Terahertz (THz) technology is gaining attention for its potential in bio-detection, but there's limited research on its use in mixtures.
Traditional 1D spectral analysis is not sensitive enough due to overlapping and distortions, so this study used the Gramian angular field (GAF) method to create 2D images from THz spectra.
The new approach, which combines histogram of oriented gradients (HOGs) and gray level histograms (GLHs) with a support vector regression (SVR) model, showed improved accuracy and stability over traditional methods, making it more effective for analyzing mixed systems.
Liver cancer is challenging to detect early due to its high malignancy and subtle symptoms, making research into early detection methods crucial.
This study introduces a biometric detection technique using terahertz time-domain spectroscopy to distinguish between normal and cancerous liver cells by analyzing five key characteristics.
The proposed method achieved an accuracy of 91.6% in classifying human hepatoma and normal cell lines, offering a promising approach for evaluating living cells in a liquid environment.