3.2.5. Hematological Cancer

Recently, impedance spectrometers have been shown to generate all-inclusive labon-a-chip platforms to detect nucleus abnormalities. The paper, presented by Ferguson et al. [117], is a proof-of-concept study on the classification of cancerous cells using a biosensor that employs impedance-based spectroscopy to identify the type of cells based on the size of their nucleus. The biosensor consists of a microfluidic channel attached to a quartz substrate containing an ultra-wideband waveguide. The cells passing through the PDMS channel are electrically trapped using a dielectrophoretic signal, and electrical signals are collected using microwave spectroscopy. The authors used statistical elimination and feature selection techniques along with SVMs and RF algorithms to achieve a 96% accuracy on multi-class classification. The study demonstrates the potential of using machine learning in combination with microwave impedance spectroscopy for single-cell classification

based on the population nucleus size, which could have significant implications for cancer diagnosis and treatment.
