Clinically Robust and Transparent AI-Assisted Medical Diagnostics: From Learning Dynamics to Real-World Deployment

Dear Colleagues,

In recent years, there has been a substantial increase in machine-learning research focused on medical diagnostics, with numerous studies reporting high accuracies on benchmark datasets. However, many of these studies do not sufficiently report the learning dynamics of model training or address critical issues such as overfitting, data leakage or the absence of external validation. As a result, models that demonstrate strong performances in controlled settings frequently fail to generalize to real-world clinical environments, often due to methodological shortcomings or insufficient transparency in the training and evaluation processes.

This Special Issue aims to foreground clinical robustness and transparency in AI-assisted medical diagnostics. We invite work that not only advance accuracy but also systematically demonstrates learning dynamics, avoids overfitting and data leakage and provides evidence for real-world clinical applicability. The issue will serve as a venue for studies that set higher methodological standards and enable reproducibility, transparency and translation into clinical workflows.

We invite contributions on (but not limited to) the following topics:

• Learning dynamics of ML models in medical diagnostics: Reporting, visualization and interpretation.
• Strategies to avoid overfitting and data leakage in medical AI studies.
• External validation and benchmarking of AI diagnostic models.
• Explainability and transparency: Methods and reporting standards.
• Generalizability and robustness across diverse patient populations and health systems.
• Clinical deployment: Workflow integration, regulatory and ethical considerations.
• Dataset curation, annotation and sharing for reproducible research.
• Interdisciplinary approaches combining clinical expertise and AI development.
• Security, privacy and data governance in AI-assisted diagnostics.
• Benchmarking and open challenges: Call for standardized datasets and tasks.

Guiding questions for prospective authors:

• How are the learning dynamics of your model reported and interpreted?
• What steps were taken to ensure generalizability and prevent overfitting or data leakage?
• How does your work contribute to transparency and reproducibility?
• What is the path to clinical deployment, and what barriers remain?

Article Types and Expectations

We welcome:

• Original empirical studies with rigorous methodological reporting.
• Systematic reviews and meta-analyses focused on methodological quality.
• Methodological and technical contributions on evaluation, transparency or deployment.
• Benchmark datasets and open challenges for the community.
• Perspective pieces discussing standards, governance or clinical translation.
• Conceptual frameworks integrating clinical and technical perspectives.

All submissions should explicitly address transparency, reproducibility and clinical relevance. We expect clear reporting of training processes, validation strategies and limitations.

Dr. Milan Toma
Prof. Dr. Constantinos Pattichis
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the Special Issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 250 words) can be sent to the Editorial Office for assessment.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-anonymized peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Machine Learning and Knowledge Extraction is an international peer-reviewed open access monthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 1800 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.