Multi-feature function-on-scalar regression for modelling stepwatch activity data

functional-on-scalar regression models;StepwatchTM Activity Monitor;transtibial amputation;limb salvage;Biostatistics

Li, HanyingZipunnikov, Vadim ;

Johns Hopkins University

关键词: functional-on-scalar regression models;    StepwatchTM Activity Monitor;    transtibial amputation;    limb salvage;    Biostatistics;   

Others  :  https://jscholarship.library.jhu.edu/bitstream/handle/1774.2/39482/LI-THESIS-2016.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: JOHNS HOPKINS DSpace Repository

【 摘 要 】

In this thesis, we propose to use function-on-scalar regression models for modeling summaries derived from minute-to-minute step count data collected with StepwatchTM Activity Monitor (SAM). The approach has been motivated by the OUTLET study that compares the transtibial amputation vs limb salvage and use SAM as an objective measure of the 18 month functional outcomes. We compare both daily summaries and temporal change in the day of different measures of walking activities of amputee and salvage groups. On the whole, amputation and salvage treatments for severe distal, tibia, ankle and/or foot trauma result in comparable functional outcomes measured with step activity monitors. However, most of the summaries that measure the stability and variability of walking activity are significantly influenced by amputation, at least for a period of time in the day.