Machine Learning-Based Pulse Wave Analysis for Early Detection of Abdominal Aortic Aneurysms Using In Silico Pulse Waves

Abstract

An abdominal aortic aneurysm (AAA) is usually asymptomatic until rupture, which is associated with extremely high mortality. Consequently, the early detection of AAAs is of paramount importance in reducing mortality; however, most AAAs are detected by medical imaging only incidentally. The aim of this study was to investigate the feasibility of machine learning-based pulse wave (PW) analysis for the early detection of AAAs using a database of in silico PWs. PWs in the large systemic arteries were simulated using one-dimensional blood flow modelling. A database of in silico PWs representative of subjects (aged 55, 65 and 75 years) with different AAA sizes was created by varying the AAA-related parameters with major impacts on PWs—identified by parameter sensitivity analysis—in an existing database of in silico PWs representative of subjects without AAAs. Then, a machine learning architecture for AAA detection was trained and tested using the new in silico PW database. The parameter sensitivity analysis revealed that the AAA maximum diameter and stiffness of the large systemic arteries were the dominant AAA-related biophysical properties considerably influencing the PWs. However, AAA detection by PW indexes was compromised by other non-AAA related cardiovascular parameters. The proposed machine learning model produced a sensitivity of 86.8 % and a specificity of 86.3 % in early detection of AAA from the photoplethysmogram PW signal measured in the digital artery with added random noise. The number of false positive and negative results increased with increasing age and decreasing AAA size, respectively. These findings suggest that machine learning-based PW analysis is a promising approach for AAA screening using PW signals acquired by wearable devices.

The effect of cardiac properties on arterial pulse waves: An in‐silico study

“…In particular, the photoplethysmogram (PPG) pulse wave is easily acquired using pulse oximeters, which are frequently used in healthcare settings to measure arterial blood oxygen saturation and pulse rate. PPG signals can also be acquired by devices available to the wider population, such as smartwatches and fitness bands 7,8 …”

“…PPG signals can also be acquired by devices available to the wider population, such as smartwatches and fitness bands. 7,8 Databases of in silico pulse waves signals representative of cohorts of real subjects can be produced using robust and efficient computational blood flow models for the development and pre-clinical testing of pulse wave analysis algorithms. 9,10 Virtual subjects are characterised by haemodynamic variables spanning the physiological range, even in disease-related conditions.…”

A machine learning model for non-invasive detection of atherosclerotic coronary artery aneurysm

“…Similar studies were performed by Liang et al [ 9 ] to investigate the relationship between morphometric properties and numerically predicted risk of aortic aneurysm growth applying an ML algorithm. Most of the performed studies related to using AI in aneurysms are supported by data sets from kinds of studying methods such as in-vivo clinical experiments [ 10 ], segmentation 1-dimensional flow [ 11 , 12 ], the DL techniques based on in silico pressure [ 13 ], and CTA-based datasets. However, some studies are focused on explaining growth [ 14 ], automatic detection according to morphometric features [ 15 – 17 ] and early detection [ 10 ] of aortic aneurysms.…”

Classification and regression of stenosis using an in-vitro pulse wave data set: Dependence on heart rate, waveform and location

“…Much of the work in previous studies on aneurysms was based on simulation data. While most approaches apply machine learning in the time domain [7][8][9], there is also recent work where the features are based on the Fourier series representation [10], resulting in a reduced parameter space when a limited number of modes is used. However, the underlying model assumptions often reduce the complexity of the generated data, including the number and statistical distribution of parameters that are varied in a virtual patient cohort.…”

“…There are no publicly available large in-vitro data sets to validate algorithms for classification and regression of structural variations within the arterial system. Most publicly available pulse wave databases are generated by computer simulations [8,22], which offer flexibility in generating a large amount of data. The development of new algorithms gives hope for applications on real-world data, since real-world data is accompanied by specific challenges that must be tackled and are difficult to infer from simulation-only data.…”