A novel photonic crystal sensor has achieved high sensitivity and over 91% classification accuracy in cervical cancer screening, offering a non-invasive approach for early detection. This optical sensor, designed as a grid of microscopic silicon rods, guides light through cervical tissue. Variations in light behavior reveal differences in healthy, infected, or cancerous tissue, enabling label-free differentiation without dyes.
The sensor precisely detects minute optical changes caused by tissue structure and composition. Researchers tested its performance across a range of temperatures, finding reliable function from 10°C to 60°C, with optimal accuracy at 25°C. This robustness suggests potential for effective operation in diverse clinical settings.
An artificial neural network was trained on the sensor's optical features, boosting tissue classification accuracy to over 91%. This advancement supports its potential for automated screening or decision support systems.
These findings indicate the photonic crystal sensor could be a rapid, accurate, and non-invasive tool for early cervical cancer detection, with significant practical advantages for medical applications.