University of Twente Student Theses
Comparison of two algorithms to detect change in aerobic capacity over time.
Jong, J.C.W. de (2022) Comparison of two algorithms to detect change in aerobic capacity over time.
|
PDF
1MB |
| Abstract: | Many people struggle to quit smoking and start running since the chance on relapse is very high amongst self-quitters. That is why Roessingh research and development, University of Twente and other partners designed the Perfect Fit program. The program uses a virtual coach that gives the participants a personalised training plan based on the change in aerobic capacity over the time. In this research two algorithms to estimate the change in aerobic capacity over time were designed and analysed. The first algorithm is based on the algorithm used by Garmin in their wearables and uses the positive linear relation between heart rate and VO2 to estimate the VO2max of the runner. The second algorithm uses the positive linear relation between heart rate and speed to calculate the heart rate-speed ratio. Both VO2max and the heart rate-speed ratio are a quantification of the aerobic capacity. A data set with 888 running trainings and two VO2max measurements is used to analyse and validate the designed algorithms. The measured change in aerobic capacity is an improvement of 25 ml/kg/min (53 %) between the two VO2max measurements. The change estimated by the Garmin algorithm is an improvement by only 11 ml/kg/min (19,64 %), where the heart rate-speed ratio algorithm also calculated an increase by only 1,68 bpm/(km/h) (11,38 %). Therefore both algorithms underestimated the change in aerobic capacity over time. The underestimation could be due to the low amount of runs containing heart rate data and to unknown external factors such as cold or hilly terrain. To conclude, it was not possible to accurately predict the change in aerobic capacity over time with the use of an algorithm for this data set. However it is not clear if the algorithms would work on data sets containing more VO2max measurements and more consistent training. Overall more validation and calibration needs to be done to use the algorithms in the Perfect Fit program. |
| Item Type: | Essay (Bachelor) |
| Faculty: | TNW: Science and Technology |
| Subject: | 42 biology |
| Programme: | Biomedical Technology BSc (56226) |
| Link to this item: | https://purl.utwente.nl/essays/89606 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page