Evaluating the Utility of a Cardiac Health Assessment Test in Predicting VO2max Obtained from CPET: A Pilot Study

Maximal oxygen uptake (VO(2)max) assessment, a pivotal metric in gauging cardiovascular fitness, has historically relied on labour-intensive protocols such as treadmill exercises. However, recent advancements in wearable technology, notably heart rate monitors, have revolutionized this process. This study leverages a game-based Cardiac Assessment Protocol, designed to simulate a exercise protocol using a wearable chest strap integrated with an electrocardiogram (ECG) sensor and accelerometer for estimation of VO(2)max. With a cohort comprising 22 subjects, the study focused on extracting features from both Heart Rate (HR) and acceleration data collected during Cardiac Assessment Protocol and Cardiopulmonary Exercise Testing (CPET) utilising the wearable chest strap. The ground truth for VO(2)max was derived from CPET. The Machine Learning (ML) model yielded VO(2)max predictions from HR data collected during CPET with a Root Mean Square Error (RMSE) of 5.37 compared with the ground truth COSMED VO(2)max data. Similarly, when VO(2)max was estimated from the HR data during the Cardiac Assessment Protocol, the model had a RMSE of 5.82 compared to the ground truth COSMED VO(2)max data. These results indicated that VO(2)max can be reasonably estimated from HR and acceleration data under a simulated incremental game-based Cardiac Assessment Protocol, providing a potential alternative to traditional exercise testing methods. Notwithstanding this, the proposed protocol demonstrates promising potential for VO(2)max estimation through a less invasive and more accessible approach to cardiovascular fitness assessment.