Thyroid Cancer Detection Using Artificial Neural Network and Photonic Sensor

Thyroid cancer is a grave disease nowadays with high mortality rate, a large number of people are diagnosed with thyroid cancer. In this work a method for detection of thyroid cancer is presented by using Photonic Crystal Waveguide (PCW) and Artificial Neural Network (ANN). In this proposed method, linear waveguide Photonic Crystal (PhC) structure 27 X 25, 2D square lattice is used. The square lattice has silicon rods engraved in and central blemish with air. Malignant thyroid cells exhibit the distinct wavelength deviation of 0.63351561 µm and the corresponding sensitivity of 12.6703122 X 10–6 RIU for the proposed sensor, which displays the accuracy in sensing of normal cells and cancerous cells. In the second stage of this method the simulated data is fed to an artificial neural network (ANN) model. The developed ANN model has an accuracy score 88 %, specificity 84%, precision 85 %, sensitivity 92 % and F1 score 88%.