Submit

Login Login

Register Register

Home / Journals / CMC / Online First / doi:10.32604/cmc.2026.086547
Journal Menu
Special Issues
Table of Content

All issues

Online First

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Open Access

REVIEW

YOLO in Medical Imaging: Evolution, Clinical Translation, and Future Directions

Hue T. Tran1,2, Nguyen Nhu Son3, Anh Nguyen Trong4, Anh Nguyen Tu5, Phat T. Nguyen6,*, Giang L. Nguyen3,*
1 Graduate University of Science and Technology, Vietnam Academy of Science and Technology, Hanoi, Vietnam
2 Institute of Information Technology, Vietnam Women’s Academy, Hanoi, Vietnam
3 Institute of Information Technology, Vietnam Academy of Science and Technology, Hanoi, Vietnam
4 Department of Information Technology, Dai Nam University, Hanoi, Vietnam
5 Managing Director, White Neuron Co., Ltd., Hanoi, Vietnam
6 Department of Circuit Theory and Measurement, Faculty of Radio-Electronic Engineering, Le Quy Don Technical University, Hanoi, Vietnam
* Corresponding Author: Phat T. Nguyen. Email: email; Giang L. Nguyen. Email: email

Computers, Materials & Continua https://doi.org/10.32604/cmc.2026.086547

Received 01 June 2026; Accepted 17 August 2026; Published online 14 September 2026

Abstract

Over the past decade, rapid advances in artificial intelligence have substantially expanded applications in medical image analysis, computer-aided diagnosis, and clinical decision support. Among object detection architectures, the YOLO (You Only Look Once) family has emerged as one of the most widely adopted approaches due to its real-time inference capability. However, transferring YOLO from general computer vision to clinical settings still faces major challenges, including small lesion detection and the gap between algorithmic performance and real-world clinical utility. This topical review provides a structured narrative synthesis of YOLO applications in medical imaging from 2016 to 2026. Unlike previous surveys, this work proposes a unified analytical framework and a four-dimensional taxonomy that categorizes studies according to task type, imaging modality, clinical context, and data regime. Based on this framework, the review analyses the architectural evolution of YOLO. To support quantitative analysis, the paper compiles a meta-benchmark table across standard datasets, together with a quality assessment table based on CLAIM 2024 and TRIPOD+AI guidelines. Another contribution is the analysis of the 2010–2015 foundations and the co-evolution of YOLO and modern medicine during 2016–2026, clarifying how medical challenges have shaped algorithmic design and vice versa. The paper also devotes separate sections to three issues often overlooked in previous reviews: (i) dataset and model bias analysis, (ii) the reproducibility crisis in medical AI, and (iii) failure analysis (“Failure modes and limitations of YOLO in medical imaging”). Finally, a four-layer conceptual framework, termed the “Clinical-Aware YOLO Framework,” is proposed to bridge the gap between algorithm development and clinical deployment. This framework remains conceptual and requires prospective validation. The review concludes that the next generation of medical YOLO systems must shift from benchmark-oriented optimization toward clinical awareness, in which algorithmic fairness, reproducibility, and prospective validation are treated as core design requirements.

Keywords

YOLO; medical imaging; object detection; deep learning; clinical AI; real-time inference; foundation models; CLAIM 2024; TRIPOD+AI; reproducibility; algorithmic fairness

  • 109

    View

  • 25

    Download

  • 0

    Like

Related articles
Copyright© 2026 Tech Science Press
© 1997-2026 TSP (Henderson, USA) unless otherwise stated

Share Link