Epanechnikov Nonparametric Kernel Density Estimation Based Feature-Learning in Respiratory Disease Chest X-Ray Images

Verónica Marsico; Antonio Quintero-Rincón; Hadj Batatia

doi:10.1007/978-3-032-06336-6_3

Epanechnikov Nonparametric Kernel Density Estimation Based Feature-Learning in Respiratory Disease Chest X-Ray Images

Verónica Marsico^*
, Antonio Quintero-Rincón
, Hadj Batatia

^*Corresponding author for this work

School of Mathematical & Computer Sciences

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

This study presents a novel method for diagnosing respiratory diseases using image data. It combines Epanechnikov’s non-parametric kernel density estimation (EKDE) with a bimodal logistic regression classifier in a statistical-model-based learning scheme. EKDE’s flexibility in modeling data distributions without assuming specific shapes and its adaptability to pixel intensity variations make it valuable for extracting key features from medical images. The method was tested on 13808 randomly selected chest X-rays from the COVID-19 Radiography Dataset, achieved an accuracy of 70.14%, a sensitivity of 59.26%, and a specificity of 74.18%, demonstrating moderate performance in detecting respiratory disease while showing room for improvement in sensitivity. While clinical expertise remains essential for further refining the model, this study highlights the potential of EKDE-based approaches to enhance diagnostic accuracy and reliability in medical imaging.

Original language	English
Title of host publication	Cloud Computing, Big Data and Emerging Topics. JCC-BD&ET 2025
Editors	Marcelo Naiouf, Laura De Giusti, Franco Chichizola, Leandro Libutti
Publisher	Springer
Pages	31-45
Number of pages	15
ISBN (Electronic)	9783032063366
ISBN (Print)	9783032063359
DOIs	https://doi.org/10.1007/978-3-032-06336-6_3
Publication status	Published - 2026
Event	13th Conference on Cloud Computing, Big Data and Emerging Topics 2025 - La Plata, Argentina Duration: 24 Jun 2025 → 26 Jun 2025

Publication series

Name	Communications in Computer and Information Science
Volume	2649
ISSN (Print)	1865-0929
ISSN (Electronic)	1865-0937

Conference

Conference	13th Conference on Cloud Computing, Big Data and Emerging Topics 2025
Abbreviated title	JCC-BD and ET 2025
Country/Territory	Argentina
City	La Plata
Period	24/06/25 → 26/06/25

Keywords

Bimodal logistic regression
Epanechnikov
kernel density estimation (KDE)

ASJC Scopus subject areas

General Computer Science
General Mathematics

Access to Document

10.1007/978-3-032-06336-6_3

Cite this

Marsico, V., Quintero-Rincón, A., & Batatia, H. (2026). Epanechnikov Nonparametric Kernel Density Estimation Based Feature-Learning in Respiratory Disease Chest X-Ray Images. In M. Naiouf, L. De Giusti, F. Chichizola, & L. Libutti (Eds.), Cloud Computing, Big Data and Emerging Topics. JCC-BD&ET 2025 (pp. 31-45). (Communications in Computer and Information Science; Vol. 2649). Springer. https://doi.org/10.1007/978-3-032-06336-6_3

Marsico, Verónica ; Quintero-Rincón, Antonio ; Batatia, Hadj. / Epanechnikov Nonparametric Kernel Density Estimation Based Feature-Learning in Respiratory Disease Chest X-Ray Images. Cloud Computing, Big Data and Emerging Topics. JCC-BD&ET 2025. editor / Marcelo Naiouf ; Laura De Giusti ; Franco Chichizola ; Leandro Libutti. Springer, 2026. pp. 31-45 (Communications in Computer and Information Science).

@inproceedings{8f83f1b7b1994302bc915dca33f68bcb,

title = "Epanechnikov Nonparametric Kernel Density Estimation Based Feature-Learning in Respiratory Disease Chest X-Ray Images",

abstract = "This study presents a novel method for diagnosing respiratory diseases using image data. It combines Epanechnikov{\textquoteright}s non-parametric kernel density estimation (EKDE) with a bimodal logistic regression classifier in a statistical-model-based learning scheme. EKDE{\textquoteright}s flexibility in modeling data distributions without assuming specific shapes and its adaptability to pixel intensity variations make it valuable for extracting key features from medical images. The method was tested on 13808 randomly selected chest X-rays from the COVID-19 Radiography Dataset, achieved an accuracy of 70.14\%, a sensitivity of 59.26\%, and a specificity of 74.18\%, demonstrating moderate performance in detecting respiratory disease while showing room for improvement in sensitivity. While clinical expertise remains essential for further refining the model, this study highlights the potential of EKDE-based approaches to enhance diagnostic accuracy and reliability in medical imaging.",

keywords = "Bimodal logistic regression, Epanechnikov, kernel density estimation (KDE)",

author = "Ver{\'o}nica Marsico and Antonio Quintero-Rinc{\'o}n and Hadj Batatia",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.; 13th Conference on Cloud Computing, Big Data and Emerging Topics 2025, JCC-BD and ET 2025 ; Conference date: 24-06-2025 Through 26-06-2025",

year = "2026",

doi = "10.1007/978-3-032-06336-6\_3",

language = "English",

isbn = "9783032063359",

series = "Communications in Computer and Information Science",

publisher = "Springer",

pages = "31--45",

editor = "Marcelo Naiouf and \{De Giusti\}, Laura and Franco Chichizola and Leandro Libutti",

booktitle = "Cloud Computing, Big Data and Emerging Topics. JCC-BD\&ET 2025",

}

Marsico, V, Quintero-Rincón, A & Batatia, H 2026, Epanechnikov Nonparametric Kernel Density Estimation Based Feature-Learning in Respiratory Disease Chest X-Ray Images. in M Naiouf, L De Giusti, F Chichizola & L Libutti (eds), Cloud Computing, Big Data and Emerging Topics. JCC-BD&ET 2025. Communications in Computer and Information Science, vol. 2649, Springer, pp. 31-45, 13th Conference on Cloud Computing, Big Data and Emerging Topics 2025, La Plata, Argentina, 24/06/25. https://doi.org/10.1007/978-3-032-06336-6_3

Epanechnikov Nonparametric Kernel Density Estimation Based Feature-Learning in Respiratory Disease Chest X-Ray Images. / Marsico, Verónica; Quintero-Rincón, Antonio; Batatia, Hadj.
Cloud Computing, Big Data and Emerging Topics. JCC-BD&ET 2025. ed. / Marcelo Naiouf; Laura De Giusti; Franco Chichizola; Leandro Libutti. Springer, 2026. p. 31-45 (Communications in Computer and Information Science; Vol. 2649).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Epanechnikov Nonparametric Kernel Density Estimation Based Feature-Learning in Respiratory Disease Chest X-Ray Images

AU - Marsico, Verónica

AU - Quintero-Rincón, Antonio

AU - Batatia, Hadj

N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.

PY - 2026

Y1 - 2026

N2 - This study presents a novel method for diagnosing respiratory diseases using image data. It combines Epanechnikov’s non-parametric kernel density estimation (EKDE) with a bimodal logistic regression classifier in a statistical-model-based learning scheme. EKDE’s flexibility in modeling data distributions without assuming specific shapes and its adaptability to pixel intensity variations make it valuable for extracting key features from medical images. The method was tested on 13808 randomly selected chest X-rays from the COVID-19 Radiography Dataset, achieved an accuracy of 70.14%, a sensitivity of 59.26%, and a specificity of 74.18%, demonstrating moderate performance in detecting respiratory disease while showing room for improvement in sensitivity. While clinical expertise remains essential for further refining the model, this study highlights the potential of EKDE-based approaches to enhance diagnostic accuracy and reliability in medical imaging.

AB - This study presents a novel method for diagnosing respiratory diseases using image data. It combines Epanechnikov’s non-parametric kernel density estimation (EKDE) with a bimodal logistic regression classifier in a statistical-model-based learning scheme. EKDE’s flexibility in modeling data distributions without assuming specific shapes and its adaptability to pixel intensity variations make it valuable for extracting key features from medical images. The method was tested on 13808 randomly selected chest X-rays from the COVID-19 Radiography Dataset, achieved an accuracy of 70.14%, a sensitivity of 59.26%, and a specificity of 74.18%, demonstrating moderate performance in detecting respiratory disease while showing room for improvement in sensitivity. While clinical expertise remains essential for further refining the model, this study highlights the potential of EKDE-based approaches to enhance diagnostic accuracy and reliability in medical imaging.

KW - Bimodal logistic regression

KW - Epanechnikov

KW - kernel density estimation (KDE)

UR - https://www.scopus.com/pages/publications/105020690313

U2 - 10.1007/978-3-032-06336-6_3

DO - 10.1007/978-3-032-06336-6_3

M3 - Conference contribution

AN - SCOPUS:105020690313

SN - 9783032063359

T3 - Communications in Computer and Information Science

SP - 31

EP - 45

BT - Cloud Computing, Big Data and Emerging Topics. JCC-BD&ET 2025

A2 - Naiouf, Marcelo

A2 - De Giusti, Laura

A2 - Chichizola, Franco

A2 - Libutti, Leandro

PB - Springer

T2 - 13th Conference on Cloud Computing, Big Data and Emerging Topics 2025

Y2 - 24 June 2025 through 26 June 2025

ER -

Marsico V, Quintero-Rincón A, Batatia H. Epanechnikov Nonparametric Kernel Density Estimation Based Feature-Learning in Respiratory Disease Chest X-Ray Images. In Naiouf M, De Giusti L, Chichizola F, Libutti L, editors, Cloud Computing, Big Data and Emerging Topics. JCC-BD&ET 2025. Springer. 2026. p. 31-45. (Communications in Computer and Information Science). Epub 2025 Oct 17. doi: 10.1007/978-3-032-06336-6_3