Survival Analysis Using Time-Frequency Analysis of Heart Rate Variability During Exercise

Abstract

Heart rate variability as an indicator of increased morbidity has been established in previous studies. It is defined by different frequency bands, corresponding to different biological mechanisms. This thesis aims to study heart rate variability indices in the time-frequency domain through extraction from the UK Biobank dataset.

Time-frequency indices for different time intervals during exercise were extracted. A time-frequency based filter analyzing the presence of high frequency noise was developed as a way of detecting and removing noisy sections of the ECG recordings. This noise reduction algorithm was coupled with time-domain noise reduction methods as a part of pre-processing the data before survival analysis. Intra-individual repeatability of indices was calculated and finally a combination of a random survival forest and Cox proportional hazards regression was used to evaluate said indices' value as indicators of cardiovascular risk.

The noise reduction performed by the combination of time and time-frequency domain noise rejection algorithms did not improve intra-patient repeatability. Selected time-frequency indices show high levels of intra-individual repeatability over a three year period. Out of the selected indices, one remained a significant predictor of cardiovascular event when combined with demographic data.

The use of time-frequency based noise filtering shows promise for detection of high frequency noise artifacts, and should be further studied and tested on other kinds of recordings. The significant relationship between one time-frequency index of heart rate variability and increased risk of cardiovascular events needs to be further investigated.

Popular Abstract

Survival Analysis Using Time-Frequency Analysis of heart Rate Variability During Exercise

Heart rate variability (HRV) refers to changes in time between heart beats over time and has been well established as a risk factor for, or associated with, all-cause mortality after experiencing myocardial infarction, post-traumatic stress disorder and irritable bowel syndrome, among other things. The work presented in this paper presents an evaluation of time-frequency indices of HRV changes extracted from over 100,000 ECG recordings for seven minutes long exercise bike tests. A link between said indices and increased risk of experiencing cardiac events was established.

Heart rate variability is used to gain a measure of neural activity in the human body, where different frequencies of change in heart rate correspond to different parts of our nervous system. The reason why HRV is an appealing measure to use for gauging nervous activity as opposed to directly measuring it is that it can be registered by non-invasive means instead of for example drawing blood samples. This can be achieved via for example electrodes on the skin or pulse oximeters.

This work is based on ECG-recordings during exercise, where participants have cycled on an exercise bike for seven minutes at different workloads while being recorded. From this recording, the time between heart beats can be extracted via thresholding algorithms. Subsequently, the changes in this beat-to-beat time is extracted. Because these recordings are taken from individuals during increasing levels of exercise, a method for handling changing (i.e. non-stationary) heart rate is necessary. As the resistance on the bike becomes higher, the participants’ heart rates will increase. With a higher heart rate comes a lower HRV, and so to study HRV over a non-stationary process a time-frequency approach was used.

Time-frequency methods allow us to see frequency contents in a signal for a given time, something which for example a Fourier transform would not be able to. By looking at the time-frequency distribution of HRV, an estimate of its near-instant changes is achievable. These estimates have been proven to remain stable over three years the part of the participants who took the exercise bike test twice, and one has also proven to be linked with increased risk of suffering from cardiac events such as heart attacks. The resulting link goes against previously established knowledge about HRV, which may be due to the fact that previous studies have not looked at such a large population before.

The individuals who partook in the study are non-clinical, that is they were not recruited because they were showing any form of symptoms. Instead, this study mostly consisted of data from healthy individuals. This means that the results from this project can be applied as a form of screening methodology, which in theory could be implemented in exercise bikes located in for example gyms or in your home.

• author: Sjögren, Carl
• supervisor: Martin Stridh
• organization: Department of Biomedical Engineering
• course: BMEM01 20191
• year: 2019
• type: H2 - Master's Degree (Two Years)
• subject: Technology and Engineering
• language: English
• additional info: 2019-05
• id: 8975358
• date added to LUP: 2019-05-13 15:56:16
• date last changed: 2019-05-13 16:09:55