Terahertz Time-Domain Spectroscopy of Blood Serum for Differentiation of Glioblastoma and Traumatic Brain Injury

Vrazhnov, D.A.; Ovchinnikova, D.A.; Kabanova, T.V.; Paulish, A.G.; Kistenev, Y.V.; Nikolaev, N.A.; Cherkasova, O.P.

Авторы: Vrazhnov Denis A., Ovchinnikova Daria A., Kabanova Tatiana V., Paulish Andrey G., Kistenev Yury V., Nikolaev Nazar A., Cherkasova Olga P.
Журнал: Applied Sciences
Том: 14
Номер: 7
Год издания: 2024
Издательство: MDPI
Местоположение издательства: Basel, Switzerland
Первая страница: 2872
Последняя страница: 2872
DOI: 10.3390/app14072872
Аннотация: The possibility of the differentiation of glioblastoma from traumatic brain injury throughblood serum analysis by terahertz time-domain spectroscopy and machine learning was studied usinga small animal model. Samples of a culture medium and a U87 human glioblastoma cell suspensionin the culture medium were injected into the subcortical brain structures of groups of mice referred toas the culture medium injection groups and glioblastoma groups, accordingly. Blood serum sampleswere collected in the first, second, and third weeks after the injection, and their terahertz transmissionspectra were measured. The injection caused acute inflammation in the brain during the first week, sothe culture medium injection group in the first week of the experiment corresponded to a traumaticbrain injury state. In the third week of the experiment, acute inflammation practically disappeared inthe culture medium injection groups. At the same time, the glioblastoma group subjected to a U87human glioblastoma cell injection had the largest tumor size. The THz spectra were analyzed usingtwo dimensionality reduction algorithms (principal component analysis and t-distributed StochasticNeighbor Embedding) and three classification algorithms (Support Vector Machine, Random Forest,and Extreme Gradient Boosting Machine). Constructed prediction data models were verified using10-fold cross-validation, the receiver operational characteristic curve, and a corresponding area underthe curve analysis. The proposed machine learning pipeline allowed for distinguishing the traumaticbrain injury group from the glioblastoma group with 95% sensitivity, 100% specificity, and 97%accuracy with the Extreme Gradient Boosting Machine. The most informative features for thesegroups’ differentiation were 0.37, 0.40, 0.55, 0.60, 0.70, and 0.90 THz. Thus, an analysis of mouseblood serum using terahertz time-domain spectroscopy and machine learning makes it possible todifferentiate glioblastoma from traumatic brain injury.
Добавил в систему: Шкуринов Александр Павлович

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

ИСТИНА

ИПМех РАН

Terahertz Time-Domain Spectroscopy of Blood Serum for Differentiation of Glioblastoma and Traumatic Brain Injuryстатья

Прикрепленные файлы