Exponential Pixelating Integral transform with dual fractal features for enhanced chest X-ray abnormality detection.

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Radiology

Background:

• Respiratory disorders pose a significant global health threat, with increased fatalities from novel coronavirus infections.
• Chest radiography is a vital, cost-effective tool for diagnosing respiratory conditions.
• Detecting abnormalities in chest X-rays is challenging due to low contrast, noise, and overlapping structures.

Purpose of the Study:

• To develop a novel analytical model for the automatic detection of respiratory infections in chest X-rays.
• To address the limitations of manual interpretation, including low contrast and noise.
• To improve the accuracy and efficiency of diagnosing diverse respiratory disorders.

Main Methods:

• Introduced the Exponential Pixelating Integral (EPI) model to enhance pixel intensities and overcome low-contrast issues.
• Applied polar transformation and fractal geometries (Mandelbrot, Julia) for feature extraction.
• Utilized non-parametric multivariate adaptive regression splines for classification, creating an ensemble model.

Main Results:

• The proposed Exponential Pixelating Integral framework achieved high classification accuracy (98.46-99.45%) and F1 scores (96.53-98.10%).
• Demonstrated superior performance compared to existing methods on large medical benchmarked datasets.
• The system proved to be a precise and interpretable automated diagnostic tool.

Conclusions:

• The novel Exponential Pixelating Integral model offers a highly accurate and efficient automated solution for detecting respiratory infections in chest X-rays.
• This approach has the potential to significantly aid radiologists in early diagnosis and intervention, improving patient outcomes.
• The integration of fractal geometries and advanced classification techniques provides robust feature distinction for medical image analysis.