Infection Diagnosis Etiology in Febrile Pediatric Patients Using Infrared Spectroscopy of Human Peripheral Blood and Machine Learning

by Mr. Adam H. Agbaria

Dept. Of Physics, Ben-Gurion University
at Biological and soft-matter physics

Thu, 23 Apr 2020, 12:00
ZOOM, Meeting ID: 515 554 168 Password: 772189

Abstract

Rapid diagnosis of the etiology of infection is highly important for an effective treatment of the infected patients. The objective of this study was to evaluate the potential of the mid infrared spectroscopic method for rapid and reliable identiļ¬cation of bacterial and viral infections based on peripheral blood samples. Fourier transform infrared (FTIR) spectroscopy has been found useful for distinguish between bacterial and viral etiology, based on an analysis of the blood components.
Four hundred and sixty events of febrile pediatric normal and oncology patients were evaluated in this study. White blood cells (WBC) and plasma were isolated from peripheral blood. WBC and plasma from patients with confirmed viral or bacterial infections were evaluated with FTIR spectroscopy. FTIR spectra were analysed for biomarkers and classified machine leaning classifiers.
By employing the FTIR spectroscopy of feature extraction machine learning analysis in order to identify the infectious type, a sensitivity of ~88% and an accuracy of ~81% for an infection type diagnosis were achieved on pediatric normal patients and a sensitivity of ~94% and an accuracy of ~90% for an infection type diagnosis were achieved on pediatric oncology patients. The present preliminary study suggests that FTIR spectroscopy of WBCs is a rapid potentially feasible tool for the diagnosis of an infection etiology.

Join Zoom Meeting
https://zoom.us/j/515554168?pwd=S0gyTXc5eXV2QWJRbGtCMkRQNFVVQT09

Meeting ID: 515 554 168
Password: 772189

Created on 05-03-2020 by Granek, Rony (rgranek)
Updaded on 03-04-2020 by Granek, Rony (rgranek)