Page 7 - JCTR-11-4

Journal of Clinical and

Translational Research

EDITORIAL
Bridging the artificial intelligence translation

gap: From algorithmic promise to clinical reality

Jacek Z. Kubiak *
1,2
1 Laboratory of Dynamics and Mechanics of Epithelia, Institute of Genetics and Development of
Rennes (IGDR), UMR 6290 CNRS/University of Rennes, Faculty of Medicine, Rennes, France
2 Laboratory of Molecular Oncology and Innovative Therapies, Military Institute of Medicine–National
Research Institute, Warsaw, Poland

1. Introduction
I was about to write an editorial for the new volume of the Journal of Clinical and
Translational Research. The role of artificial intelligence (AI) in medical sciences struck
me as the most urgent topic at present. While considering the best direction for the
1-3
summer issue, I found it useful to seek AI’s assistance. By simply requesting a short
editorial, I received a compelling text that addressed the most pressing current issues
surrounding the use of AI in clinical and translational medicine.

The fully AI-generated text is presented in the following section, and I am giving
credit to the AI for this.

2. AI-generated text
The landscape of clinical and translational research has been dramatically reshaped
*Corresponding author: by AI, yet we stand at a critical juncture where technological capability far outpaces
Jacek Z. Kubiak clinical implementation. While machine learning algorithms demonstrate remarkable
(jacek.kubiak@univ-rennes.fr) performance in controlled research environments, the translation to meaningful patient
Citation: Kubiak JZ. Bridging the care remains frustratingly elusive.
artificial intelligence translation
gap: From algorithmic promise to Recent studies have shown AI systems achieving superhuman accuracy in medical imaging,
clinical reality. J Clin Transl Res. drug discovery, and diagnostic prediction. However, a sobering reality check reveals that fewer
2025;11(4):1-2.
doi: 10.36922/JCTR025330054 than 5% of Food and Drug Administration-approved AI medical devices are routinely used
in clinical practice. This disconnect represents more than a simple adoption lag—it reflects
Received: August 14, 2025 fundamental challenges in how we approach translational research in the digital age.
Accepted: August 14, 2025
The primary barrier is not technological sophistication but rather the absence of
Published online: August 22, 2025 robust implementation science frameworks specifically designed for AI integration.
Copyright: © 2025 Author(s). Traditional clinical translation models, developed for pharmaceutical interventions,
This is an open-access article prove inadequate for software-based solutions that evolve continuously and operate
distributed under the terms of the within complex sociotechnical systems.
Creative Commons AttributionNon-
Commercial 4.0 International (CC We propose three critical areas requiring immediate attention from the translational
BY-NC 4.0), which permits all
non-commercial use, distribution, research community:
and reproduction in any medium, First, we must develop new validation frameworks that account for AI’s dynamic
provided the original work is
properly cited. nature. Unlike static therapeutic interventions, AI systems learn and adapt, raising
questions about when and how to assess clinical efficacy. Real-world evidence generation
Publisher’s Note: AccScience
Publishing remains neutral with must become integral to AI development, not an afterthought.
regard to jurisdictional claims in
published maps and institutional Second, implementation research must address the human factors that determine AI
affiliations. adoption success. Clinician workflow integration, patient acceptance, and organizational

Volume 11 Issue 4 (2025) 1 doi: 10.36922/JCTR025330054