland-Altman plots were used to compare its performance with both marked DN detection and an established DN detection technique (second derivative). The marking of the DN temporal location was carried out by an experienced researcher using the help of the 'find_peaks' function from the scipy Python package, serving as a reference for the evaluation. The marking was visually validated by both an engineer and an anesthesiologist. The robustness of the algorithm was evaluated as the DN was made less visually distinct across signal-to-noise ratios (SNRs) ranging from -30 dB to -5 dB in both ABP and PPG waveforms. ResultsThe correlation between SPD estimated by the algorithm and that marked by the researcher is strong for both ABP (R2(87,343) =0.99, p<.001) and PPG (R2(86,764) =0.98, p<.001) waveforms. The algorithm had a lower mean error of DN detection (s): 0.0047 (0.0029) for ABP waveforms and 0.0046 (0.0029) for PPG waveforms, compared to 0.0693 (0.0770) for ABP and 0.0968 (0.0909) for PPG waveforms for the established 2nd derivative method. The algorithm has high rate of detectability of DN detection for SNR of >= -9 dB for ABP waveforms and >= -12 dB for PPG waveforms indicating robust performance in detecting the DN when it is less visibly distinct. ConclusionOur proposed IEM- based algorithm can detect DN in both ABP and PPG waveforms with low computational cost, even in cases where it is not distinctly defined within a cardiac cycle of the waveform (‘DN-less signals’). The algorithm can potentially serve as a valuable, fast, and reliable tool for extracting features from ABP and PPG waveforms. It can be especially beneficial in medical applications where DN-based features, such as SPD, diastolic phase duration, and DN amplitude, play a significant role.

Wave Separation Analysis to Assess Cardiovascular Alterations Induced by Sepsis.

Sepsis induces a severe decompensation of arterial and cardiac functional properties, leading to important modifications of arterial blood pressure (ABP) waveform, not resolved by recommended therapy, as shown by previous works. The aim of this study is to quantify the changes in ABP waveform morphology and wave reflections during a long-term swine experiment of polymicrobial sepsis and resuscitation, to deepen the understanding of the cardiovascular response to standard resuscitation therapy. We analyzed 14 pigs: polymicrobial sepsis was induced in 9 pigs followed by standard resuscitation and 5 pigs were treated as sham controls. Septic animals were studied at baseline (T1), after sepsis development (T2), and after 24 h (T3) and 48 h (T4) of therapy administration, and sham controls at the same time points. ABP and arterial blood flow were measured in the left and right carotid artery, respectively. Pulse wave analysis and wave separation techniques were used to estimate arterial input impedance, carotid characteristic impedance, forward and backward waves, indices of wave reflections such as reflection magnitude and reflection index, and augmentation index. Sepsis led to an acute alteration of ABP waveform passing from type A to type B or C; consistently, the reflection phenomena were significantly reduced. The resuscitation was successful in reaching targeted hemodynamic stability, but it failed in restoring a physiological blood propagation and reflection. Septic pigs persistently showed altered reflected waves even after 48 hours of successful therapy according to guidelines, suggesting a persistent hidden cardiovascular disorder. The proposed indices may be useful to unravel the complex cardiovascular response to therapy administration in septic patients and could potentially be used for risk stratification of patient deterioration. Whether alterations of blood propagation and reflection contribute to persisting organ dysfunction after hemodynamic stabilization should be further investigated.

Comparison of Left Ventricular End-Diastolic Volume Approximated from Mean Blood Pressure and Stroke Volume and End-Diastolic Volume Calculated from Left Ventricular-Aortic Coupling.

Objectives: The purpose of this study was to compare left ventricular end-diastolic volume (EDV), derived from left ventricular arterial coupling (Ees/Ea), and mean arterial blood pressure. Both of these methods of measuring EDV require some invasive procedure. However, the method of measuring EDV approximate is less invasive than the EDV coupling measuring method. This is because EDV approximate only requires arterial pressure waveform as an invasive procedure. Methods: This study included 14 patients with normal cardiac function who underwent general anesthesia. The point when blood pressure stabilized after the induction of anesthesia was taken as a baseline according to the study protocol. At the point when systolic arterial blood pressure fell 10% or more from the baseline blood pressure, 300 mL of colloid solution was administered over 15 min. EDV approximate and EDV coupling were calculated for each of the 14 patients at three points during the course of anesthetic. Each value was obtained by calculating a 5 min average. The timing of these three points was 5 min before, 5 min during, and 5 min after infusion loading. Results: The total number of comparable points was 42; 3 points were taken from each of the 14 participants. Both EDV approximate and EDV coupling increased through the infusion load testing. Scatter plots were prepared, and regression lines were calculated from the obtained values. A high correlation was shown between EDV approximate and EDV coupling (R2 = 0.96, p < 0.05). Conclusions: In patients with good cardiac function, EDV approximate can be substituted for EDV coupling, suggesting the possibility that EDV can be continuously and less invasively calculated under the situation of general anesthesia.

Increased left atrial pressure distorts pulmonary arterial impedance estimates in the low-range of frequency in rats

BACKGROUD: Post-capillary pulmonary hypertension (PH), characterized as increased downstream pressure such as left atrial pressure (LAP), is the most prevalent type of PH and a significant contributor to right heart failure. Accurate estimation of pulmonary arterial characteristics, serving as right ventricular afterload, is crucial for management of PH. These characteristics can be captured comprehensively through pulmonary arterial impedance (PAZ), which elucidates pulmonary arterial pressure-flow relationship within the frequency domain. Although LAP waveforms are potential to affect the contours of pulmonary arterial pressure (PAP) and flow (PAF) waveforms, it remains unknown whether increased LAP affects the accuracy of PAZ measurement. This study aimed to investigate the impact of LAP on the accuracy of PAZ using an acute rodent model. METHODS: We measured pulmonary arterial blood flow (PAF), PAP, and LAP simultaneously in eight male Sprague-Dawley rats. PAZ was calculated as a reciprocal of admittance. Admittance was calculated in the following two ways: 1) one-input (PAP) and one-output (PAF) analysis (I1O1 analysis): the ratio of PAP(f) to PAF(f) at each frequency, where PAP(f) and PAF(f) are frequency spectra of PAP and PAF, respectively, and 2) two-input (PAP and LAP) and one-output (PAF) analysis (I2O1 analysis): excluding the potential effect of LAP from PAP and PAF. We measured PAZ under both normal physiological condition and post-capillary PH condition. To induce acute post-capillary PH by raising LAP &gt;15 mmHg, we ligated aortic arch at the site between brachiocephalic artery and left common carotid artery and performed blood transfusion from a donor rat (TAL/BT). The accuracy of PAZ was examined as the coherence function representing the linearity between the input and output signals. Results: TAL/BT significantly increased LAP from 3 ± 1mmHg to 21 ± 4 mmHg and PAP from 17 ± 3 mmHg to 34 ± 3 mmHg (p&lt;0.01 for both). The I2O1 analysis provided a consistently accurate estimation of PAZ under both normal and elevated LAP conditions with high coherence function (&gt;0.9) in the wide range of frequency between 0.12 and 10.7 Hz. The I1O1 analysis also estimated the PAZ accurately in normal condition with high coherence function (&gt;0.9) in the wide range of frequency, however, the I1O1 analysis overestimated PAZ in the low-range frequency between 0.12 and 1.0 Hz compared with the I2O1 analysis (I1O1: 0.32±0.10 vs. I2O1 0.23±0.06 mmHg/ml/min, p&lt;0.01), with significantly lower coherence function (I1O1: 0.77±0.05 vs. I2O1: 0.96±0.01, p&lt;0.01). CONCLUSION: Increased LAP may have the impact on the accuracy of PAZ estimation in the low-range frequency. The I2O1 analysis can lead to more accurate assessment of right ventricular afterload in post-capillary PH, compared with the I1O1 analysis. Grant-in-Aid for Young Scientists of the Japanese Society of Cardiovascular Anesthesiologists Project Number 2271000048. This is the full abstract presented at the American Physiology Summit 2024 meeting and is only available in HTML format. There are no additional versions or additional content available for this abstract. Physiology was not involved in the peer review process.

Cerebrovascular Compliance during Progressive Hypotension in Patients with Autonomic Failure

Background: Cerebral perfusion during reductions in arterial pressure (AP) can be supported transiently by a Windkessel mechanism that increases cerebrovascular compliance (Ci). This defense mechanism exists in healthy participants during rapid standing and in patients during vasovagal syncope. However, it remains unknown whether the rise in Ci with falling AP exhibits a threshold effect. Thus, we tested the hypothesis that increased Ci is passive to a reduction in AP and starts to increase only after a threshold has been reached. To test this hypothesis, it was necessary to study a population that tolerates progressive large reductions in BP without a loss of consciousness. Therefore, we analyzed Ci changes during a graded head-up tilt (HUT) from patients with autonomic failure (AF). Methods: Finger AP and right middle cerebral artery blood velocity (MCAv) were recorded from 5 AF patients (n = 5 male, 61 ± 23 years, 25 ± 5 kg/m2) at supine rest and during graded-HUT (30, 45, 60 degrees). Tilt gradients increased incrementally every 5 minutes until AP reached a critically low value prompting termination of HUT. The total time in HUT was 11 ± 4 min. Individual AP and MCAv waveforms during supine rest and graded HUT (one waveform every 5 s) were assessed with a modified Windkessel model to calculate Ci and cerebrovascular resistance (CVR). Pulse pressure (PP) was calculated as systolic AP − diastolic AP. A threshold value for the increase in Ci was determined as five consecutive cardiac cycles greater than the average supine Ci plus 2 standard deviations. Results: Graded HUT resulted in an overall reduction in AP (85 ± 18 to 50 ± 7 mmHg), PP (72 ± 22 to 32 ± 19 mmHg), mean MCAv (62 ± 12 to 35 ± 7 cm/s), and CVR (1.4 ± 0.1 to 1.0 ± 0.3 mmHg/cm/s; all p &lt; 0.05 vs supine). Ci increased by 905 ± 756% (p = 0.04 vs supine). The threshold effect was evident as Ci began to increase progressively after PP fell by 15 ± 3 mmHg and AP fell by 12 ± 2 mmHg. Conclusions: Our results provide evidence that Ci increases after a threshold reduction in PP and AP in the apparent absence of autonomic support. These data support the idea that increasing Ci during reductions in AP represents a pressure-dependent Windkessel mechanism. Of interest, our findings provide clinical insight into how AF patients compensate for such large reductions in pressure without the assistance of CVR as a defense mechanism. NSERC USRA — Western University, Natural Sciences and Engineering Research Council of Canada (NSERC, Grant No. RGPIN-2018-06255). This is the full abstract presented at the American Physiology Summit 2024 meeting and is only available in HTML format. There are no additional versions or additional content available for this abstract. Physiology was not involved in the peer review process.

Reconstruction of arterial blood pressure waveforms based on squeeze-and-excitation network models using electrocardiography and photoplethysmography signals

Arterial blood pressure (ABP) waveforms indicate the efficiency with which a patient’s blood responds to changes in the arterial flow. ABP waveforms can be used to analyze the cardiovascular health of patients, including heart rate analysis, stress analysis, and cardiovascular disease diagnosis. However, obtaining the ABP waveform is more difficult than recording the electrocardiography (ECG) and photoplethysmography (PPG) signals. Several studies have focused on the reconstruction of the ABP waveform by recombining bio-signals. Reconstructing ABP waveforms is considered a major challenge because of signals being distorted by waveform destruction, owing to the normalization and standardization of the input ABP waveform. This paper presents a model that can simultaneously output blood pressure estimation and ABP waveforms using the raw ABP signal as the input. Herein, the squeeze-and-excitation network (SE-Net) is coupled with the multi-task learning architecture. Three signals, namely, the ECG, PPG, and ECG–PPG, were trained as feature vectors using different variants of the U-Net model, and the results are compared. Additionally, we introduce a novel method for incorporating white Gaussian noise at different signal-to-noise ratios (SNRs) to augment the training data. The objective of the proposed approach is to evaluate the ABP signal reconstruction capability of the SE-Net at various SNR levels and analyze the overall performance and robustness of the model. In the case of systolic blood pressure, the root mean square error (RMSE) of the PPG-based values is 2.58 mmHg, and the Pearson correlation coefficient is r = 0.92 (p≤ 0.001). In the case of diastolic blood pressure(DBP), the RMSE was 3.35 mmHg, and the Pearson correlation coefficient was r = 0.89 (p ≤ 0.001). The results indicate the potential for replacing ECG signals with PPG signals to overcome the constraints of optimizing the non-invasive reconstruction of ABP waveforms using SE-Net. Furthermore, we demonstrate that the SE-Net-combined multi-task learning architecture model can simultaneously perform blood pressure estimation and output ABP waveforms without damaging the raw signal.

Deep-learning-based real-time individualization for reduce-order haemodynamic model

The reduced-order lumped parameter model (LPM) has great computational efficiency in real-time numerical simulations of haemodynamics but is limited by the accuracy of patient-specific computation. This study proposed a method to achieve the individual LPM modeling with high accuracy to improve the practical clinical applicability of LPM.Clinical data was collected from two medical centres comprising haemodynamic indicators from 323 individuals, including brachial artery pressure waveforms, cardiac output data, and internal carotid artery flow waveforms. The data were expanded to 5000 synthesised cases that all fell within the physiological range of each indicator. LPM of the human blood circulation system was established. A double-path neural network (DPNN) was designed to input the waveforms of each haemodynamic indicator and their key features and then output the individual parameters of the LPM, which was labelled using a conventional optimization algorithm. Clinically collected data from the other 100 cases were used as the test set to verify the accuracy of the individual LPM parameters predicted by DPNN.The results show that DPNN provided good convergence in the training process. In the test set, compared with clinical measurements, the mean differences between each haemodynamic indicator and the estimate calculated by the individual LPM based on the DPNN were about 10 %. Furthermore, DPNN prediction only takes 4 s for 100 cases.The DPNN proposed in this study permits real-time and accurate individualization of LPM's. When facing medical issues involving haemodynamics, it lays the foundation for patient-specific numerical simulation, which may be beneficial for potential clinical application.

Impact of respiratory cycle during mechanical ventilation on beat-to-beat right ventricle stroke volume estimation by pulmonary artery pulse wave analysis.

The same principle behind pulse wave analysis can be applied on the pulmonary artery (PA) pressure waveform to estimate right ventricle stroke volume (RVSV). However, the PA pressure waveform might be influenced by the direct transmission of the intrathoracic pressure changes throughout the respiratory cycle caused by mechanical ventilation (MV), potentially impacting the reliability of PA pulse wave analysis (PAPWA). We assessed a new method that minimizes the direct effect of the MV on continuous PA pressure measurements and enhances the reliability of PAPWA in tracking beat-to-beat RVSV. Continuous PA pressure and flow were simultaneously measured for 2-3min in 5 pigs using a high-fidelity micro-tip catheter and a transonic flow sensor around the PA trunk, both pre and post an experimental ARDS model. RVSV was estimated by PAPWA indexes such as pulse pressure (SVPP), systolic area (SVSystAUC) and standard deviation (SVSD) beat-to-beat from both corrected and non-corrected PA signals. The reference RVSV was derived from the PA flow signal (SVref). The reliability of PAPWA in tracking RVSV on a beat-to-beat basis was enhanced after accounting for the direct impact of intrathoracic pressure changes induced by MV throughout the respiratory cycle. This was evidenced by an increase in the correlation between SVref and RVSV estimated by PAPWA under healthy conditions: rho between SVref and non-corrected SVSD -0.111 (0.342), corrected SVSD 0.876 (0.130), non-corrected SVSystAUC 0.543 (0.141) and corrected SVSystAUC 0.923 (0.050). Following ARDS, correlations were SVref and non-corrected SVSD -0.033 (0.262), corrected SVSD 0.839 (0.077), non-corrected SVSystAUC 0.483 (0.114) and corrected SVSystAUC 0.928 (0.026). Correction also led to reduced limits of agreement between SVref and SVSD and SVSystAUC in the two evaluated conditions. In our experimental model, we confirmed that correcting for mechanical ventilation induced changes during the respiratory cycle improves the performance of PAPWA for beat-to-beat estimation of RVSV compared to uncorrected measurements. This was demonstrated by a better correlation and agreement between the actual SV and the obtained from PAPWA.

Use of a Non-invasive Central Arterial Pressure Waveform Analysis Device in Stroke Patients – A Feasibility Study (P9-5.008)

Use of a Non-invasive Central Arterial Pressure Waveform Analysis Device in Stroke Patients – A Feasibility Study (P9-5.008)

The Edwards Acumen IQ system using peripheral arterial catheter-based waveforms to estimate cardiac output is not accurate as compared to thermodilution in dogs.

To assess the agreement between cardiac output (CO) estimated via evaluation of the arterial pressure waveform by a novel monitoring system (Edwards Acumen IQ sensor and HemoSphere Advanced Monitor Platform [HS-IQ]; Edwards LifeSciences) and measured by thermodilution (TD) in anesthetized, normovolemic, and hypovolemic dogs. To assess the agreement between the HS-IQ CO measurements in the radial artery and dorsal metatarsal artery. 8 purpose-bred Beagles. Dogs were placed under general anesthesia. CO was measured via TD and via the HS-IQ at radial and dorsal metatarsal arterial catheters. CO measurements were obtained at 4 time points including normovolemic and multiple hypovolemic states. Paired measurements of CO were evaluated via the method of Bland and Altman with acceptable limits of agreement (LOA) defined as < 30%. A total of 24 (dorsal metatarsal) and 21 (radial) paired measurements were collected in 8 dogs. The overall bias (CI) for comparison of TD to radial arterial HS-IQ CO measurements was -0.09 L/min. LOA and proportional LOA were -2.66 to 2.49 L/min and -140.72% to 104.94%. The overall bias (CI) for comparison of TD to dorsal metatarsal arterial HS-IQ CO measurements was -0.26 L/min. LOA and proportional LOA were -2.76 to 2.24 L/min and -135.96% to 93.25%. The overall proportional error for radial arterial was -17.9% and for dorsal metatarsal was -21.4%. CO measurements with the HS-IQ were easy to obtain but did not produce results within a clinically acceptable range for either measurement site, with a very wide LOA. The CO estimations from the HS-IQ are not appropriate for clinical use at this time.

Detecting aortic valve stenosis based on the non-invasive blood pressure waveform—a proof of concept study

The incidence of aortic valve stenosis (AoS) increases with age, and once diagnosed, symptomatic severe AoS has a yearly mortality rate of 25%. AoS is diagnosed with transthoracic echocardiography (TTE), however, this gold standard is time consuming and operator and acoustic window dependent. As AoS affects the arterial blood pressure waveform, AoS-specific waveform features might serve as a diagnostic tool. Aim of the present study was to develop a novel, non-invasive, AoS detection model based on blood pressures waveforms. This cross-sectional study included patients with AoS undergoing elective transcatheter or surgical aortic valve replacement. AoS was determined using TTE, and patients with no or mild AoS were labelled as patients without AoS, while patients with moderate or severe AoS were labelled as patients with AoS. Non-invasive blood pressure measurements were performed in awake patients. Ten minutes of consecutive data was collected. Several blood pressure-based features were derived, and the median, interquartile range, variance, and the 1st and 9th decile of the change of these features were calculated. The primary outcome was the development of a machine-learning model for AoS detection, investigating multiple classifiers and training on the area under the receiver-operating curve (AUROC). In total, 101 patients with AoS and 48 patients without AoS were included. Patients with AoS showed an increase in left ventricular ejection time (0.02 s, p = 0.001), a delayed maximum upstroke in the systolic phase (0.015 s, p < 0.001), and a delayed maximal systolic pressure (0.03 s, p < 0.001) compared to patients without AoS. With the logistic regression model, a sensitivity of 0.81, specificity of 0.67, and AUROC of 0.79 were found. The majority of the population without AoS was male (85%), whereas in the population with AoS this was evenly distributed (54% males). Age was significantly (5 years, p < 0.001) higher in the population with AoS. In the present study, we developed a novel model able to distinguish no to mild AoS from moderate to severe AoS, based on blood pressure features with high accuracy. Clinical registration number: The study entailing patients with TAVR treatment was registered at ClinicalTrials.gov (NCT03088787, https://clinicaltrials.gov/ct2/show/NCT03088787). The study with elective cardiac surgery patients was registered with the Netherland Trial Register (NL7810, https://trialsearch.who.int/Trial2.aspx?TrialID=NL7810).