Privacy Enhanced Lung Cancer Detection by Hi-SN-Net-SSOA Classifier with Attribute Based Encryption Technique

For individuals Applying pre-processing using multilayer brightness-preserving method to the acquired photos allows for a thorough examination of each pixel, the removal of noise, and an development in the excellence of the lung image. Features are extracted using gray-level cooccurrence matrix (GLCM) from the noise-removed lung CT picture. Then features are optimally selected using krill herd optimization algorithm (KHOA). Additionally, a hierarchical spiking neural network (Hi-SN-Net) technique is used for classification with hyper parameter tuning based on shark smell optimization algorithm (SSOA). Using CT scans, the proposed Hi-SN-Net-SSOA approach seeks to identify and classify the presence of illnesses. Finally, in order to maintain privacy in our suggested architecture, this study suggests an attribute-based encryption (ABE) technique based on cloud. The proposed Hi-SN-Net-SSOA technique under discussion improves the accuracy up to 99%, precision, sensitivity, and specificity of lung cancer prediction.