Processed Datasets for Segmentation Mask Guided Hybrid CNN Vision Transformer Framework

This Zenodo repository contains processed representatives samples of the dasets generated during the study entitled “Segmentation Mask Guided Hybrid CNN Vision Transformer Framework for Diagnosing Communicable Lung Diseases.” The datasets were created to support deep learning based analysis of chest X-ray images for the automated diagnosis of communicable lung diseases. The repository includes representative samples of processed data corresponding to all stages of the proposed pipeline, including lungs segmented chest X-ray images, enhanced image datasets, augmented image datasets, and predicted lung mask datasets. In addition, the numerical data underlying the reported experimental results such as epoch-wise training metrics, evaluation metrics and confusion matrix values are provided as CSV files to ensure transparency and reproducibility. The processed datasets were derived from publicly available, anonymized chest X-ray repositories, including the COVID-19 Radiography Database and the Montgomery County and Shenzhen Tuberculosis chest X-ray datasets released by the U.S. National Library of Medicine. No personally identifiable information is included in this repository. All data are released under the Creative Commons Attribution 4.0 International (CC BY 4.0) license and are intended to support open research, reuse, and further development of computer-aided diagnostic methods for medical imaging.