Research
Cancer Epidemiology
survival; competing risks; case-cohort; nested case-control; sampling; missing
Cohort studies provide information on the risk of cancers and other diseases. For a large cohort and rare outcomes, it is cost-effective to measure expensive covariates on all cases and a small subset of individuals sampled by two-phase cohort designs. The two most popular sub-sampling strategies are the nested case-control design (NCC) and the case-cohort design (CC). In NCC, controls are sampled for each case arising in a cohort from among the set of individuals at risk at the time of case diagnosis. CC collects all cases and a random sample of the entire cohort which may include cases.
Data Structure of Two-phase Epidemiological Cohort Design
ALS Disease
spatiotemporal; network; hidden Markov model; pseudo-likelihood; functional
Amyotrophic lateral sclerosis (ALS; 근위축성 측삭경화증), also known as Lou Gehrig's disease (루게릭병), is a neurological disease that mainly affects the nerve cells in the brain and the spinal cord that control voluntary muscle movements. The symptoms typically start from a particular group of muscles and then spread as the disease progresses. To study spreading patterns of muscle weakness, I analyze muscle strength data by modeling spatial association and/or temporal progression.
Nerve cells in brain and spinal cord
muscles where strength measured
Wearable Devices
digital biomarker; hidden Markov; machine learning; functional; longitudinal
Understanding how the improvement of physical activity and/or exposure to light in daily life can reduce disease risk is an important area in public health. Most literature addressing this question uses self-reported assessments of these exposures resulting in inconsistencies between studies. The use of wearable devices provides researchers the opportunity to explore the relationship between complex features of these exposures on disease risk in a more precise and systematic fashion.
Circadian Rhythm Analysis