Deep learning for prediction of cardiac indices from photoplethysmographic waveform: A virtual database approach

Abstract

Deep learning methods combined with large datasets have recently shown significant progress in solving several medical tasks. However, collecting and annotating large datasets can be a very cumbersome and expensive task. We tackle these problems with a virtual database approach where training data is generated using computer simulations of related phenomena. Specifically, we concentrate on the following problem: can cardiovascular indices such as aortic elasticity, diastolic and systolic blood pressures, and blood flow from heart be predicted continuously using wearable photoplethysmographic sensors? We simulate the blood flow using a haemodynamic model consisting of the entire human circulation. Repeated evaluation of the simulator allows us to create a database of “virtual subjects” with size that is only limited by available computational resources. Using this database, we train neural networks to predict the cardiac indices from photoplethysmographic signal waveform. We consider two approaches: neural networks based on predefined input features and deep convolutional neural networks taking waveform directly as the input. The performance of the methods is demonstrated using numerical examples, thus carrying out a preliminary assessment of the approaches. The results show improvements in accuracy compared with the previous methods. The improvements are especially significant with indices related to aortic elasticity and maximum blood flow. The proposed approach would provide new means to measure cardiovascular health continuously, for example, with a simple wrist device.

Towards enabling a cardiovascular digital twin for human systemic circulation using inverse analysis

“… 2019 ; Huttunen et al. 2019 ). Two key components are required to build such realistic models, a validated one-dimensional haemodynamic model and realistic arterial networks based on anthropometric and haemodynamic parameters to represent a reliable human cohort.…”

“…The objective of virtual patient generation for a cardiovascular system is to build a numerical model for blood flow that closely resembles a human patient in terms of blood flow parameters and vascular network. Previous research in virtual patient database generation methods can be found in references (Willemet et al 2015;Charlton et al 2019;Huttunen et al 2019). Two key components are required to build such realistic models, a validated one-dimensional haemodynamic model and realistic arterial networks based on anthropometric and haemodynamic parameters to represent a reliable human cohort.…”

Determination of Aortic Characteristic Impedance and Total Arterial Compliance From Regional Pulse Wave Velocities Using Machine Learning: An in-silico Study

“…Yet, the in-silico data allow us for performing an initial validation of the proposed methodology, whose results will allow to proceed with the clinical validation. Previous works have used a similar approach to validate ML-based techniques using virtual patients when real clinical data were not available ( Huttunen et al, 2019 ; Bikia et al, 2020b , 2021 ; Huttunen et al, 2020 ). Hence, the present study proposes the methodology rather than the model per se .…”

Noninvasive estimation of aortic hemodynamics and cardiac contractility using machine learning

“…They have provided a valuable alternative for the assessment of pressure and flow in the entire arterial network providing additional pathophysiological insights, which are difficult to acquire in-vivo. Numerous previous studies have used in-silico data for the estimation of aortic BP, cardiac output, aortic PWV and many more [56][57][58][59][60] . Importantly, in-silico studies allow for the preliminary evaluation of predictive models across a wide range of cardiovascular parameters 61 in a quick and cost-efficient way, while their results can be rather informative of the design of clinical studies 62,63 .…”