Brain Tumor Detection using YOLOv10

This project demonstrates how to use the YOLOv10 model for brain tumor detection using MRI images. The model has been trained and fine-tuned on a dataset to detect brain tumors efficiently.

---

Table of Contents

• Brain Tumor Detection using YOLOv10
  • Table of Contents
  • Overview
    • Tools and Libraries Used:
  • Installation and Setup
    • YOLOv10 Installation
    • Roboflow Integration
    • Training the Model
  • Model Inference
  • Visualization
  • Web Application
  • Results
  • License
  • References

---

Overview

This project uses YOLOv10 for detecting brain tumors from MRI scans. We employ Roboflow for dataset management and Gradio for creating a web-based interface.

Tools and Libraries Used:

• YOLOv10
• Roboflow
• Gradio
• Matplotlib
• OpenCV

---

Installation and Setup

YOLOv10 Installation

1. Install YOLOv10 from the official GitHub repository:

   !pip install -q git+https://github.com/THU-MIG/yolov10.git

2. Download the pre-trained YOLOv10n model weights:

   !wget -P -q https://github.com/jameslahm/yolov10/releases/download/v1.0/yolov10n.pt

3. Install Roboflow for dataset handling:

   !pip install -q roboflow

---

Roboflow Integration

1. Install and set up Roboflow to manage and download the dataset:

   from roboflow import Roboflow
   rf = Roboflow(api_key="2A2n8FdjhCq768GBy1Yw")
   project = rf.workspace("brain-mri").project("mri-rskcu")
   version = project.version(3)
   dataset = version.download("yolov9")

---

Training the Model

1. Start training the YOLOv10 model with the following command. You can modify parameters such as epochs, batch, and model:

   !yolo task=detect mode=train epochs=25 batch=32 plots=True \
   model=/content/yolov10n.pt \
   data=/content/MRI-3/data.yaml

---

Model Inference

1. Load the trained model:

   from ultralytics import YOLOv10
   
   model_path = "/content/runs/detect/train9/weights/best.pt"
   model = YOLOv10(model_path)

2. Run inference on the dataset with a confidence threshold:

   result = model(source="/content/MRI-3/train/images", conf=0.25, save=True)

---

Visualization

1. Visualize predictions from multiple images using Matplotlib:

   import glob
   import matplotlib.pyplot as plt
   import matplotlib.image as mpimg
   
   images = glob.glob("/content/runs/detect/predict/*.jpg")
   images_to_display = images[:10]
   
   fig, axes = plt.subplots(2, 5, figsize=(20, 10))
   
   for i, ax in enumerate(axes.flat):
       if i < len(images_to_display):
           img = mpimg.imread(images_to_display[i])
           ax.imshow(img)
           ax.axis('off')
       else:
           ax.axis('off')
   plt.tight_layout()
   plt.show()

2. Visualize predictions on a single image:

   result = model.predict(source="/content/MRI-3/valid/images/Tr-gl_0228_jpg.rf.b9ecef834d39f770e41b0585b63bdc1a.jpg", imgsz=640, conf=0.25)
   annotated_img = result[0].plot()
   annotated_img[:, :, ::-1]

---

Web Application

1. Install Gradio:

   !pip install gradio

2. Define the predict() function and launch the app:

   import gradio as gr
   import cv2
   import numpy as np
   
   def predict(image):
       result = model.predict(source=image, imgsz=640, conf=0.25)
       annotated_img = result[0].plot()
       annotated_img = annotated_img[:, :, ::-1]
       return annotated_img
   
   app = gr.Interface(
       fn=predict,
       inputs=gr.Image(type="numpy", label="Upload an image"),
       outputs=gr.Image(type="numpy", label="Detect Brain Tumor"),
       title="Brain Tumor Detection Using YOLOv10",
       description="Upload an image and the YOLOv10 model will detect and annotate the brain tumor."
   )
   
   app.launch()

---

Results

---

License

This project is licensed under the MIT License.

---

References

• YOLOv10 Official GitHub Repository
• Roboflow
• Gradio

Feel free to contribute or suggest improvements!

---

Note: Replace the images/ paths with the actual paths to the images used in your project. If you don't have these images, you might need to create or capture them based on your project's outputs.