Illuminating Pneumonia Diagnosis and Severity Grading in Chest X-rays with EfficientNetB0

Deep Learning

Pneumonia, an infection inflaming the air sacs in the lungs, poses a significant health concern. Accurate and efficient diagnosis is crucial for timely treatment. Detecting pneumonia involves multiple approaches. Chest X-rays and CT scans provide initial assessments, while pulse oximetry assesses blood oxygen levels. Laboratory tests, including sputum analysis and blood tests for infection markers, further aid diagnosis. Due to its inherent advantages in accessibility, speed, and cost-effectiveness, chest X-ray is our chosen method for initial pneumonia detection in this project. Our model leverages EfficientNetB0, a deep learning architecture known for striking a balance between high accuracy and computational efficiency, to automate pneumonia detection from chest X-ray images. The study utilizes EfficientNetB0 for automating X-ray image analysis, facilitating fast and accurate detection of pneumonia cases, alongside introducing severity stages such as low, mild, and high. Evaluation of the model's performance, based on metrics like accuracy, precision, recall, and F1-score, demonstrates superiority over conventional diagnostic methods.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.