Abstract:Brain tumor detection in multiplane Magnetic Resonance Imaging (MRI) slices is a challenging task due to the various appearances and relationships in the structure of the multiplane images. In this paper, we propose a new You Only Look Once (YOLO)-based detection model that incorporates Pretrained Knowledge (PK), called PK-YOLO, to improve the performance for brain tumor detection in multiplane MRI slices. To our best knowledge, PK-YOLO is the first pretrained knowledge guided YOLO-based object detector. The main components of the new method are a pretrained pure lightweight convolutional neural network-based backbone via sparse masked modeling, a YOLO architecture with the pretrained backbone, and a regression loss function for improving small object detection. The pretrained backbone allows for feature transferability of object queries on individual plane MRI slices into the model encoders, and the learned domain knowledge base can improve in-domain detection. The improved loss function can further boost detection performance on small-size brain tumors in multiplanar two-dimensional MRI slices. Experimental results show that the proposed PK-YOLO achieves competitive performance on the multiplanar MRI brain tumor detection datasets compared to state-of-the-art YOLO-like and DETR-like object detectors. The code is available at <a class="link-external link-https" href="https://github.com/mkang315/PK-YOLO" rel="external noopener nofollow">this https URL</a>.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the challenge of automatically detecting brain tumors in multi - planar magnetic resonance imaging (MRI) slices. Specifically, due to the various appearances of tumors and the relationships of their internal structures in multi - planar images, it is very difficult to accurately and automatically detect these lesions. In particular, when converting from 3D multi - planar MRI scans to 2D slices, the number of small - sized lesions usually increases, which further increases the detection difficulty. Therefore, this paper proposes a new YOLO (You Only Look Once) - based object detection model - PK - YOLO, which improves the brain tumor detection performance in multi - planar MRI slices by introducing pre - trained knowledge. ### Main problems 1. **Brain tumor detection in multi - planar MRI slices**: The diversity and complexity of tumors in multi - planar MRI images make it difficult for existing deep - learning methods to achieve high - precision detection effects on all planar images simultaneously. 2. **Detection of small - sized lesions**: In multi - planar MRI slices, there are a large number of small - sized lesions, and existing detection methods perform poorly when dealing with these small targets. 3. **Feature extraction and model generalization**: How to effectively extract and utilize the features in multi - planar MRI slices and improve the generalization ability of the model on different datasets. ### Solutions To address the above challenges, the paper proposes the following solutions: 1. **Pre - trained lightweight convolutional neural network (CNN) backbone network**: Use the RepViT backbone network and pre - train it through sparse mask modeling (SparK) to inject domain knowledge and enhance the model's ability to recognize tumor features in multi - planar MRI slices. 2. **Improved YOLO architecture**: Combine the pre - trained backbone network with an improved regression loss function (Focaler - IoU) to improve the model's performance in small - target detection. 3. **Multi - level information fusion**: Provide supplementary information through auxiliary branches to alleviate the problem of information loss in deep neural networks and improve the model's ability to detect targets of different scales. ### Experimental results The experimental results show that the proposed PK - YOLO model achieves better performance than the existing state - of - the - art YOLO - like and DETR - like object detectors on the multi - planar MRI brain tumor detection dataset. Especially in the detection of small - sized tumors, PK - YOLO shows significant advantages. ### Conclusion By introducing pre - trained knowledge and an improved model architecture, PK - YOLO effectively solves the problem of brain tumor detection in multi - planar MRI slices and improves the detection accuracy and robustness. This research provides new ideas and technical support for the automated detection of brain tumors.

PK-YOLO: Pretrained Knowledge Guided YOLO for Brain Tumor Detection in Multiplanar MRI Slices

Automated Brain Tumor Segmentation and Classification in MRI Using YOLO-Based Deep Learning

RCS-YOLO: A Fast and High-Accuracy Object Detector for Brain Tumor Detection

Brain Tumor Detection Based on Deep Learning Approaches and Magnetic Resonance Imaging

Accurate Detection of Brain Tumor Lesions From Medical Images Based on Improved YOLOv8 Algorithm

BGF-YOLO: Enhanced YOLOv8 with Multiscale Attentional Feature Fusion for Brain Tumor Detection

Analysis of MRI brain tumor images using deep learning techniques

Enhancing brain tumor detection in MRI images using YOLO-NeuroBoost model

Brain Tumor Detection with YOLOv8

Computer-aided detection and classification of brain tumor using YOLOv3 and deep learning

YOLO Object Detector and Inception-V3 Convolutional Neural Network for Improved Brain Tumor Segmentation

Advanced Brain Tumor Segmentation and Detection using YOLOv8

An improved framework for brain tumor analysis using MRI based on YOLOv2 and convolutional neural network

Automatic Detection of Brain Tumor on MRI Images Using a YOLO-Based Algorithm

Intraoperative Glioma Segmentation with YOLO + SAM for Improved Accuracy in Tumor Resection

A deep learning model for the localization and extraction of brain tumors from MR images using YOLOv7 and grab cut algorithm

Brain Cancer Segmentation Using YOLOv5 Deep Neural Network

A Review of Brain Tumor Detection Techniques Using YOLOv8

Brain tumor detection and segmentation using deep learning

Enhancing Automated Brain Tumor Detection Accuracy Using Artificial Intelligence Approaches for Healthcare Environments

A Deep Learning Approach for Brain Tumor Firmness Detection Using YOLOv4