Validation of non-empirical fat-free mass estimation model for a wrist-worn device

Polokhin, A.; Pronina, A.; Boev, A.; Gorbunov, S.

Авторы: Polokhin Aleksandr, Pronina Anna, Boev Andrey, Gorbunov Stas
Журнал: Journal of Electrical Bioimpedance
Том: 13
Номер: 1
Год издания: 2022
Издательство: Sciendo
Местоположение издательства: Warsaw, Poland
Первая страница: 31
Последняя страница: 38
DOI: 10.2478/joeb-2022-0006
Аннотация: Fat-free mass (FFM) estimation has dramatic importance for body composition evaluation, often providing a basis for treatment of obesity and muscular dystrophy. However, current methods of FFM estimation have several drawbacks, usually related to either cost-effectiveness and equipment size (dual-energy X-ray absorptiometry (DEXA) scan) or model limitations. In this study, we present and validate a new FFM estimation model based on hand-to-hand bioimpedance analysis (BIA) and arm volume. Forty-two participants underwent a full-body DEXA scan, a series of anthropometric measurements, and upper-body BIA measurements with the custom-designed wearable wrist-worn impedance meter. A new two truncated cones (TTC) model was trained on DEXA data and achieved the best performance metrics of 0.886 ± 0.051 r2, 0.052 ± 0.009 % mean average error, and 6.884 ± 1.283 kg maximal residual error in FFM estimation. The model further demonstrated its effectiveness in Bland-Altman comparisons with the skinfold thickness-based FFM estimation method, achieving the least mean bias (0.007 kg). The novel TTC model can provide an alternative to full-body BIA measurements, demonstrating an accurate FFM estimation independently of population variables.
Добавил в систему: Пронина Анна Сергеевна

	ИСТИНА	Войти в систему Регистрация
	ИСТИНА ЦЭМИ РАН
	Главная Поиск Статистика О проекте Помощь

ИСТИНА

ИСТИНА ЦЭМИ РАН

Validation of non-empirical fat-free mass estimation model for a wrist-worn deviceстатья

Прикрепленные файлы