Automated Lungs Disease Detection Using Hybrid Deep Learning Technique

Abstract

This paper presents a current state of the art approach to the early detection and classification of lung diseases using medical images based on the use of deep learning models – Convolutional Neural Networks (CNNs). Diagnosis of lung diseases is difficult in their early stage due to asymptomatic presentation of the diseases and this explains the importance of advanced diagnostic tests. One of deep learning’s benefits is that it can accurately recognise complex patterns in medical images to improve diagnoses and spot potential issues early. The proposed method can be divided into Image Acquisition, Preprocessing, Training, and Classification stages using deep learning algorithms like CNN. Transfer learning is utilized to tailor pre-trained models to the task of lung disease recognition. Besides, the study suggests a classification system based on image types, features, data augmentation, and deep learning algorithm types describing the state of the art in this field. Data preparation, model building, training, fine-tuning, and model evaluation are steps of the training process that seeks to improve model efficiency for predicting lung diseases. In conclusion, the application of deep learning methods in research demonstrates practical benefits for diagnosing and treating lung disease at earlier stages and better results in treatments.