Predicting Cardiovascular Rehabilitation of Patients with Coronary Artery Disease Using Transfer Feature Learning

Abstract

Cardiovascular diseases represent the leading cause of death worldwide. Thus, cardiovascular rehabilitation programs are crucial to mitigate the deaths caused by this condition each year, mainly in patients with coronary artery disease. COVID-19 was not only a challenge in this area but also an opportunity to open remote or hybrid versions of these programs, potentially reducing the number of patients who leave rehabilitation programs due to geographical/time barriers. This paper presents a method for building a cardiovascular rehabilitation prediction model using retrospective and prospective data with different features using stacked machine learning, transfer feature learning, and the joint distribution adaptation tool to address this problem. We illustrate the method over a Chilean rehabilitation center, where the prediction performance results obtained for 10-fold cross-validation achieved error levels with an NMSE of 0.03±0.013 and an R2 of 63%±19%, where the best-achieved performance was an error level with a normalized mean squared error of 0.008 and an R2 up to 92%. The results are encouraging for remote cardiovascular rehabilitation programs because these models could support the prioritization of remote patients needing more help to succeed in the current rehabilitation phase.

Machine-Learning based Prediction Models for Healthcare Outcomes in Patients Participating in Cardiac Rehabilitation: A Systematic Review

“…From the initial search that yielded 151 records, 81 were excluded following a review of the title and abstract, 60 records were excluded after full-text review. In the end, 7 studies met the inclusion criteria and were included in our systemic review (Figure 1 shows process of screening and study selection) [35][36][37][38][39][40][41] . [35][36][37][38]41 , one cross-sectional study 40 , and one that combines retrospective and prospective approaches 39 .…”

“…In the end, 7 studies met the inclusion criteria and were included in our systemic review (Figure 1 shows process of screening and study selection) [35][36][37][38][39][40][41] . [35][36][37][38]41 , one cross-sectional study 40 , and one that combines retrospective and prospective approaches 39 . Most studies target phase II of CR, with one addressing phases II and III 35 and another focusing solely on phase III 37 .…”

“…38 , various physiological and psychological outcome 35 . Additionally, two studies focused on predicting patient disposition post-CR, such as likelihood of returning to work 36 as well as the transition from one phase of CR to another 39 .…”

“…Laboratory test results and physical fitness levels were commonly used. Two studies harnessed specific imaging like EKG and TTE 38,39 .…”

“…Most studies used pre-processing techniques for feature selection to reduce the number of input variables to those that are most important to the predictive model to improve performance and reduce computational cost. RF and principal component analysis (PCA) were most employed 35,39,40 .…”