RRFNet - Reparameterization Technology - A free-anchor brain tumor detection and classification network

Sep 23, 2026

Medical imaging advancements enable CT and MRI to capture detailed brain structures for more effective tumor diagnosis, yet manual interpretation remains time-consuming. This research applies deep learning to brain tumor detection and classification, examining how model parameters, input batch size, and anchor-box design affect performance under complex medical imaging conditions. Experiments show excessive parameters and anchor boxes can reduce accuracy, while more tumor samples improve results. The proposed RRFNet uses RepConv-RepC3 backbone with FGConcat splicing to improve accuracy and inference speed, reaching 79.2% mAP versus YOLOv8.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages23

Reading Time58 min

AccessFree to download

Abstract

Introduction

PLOS ONE OPEN ACCESS Citation: Liu W, Guo X (2025) RRFNet: A free-anchor brain tumor detection and classification network based on reparameterization technology. PLoS One 20(6): e0325483 . [https://doi.org/10.1371/journal.pone.0325483](https://doi.org/10.1371/journal.pone.0325483) Editor: Hirenkumar Kantilal Mewada, Prince Mohammad Bin Fahd University, SAUDI ARABIA Received: September 21, 2024 Accepted: May 12, 2025 Published: June 16, 2025 Copyright: © 2025 Liu, Guo. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data availability statement: The codes and datasets related to this paper have been published on Github. Please refer to [https://github.com/Mr-cream/RRFNet/](https://github.com/Mr-cream/RRFNet/) . Funding: This work was funded by the Natural Science Foundation of Hunan Province, 2022JJ30438 awarded to W.L., Natural Science Foundation of Changsha, kq2202260, awarded to W.L., and Hunan Province Traditional Chinese Medicine Research Project, B2023039 awarded to W.L. RESEARCH ARTICLE RRFNet: A free-anchor brain tumor detection and classification network based on reparameterization technology Wei Liu∗ , Xingxin Guo School of Informatics, Hunan University of Chinese Medicine, Changsha, Hunan, China ∗ [weiliu@hnucm.edu.cn](weiliu@hnucm.edu.cn) Abstract Advancements in medical imaging technology have facilitated the acquisition of highquality brain images through computed tomography (CT) or magnetic resonance imaging (MRI), enabling professional brain specialists to diagnose brain tumors more effectively. However, manual diagnosis is time-consuming, which has led to the growing importance of automatic detection and classification through brain imaging. Conventional object detection models for brain tumor detection face limitations in brain tumor detection owing to the significant differences between medical images and natural scene images, as well as challenges such as complex backgrounds, noise interference, and blurred boundaries between cancerous and normal tissues. This study investigates the application of deep learning to brain tumor detection, analyzing the effect of three factors, the number of model parameters, input data batch size, and the use of anchor boxes, on detection performance. Experimental results reveal that an excessive number of model parameters or the use of anchor boxes may reduce detection accuracy. However, increasing the number of brain tumor samples improves detection performance. This study, introducesa backbone network built using RepConv and RepC3, along with FGConcat feature map splicing module to optimize the brain tumor detection model. The experimental results show that the proposed RepConv-RepC3-FGConcat Network (RRFNet) can learn underlying semantic information about brain tumors during training stage, while maintaining a low number of parameters during inference, which improves the speed of brain tumor detection. Compared with YOLOv8, RRFNet achieved a higher accuracy in brain tumor detection, with a mAP value of 79.2% . This optimized approach enhances both accuracy and efficiency, which is essential in clinical settings where time and precision are critical. Introduction A brain tumor is a collection or mass of abnormal cells in the brain. When a benign or malignant tumor grows within the confined space of the brain, it increases intracranial pressure, leading to brain damage and potentially life-threatening conditions. The brain is also a common site for metastasis from other types of cancer, such as lung, breast, skin (melanoma), Competing interests: The authors hereby declare that they have no financial or non-financial interests, including personal relationships, that could be perceived as influencing the work reported in this paper. colon, kidney, an

FAQs

Why is automatic brain tumor detection important compared with manual diagnosis?

Manual brain tumor diagnosis is time-consuming and requires specialized radiologists to process large MRI datasets. Automatic detection reduces workload and can improve decision support in clinical workflows.

Which factors does the study analyze for detection performance?

The study analyzes the effect of three factors: the number of model parameters, input data batch size, and the use of anchor boxes.

What is the key architectural idea behind RRFNet and its benefit?

RRFNet introduces a backbone built using RepConv and RepC3, together with an FGConcat feature map splicing module. It aims to learn semantic tumor information during training while keeping inference parameters low to improve detection speed.

Sep 23, 2026

Uploaded by นรินทร์