Design and analysis of a GaN-based 2D photonic crystal biosensor integrated with machine learning techniques for detection of skin diseases.

Photonic crystals are prevalent in the detection of assorted diseases and malignancies such as vitiligo and cutis laxa. A 2D photonic crystal utilizing GaN is demonstrated to detect skin diseases, highlighting its substantial relevance to the photonic sensing community. Various parameters analysed are quality factor, wavelength sensitivity, FWHM, figure of merit and detection limit. The analysis of sensor characteristics demonstrated that GaN is a highly suitable material for detecting vitiligo and cutis laxa. Topology of design is crucial to focus the light on to the sensing region. Opti FDTD tool was used for the design and simulation of the sensor. The photonic band gap was simulated and it was observed that it contained one band gap region. The design provided high transmission efficiency and sensitivity. The various skin abnormalities related to vitiligo and cutis laxa could be easily detected from the results. Machine learning models such as K-nearest neighbor, Random Forest, Support Vector Machine and Multi-Layer Perceptron were adopted to enhance the sensor system to classify the data with higher accuracy.