[email protected] +91 9840521421 Scholarly Open Access, Peer-reviewed, Refereed Journal e-ISSN 2583-5300 Since 2022
  1. Home
  2. Archives
  3. 2026
  4. Vol. 5 Issue 3
  5. Article
Review Article

Med trust: A Trust-Aware and Explainable Artificial Intelligence Framework for Reliable Multi-Disease Chest X-Ray Diagnosis

Open Access Peer Reviewed Crossref DOI

Authors

  • Dr. P. R. Gavali1
  • Om C. Joshi2
  • Snehal D. Kumbhar3
  • Sanika S Kumbhar4
  • Vishal G. Biradar5
  • Atharv A. Lite6

1 Assistant Professor, Department of Computer Science and Engineering, D. Y. Patil College of Engineering and Technology, Kolhapur, India. 2 3 4 5 6 Department of Computer Science and Engineering, D. Y. Patil College of Engineering and Technology, Kolhapur, India.

Article Information

Published Online
September-December 2026
Pages
892-898
Article Type
Review Article
Check for updates
check update

Abstract

View PDF

Although AI has shown encouraging findings for automated chest X-ray diagnosis, precision on its own isn't enough for clinical use, in real-world scenarios. MedTrust, a trust-aware and explainable framework that's checking how consistent each chest X-ray prediction is mainly is being provided by the work, on the whole. The framework is combining image quality assessment, multi-disease grouping, predictive uncertainty estimation, out-of-distribution (OOD) detection, temperature-scaled probability calibration, similarity-based retrieval, and explainability, broadly speaking. Together these elements are outlining whether an image is fit for use, how sure the model is, whether the case resembles the data the model was devised on, and what proof supports the finding, to a significant degree. The signals are together brought by a trust-aware decision layer and marks predictions that merit extra review, in actual application. Performance is measured with standard classification metrics and the space, area under the receiver operating characteristic curve (AUROC), in many respects. Calibration is assessed with negative log-likelihood (NLL), the Brier score, and expected calibration error (ECE), while reliability diagrams show how closely stated confidence matches real outcomes. The behavior of uncertainty and OOD signals is also inspected to spot inputs that are usually unreliable or unfamiliar, in many respects. By treating accuracy, confidence reliability, and interpretability together, a more careful and informed use of AI in chest X-ray interpretation is being supported by MedTrust, in real-world scenarios.

APC from ₹1200 + 18% GST

Peer-reviewed open access₹1200 Indian (₹1400 with DOI)$80 USD internationalpaid after acceptance Peer-reviewed open access₹1200 Indian$80 USD international ₹1200 Indian · $80 intl

  • CrossRef DOI
  • ISSN 2583-5300
  • Open Access
  • Peer Reviewed
  • Plagiarism Checked
  • Google Scholar
Contact Us Submit Paper