Terahertz Time-Domain Spectroscopy of Glioma Patient Blood Plasma: Diagnosis and Treatment

Gliomas, one of the most severe malignant tumors central nervous system, have a high mortality rate and an increased risk recurrence. Therefore, early glioma diagnosis control treatment great significance. The blood plasma samples patients, patients with skull craniectomy defects, healthy donors were studied using terahertz time-domain spectroscopy (THz-TDS). An analysis experimental THz data was performed by machine learning (ML). ML pipeline included (i) spectra smoothing Savitzky–Golay filter, (ii) dimension reduction principal component t-distribution stochastic neighborhood embedding methods; (iii) separability analyzed Support Vector Machine (SVM), Random Forest (RF), Extreme Gradient Boosting (XGBoost). models’ performance evaluated k-fold cross validation technique ROC-AUC, sensitivity, specificity metrics. It shown that tree-based ensemble methods work more accurately than SVM. RF XGBoost provided better differentiation group from defects. THz-TDS combined to make it possible separate before after tumor removal surgery (AUC = 0.92). Thus, applicability for monitoring has been shown.