Probabilistic Image Analysis for Chest X-Ray Classification: Integrating Density Functions as Descriptive Features
DOI:
https://doi.org/10.22401/Keywords:
Chest X-ray , GEV , CNN , Image ClassificationAbstract
In this study, we present a technique for chest X-ray image classification by integrating Generalized Extreme Value (GEV) probability density functions (PDFs) with Convolutional Neural Networks (CNNs). The traditional method is related to our proposed method, and the results show significant enhancements in key performance metrics. The proposed method reaches an accuracy of 91.95%, higher than the traditional method's 89.85%, typically enhanced capabilities in accurate classification. Precision is higher in the proposed method at 91.13%, emphasizing its proficiency in correctly identifying positive cases. Specificity is improved in the proposed method (67.22%) compared to the traditional method (61.20%), which represents a lower risk of false positives. The F1-score of 95.01% in the proposed method indicates a consistent balance between precision and recall, underlining its effectiveness in minimizing both false positives and false negatives. These findings suggest that the integration of GEV PDFs and CNNs holds potential for progressing chest X-ray image classification accuracy and reliability, with possible suggestions for improving diagnostic procedures in clinical settings.
References
[1] Huang, SC., Pareek, A., Seyyedi, S. Imon B., and Matthew P.; "Fusion of Medical Imaging and Electronic Health Records Using Deep Learning: A Systematic Review and Implementation Guidelines". Nat. Digit. Med., 3(1), 45-60, 2020.
[2] Svu Ha, Ct Nguyen, Hn Phan, NM Chung, and Phuong HH.; "CDN-MEDAL: Two-stage Density and Difference Approximation Framework for Motion Analysis". arxiv.org, 2021.
[3] H Huynh-Van, T Le-Hoang, and T Vo-Van; "Classifying for Images Based on The Extracted Probability Density Function and the Quasi Bayesian Method". Comput Stat 39, 2677–2701, 2024.
[4] Kotei, E., and Thirunavukarasu, R.; "A Comprehensive Review on Advancement in Deep Learning Techniques for Automatic Detection of Tuberculosis From Chest X-Ray Images". Arch Comput Methods Eng 31, 455–474, 2024.
[5] Wang, B., Pan, H., and A Aboah, Z; "Gazegnn: A Gaze-Guided Graph Neural Network for Chest X-Ray Classification". IEEE Workshop on Applications of Computer Vision (WACV), 2024.
[6] Jeong, J., Tian, K., and Li, S.; "Multimodal Image-Text Matching Improves Retrieval-Based Chest X-Ray Report Generation". Comput Lang., 2, 978–990, 2024.
[7] Maniruzzaman, M., Sami, A., Hoque, R., and Mandal, P.,; "Pneumonia Prediction Using Deep Learning in Chest X-Ray Images". Int. J. Sci. Res. Arch., 12, 767–773, 2024.
[8] Asnake, N.W., Salau, A.O., and Ayalew, A.M.; "X-Ray Image-Based Pneumonia Detection and Classification Using Deep Learning". Multimed Tools Appl, 83, 60789–60807, 2024.
[9] Aseel, M.E., and Mohammad, K.L.; "Using Kernel Functions to Estimate the Probability Density Function for Segmentation Images". J. Stat. Manag. Syst., 27, 3, 645–654, 2024.
[10] Stotts, L.B., and Andrews, L.C.; "Probability Density Function Models for Adaptive Optical Systems Operating in Turbulence". Opt. Eng. - OE, 63, 8, 1-10, 2024.
[11] Albahli S.; "A Deep Neural Network to Distinguish Covid-19 from Other Chest Diseases Using X-Ray Images". Curr Med Imaging, 17(1):109-119, 2021.
[12] Varon, E., Brun, A.L., Claessens, Y.E., Duval, E.; "Ct Features of Community-Acquired Pneumonia at the Emergency Department" Resp Medic and Research, 2022.
[13] Arai, K.; "Image Classification Method Considering Overlapping and Correlation Between Probability Density Functions of Features". Intell. Syst. Appl .,1(3), 24–45, 2024.
[14] El-Genidy, M.M.; and Hebeshy, E.A.; "An Accurate Method for Estimating the Parameters of the Generalized Extreme Value Distribution Using its Moments". Alfarama J. Basic Appl. Sci., 2024.
[15] Jiefeng, L., Siyuan, B., Ailing, Z., Can W., Bo P., Wentao L., and Cewu L.; "Human Pose Regression With Residual Log-Likelihood Estimation". CVF Int. Conf. Comput. Vis., 3, 11025-11034 2021.
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Manhal Elias Polus
This work is licensed under a Creative Commons Attribution 4.0 International License.
.jpg)
