Attention-Enhanced Inception Next-Based Hybrid Deep Learning Model For Lung Cancer Detection

Abstract

Due to late discovery and a lack of effective early diagnostic methods, it takes the lives of many people. Early diagnosis could save lots of people, but typical methods like CT scan analysis depend on the experience of radiologists, which cause variability and errors. This paper offers an AttentionEnhanced HDL model for automatic lung cancer detection system using the LIDC-IDRI dataset to solve above problems. This classification blocs CNN and care instruments to extract and concentration on the imperative fragments of lung cancer images. On the other hand, preprocessing methods like normalization, resizing, and data augmentation help the model to be well trained and robust. Performance metrics such as accuracy, precision, recall, and F1-score show that the model has achieved 95% accuracy, 94% precision, 96% recall, and 95% F1-score which is recovering than other traditional ML reproductions and standard CNN-based architectures. This paper shows that the method presented can improve the doctor's diagnosis performance and at the same time be useful to the doctors by making the lung cancer detection early and easy and it has also the potential to be used in a hospital set up.