Computed Tomography-Based Lung Cancer Revelation and Collocation Utilizing Optimized Anti-Interference Dynamic Integral Neural Networks for Superior Diagnostic Accuracy and Quality Assessment

Lung cancer (LC) is one of the common causes of death worldwide. Early detection is a must to protect people from this disease. Nowadays, computed tomography (CT) scan is the main imaging technique for LC diagnosis. But, manual CT scan analysis takes a lot of time, does not provide enough accuracy and is prone to mistakes. Therefore, a Computed Tomography Based Lung Cancer Revelation and Collocation using Optimized Anti-Interference Dynamic Integral Neural Networks for Superior Diagnostic Accuracy and Quality Assessment (CTLCC-ADINN-DAQ) is proposed in this paper. Here, the input images are collected from the IQ-OTHNCCD LC dataset and undergo pre-processing using the Modified Hamilton Filter (MHF) to reduce noise. Subsequently, the pre-processed images are segmented utilizing the Deep fuzzy variable C-means clustering (DFCC) method to identify the ROI in lung images. The segmented images are fed into the Anti-Interference Dynamic Integral Neural Network (ADINN) for classifying LC as benign, malignant and normal. The ADINN framework is optimized using the Hunger Game Search Optimization Algorithm (HGSOA), enhancing its accuracy in LC classification. Performance evaluation metrics like accuracy, recall, precision, F1-score, and computational time are utilized to evaluate the CTLCC-ADINN-DAQ method. The proposed technique attains 22.37%, 23.35%, and 21.45% higher accuracy when compared with the existing techniques: Deep Learning Approach for Early Stage Lung Cancer Detection (DLA-ES-LCD), Transfer Learning along GoogleNet for Detection of Lung Cancer (TL-GND-LC), and Evaluation of SVM Performance in the Detection of Lung Cancer in Marked CT Scan Dataset (SVM-LC-CTSD), respectively.