About | Contact Us | Register | Login
ProceedingsSeriesJournalsSearchEAI
Computer Science and Education in Computer Science. 19th EAI International Conference, CSECS 2023, Boston, MA, USA, June 28–29, 2023, Proceedings

Research Article

AI-Enabled Infrared Thermography: Machine Learning Approaches in Detecting Peripheral Arterial Disease

Cite
BibTeX Plain Text
  • @INPROCEEDINGS{10.1007/978-3-031-44668-9_12,
        author={Georgi Kostadinov},
        title={AI-Enabled Infrared Thermography: Machine Learning Approaches in Detecting Peripheral Arterial Disease},
        proceedings={Computer Science and Education in Computer Science. 19th EAI International Conference, CSECS 2023, Boston, MA, USA, June 28--29, 2023, Proceedings},
        proceedings_a={CSECS},
        year={2023},
        month={10},
        keywords={Peripheral Arterial Disease Machine Learning Thermal Imaging Predictive Models XGBoost LightGBM},
        doi={10.1007/978-3-031-44668-9_12}
    }
    
  • Georgi Kostadinov
    Year: 2023
    AI-Enabled Infrared Thermography: Machine Learning Approaches in Detecting Peripheral Arterial Disease
    CSECS
    Springer
    DOI: 10.1007/978-3-031-44668-9_12
Georgi Kostadinov1,*
  • 1: New Bulgarian University, 21 Montevideo str
*Contact email: georgi.kostadinov@kelvin.health

Abstract

Peripheral Arterial Disease (PAD) is a common circulatory problem that, if undetected or untreated, can lead to severe health consequences, including amputation. This study presents a novel approach to PAD detection using thermal data collected via a mobile thermal camera, processed, and analysed through various machine learning algorithms. The investigation focused on six machine learning models: Linear Regression, Decision Trees, Random Forest, Neural Network, XGBoost, and LightGBM, and their ability to predict the presence of PAD based on thermal features extracted from different angiosomes of the legs. Each model was trained and validated on a dataset consisting of thermal data from 42 patients, annotated with PAD status based on angiography diagnostics. The performance of each model was evaluated using eight metrics, including accuracy, sensitivity, and specificity. The results indicate that ensemble methods, particularly XGBoost and LightGBM, outperformed the other models with an accuracy of 96.8%. This research demonstrates the potential of thermal imaging coupled with machine learning for the detection of PAD, offering a non-invasive, accessible, and cost-effective diagnostic tool.

Keywords
Peripheral Arterial Disease Machine Learning Thermal Imaging Predictive Models XGBoost LightGBM
Published
2023-10-11
Appears in
SpringerLink
http://dx.doi.org/10.1007/978-3-031-44668-9_12
Copyright © 2023–2025 ICST
EBSCOProQuestDBLPDOAJPortico
EAI Logo

About EAI

  • Who We Are
  • Leadership
  • Research Areas
  • Partners
  • Media Center

Community

  • Membership
  • Conference
  • Recognition
  • Sponsor Us

Publish with EAI

  • Publishing
  • Journals
  • Proceedings
  • Books
  • EUDL