Abstract: Transformers are recent deep learning (DL) models used to capture the dependence between parts of sequential data. While…
ISOQOL 29th Annual Conference, Prague, Czech Republic, October 2022
On Data Science Day, University of Geneva, Switzerland, September 15, 2022. Igor Matias1, Matthias Kliegel1, Katarzyna Wac1 1University of Geneva, Switzerland Poster: here
Abstract:
Background
Heart rate (HR), especially at nighttime, is an important biomarker for cardiovascular health. It is known to be influenced by overall physical fitness, as well as daily life physical or psychological stressors like exercise, insufficient sleep, excess alcohol, certain foods, socialization, or air travel causing physiological arousal of the body. However, the exact mechanisms by which these stressors affect nighttime HR are unclear and may be highly idiographic (i.e. individual-specific). A single-case or “n-of-1” observational study (N1OS) is useful in exploring such suggested effects by examining each subject’s exposure to both stressors and baseline conditions, thereby characterizing suggested effects specific to that individual.
Objective
Our objective was to test and generate individual-specific N1OS hypotheses of the suggested effects of daily life stressors on nighttime HR. As an N1OS, this study provides conclusions for each participant, thus not requiring a representative population.
Methods
We studied three healthy, nonathlete individuals, collecting the data for up to four years. Additionally, we evaluated model-twin randomization (MoTR), a novel Monte Carlo method facilitating the discovery of personalized interventions on stressors in daily life.
Results
We found that physical activity can increase the nighttime heart rate amplitude, whereas there were no strong conclusions about its suggested effect on total sleep time. Self-reported states such as exercise, yoga, and stress were associated with increased (for the first two) and decreased (last one) average nighttime heart rate.
Conclusions
This study implemented the MoTR method evaluating the suggested effects of daily stressors on nighttime heart rate, sleep time, and physical activity in an individualized way: via the N-of-1 approach. A Python implementation of MoTR is freely available.
In CUSO Winter School, Champéry, Switzerland, January 31 – February 4, 2022. Igor Matias1, Matthias Kliegel1, Katarzyna Wac1 1University of Geneva, Switzerland Poster: here
In UNIGE’s Data Science Day round table “Beyond SDGs: measuring and forecasting for taking action”, Geneva, Switzerland, September 16, 2021. Igor Matias1 (September 2021)1University of Geneva, Switzerland https://datascience.unige.ch/recherche/uniges-data-science-days
On UNIGE’s Data Science Day, Geneva, Switzerland, September 16, 2021. Vlad Manea1, Igor Matias2, Katarzyna Wac1,2 1University of Copenhagen, Denmark; 2University of Geneva, Switzerland https://datascience.unige.ch/recherche/uniges-data-science-days
Abstract: Inactivity, lack of sleep, and poor nutrition predispose individuals to health risks. Momentary Patient-Reported Outcomes (PROs) assess short-term physical behaviors and psychological states for a given recall period (ranging from weeks to seconds) but suffer from reporting biases. Conversely, wearables are
More
Abstract: N-of-1 randomized trials help better understand the effects of an intervention, e.g., behavioral changes in a specific person, allowing for the correlation and causation between, for example, duration of sleep and the level of physical exercise the following day.By leveraging the
More
Abstract: Atrial Fibrillation (AF) is a type of arrhythmia characterized by irregular heartbeats, with four types, two of which are complicated to diagnose using standard techniques such as Electrocardiogram (ECG). However, and because smart wearables are increasingly a piece of commodity equipment, there are several ways of detecting and predicting AF episodes using only an ECG exam, allowing physicians easier diagnosis. By searching several databases, this study presents a review of the articles published in the last ten years, focusing on those who reported studies using Artificial Intelligence (AI) for prediction of AF. The results show that only twelve studies were selected for this systematic review, where three of them applied deep learning techniques (25%), six of them used machine learning methods (50%) and three others focused on applying general artificial intelligence models (25%). To conclude, this study revealed that the prediction of AF is yet an under-developed field in the context of AI, and deep learning techniques are increasing the accuracy, but these are not as frequently applied as it would be expected. Also, more than half of the selected studies were published since 2016, corroborating that this topic is very recent and has a high potential for additional research.