Page 53 - חוברת תזות הנדסה ירוקה 2023

Diagnosis of the etiology of infection and monitoring the efficacy of antibiotic therapy in patients with bacterial infections using infrared spectroscopy
Gal Cohen Logasi1 Galco2652@gmail.com
Prof. Ahmad Salman2, Prof.Mahmoud Huleihel3, Prof. Joseph Kapelushnik4
1, 2SCE-Sami Shamoon College of Engineering, 3Ben-Gurion University of the Negev, 4Soroka University Medical Center and Faculty of Health Sciences
Every year, many patients arrive at clinics and emergency rooms with infectious bacterial and viral diseases sharing similar symptoms. Patients with viral infections will usually recover after a few days, while patients with bacterial infections should be treated properly with antibiotics immediately to prevent complications of the disease. It is very important to monitor the effectiveness of antibiotic therapy to ensure adequate treatment and to limit the development of new antibiotic-resistant mutants. Unfortunately, antibiotics are not fully broken down in our bodies, and high percentages of their active substances are excreted from the body in their original form. A significant part of these medical waste materials ends up in the Earth’s soil and water. The serious damage may be done to marine creatures, and people who eat seafood may also be negatively affected by such antibiotic residue.
The main goals of this study are to evaluate the potential use of infrared spectroscopy to test white blood cells and serum extracted from the blood of patients from the Oncology Department: 1( in order to attain a rapid and objective diagnosis of the etiology of accessible and inaccessible infections )bacterial and viral( and 2( to monitor the effectiveness of the specific antibiotic treatment of patients with certain bacterial infections and confirm the efficacy of the treatment and prevent any unnecessary use of antibiotics or the prescription of excessive dosages. We hypothesize that the immune system reacts differently to bacterial and viral infections, and it is possible to monitor these changes by measuring various blood components, such as WBCs and serum by means of mid-infrared spectroscopy. We applied machine learning methods to diagnose the etiology of infections )bacterial or viral( based on the measured spectra from our test samples.
Mid-infrared spectroscopy is known to be a fast, accurate, and sensitive method for detecting even the smallest biochemical changes associated with the development of abnormalities, including cancer, antibiotic resistance, and infections. The energy of the absorbed wavelengths corresponds to the vibrational modes of the functional groups of the molecules that compose the measured samples. Therefore, the absorption spectrum for each tested sample serves as its biochemical fingerprint. In this study, first, Fourier transform infrared )FTIR( spectroscopy was used to measure the WBCs and serum of oncology patients. Then, the spectra of these samples were analyzed, using chemometric and machine learning methods,
53