File size: 3,695 Bytes
172cc9a e1f0de7 172cc9a 8591b97 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 |
---
language: en
license: mit
tags:
- deep-learning
- medical-imaging
- tumor-detection
- MRI
- h5
model-index:
- name: MRI_LLM
results:
- task:
type: image-classification
dataset:
name: Public MRI datasets (Kaggle, NIH, TCIA)
type: image
metrics:
- type: accuracy
value: 95.2
---
# MRI_LLM: Brain, Breast, and Lung Tumor Detection Models
📌 **Author**: Vijayendher Gatla (@wizaye)
📌 **Repository**: [https://huggingface.co/wizaye/MRI_LLM](https://huggingface.co/wizaye/MRI_LLM)
📌 **License**: MIT
📌 **Tags**: `deep-learning`, `medical-imaging`, `tumor-detection`, `MRI`, `h5`
---
## **Model Overview**
The **MRI_LLM** repository contains three deep learning models trained for **tumor detection** in **brain, breast, and lung MRIs**. These models leverage deep neural networks to assist in the automated diagnosis of tumors from medical imaging data.
### **Models Included**
- **Brain Tumor Model (`brain_model.h5`)**: Detects tumors in MRI brain scans.
- **Breast Tumor Model (`breast_tumor.h5`)**: Identifies malignant and benign breast tumors.
- **Lung Tumor Model (`lung_tumor.h5`)**: Predicts lung tumors using CT/MRI scans.
---
## **Intended Use**
These models are designed for **research and educational purposes**. They can be used for:
✅ Assisting radiologists in medical image analysis
✅ Experimenting with deep learning in healthcare
✅ Further fine-tuning on custom datasets
**⚠️ Disclaimer:** These models are **not** FDA/CE-approved and should not be used for clinical diagnosis.
---
## **Model Architecture**
Each model is based on **Convolutional Neural Networks (CNNs)**, specifically optimized for medical image classification. The architecture includes:
- **Feature extraction** layers for capturing patterns in MRI scans
- **Fully connected** layers for classification
- **Softmax/Sigmoid activation** depending on the number of classes
---
## **Dataset**
- The models were trained on **publicly available MRI datasets** (e.g., Kaggle, NIH, TCIA).
- Data preprocessing included **normalization, augmentation, and resizing**.
- If you are using these models, make sure to verify dataset compatibility.
---
## **How to Use**
### **Load the Model**
```python
from tensorflow.keras.models import load_model
# Load Brain Tumor Model
model = load_model("brain_model.h5")
# Predict on new images
import numpy as np
from tensorflow.keras.preprocessing import image
img_path = "sample_mri.jpg"
img = image.load_img(img_path, target_size=(224, 224))
img_array = image.img_to_array(img) / 255.0
img_array = np.expand_dims(img_array, axis=0)
prediction = model.predict(img_array)
print("Tumor Detected" if prediction > 0.5 else "No Tumor")
```
---
## **Performance Metrics**
| Model | Accuracy | Precision | Recall |
|--------|----------|------------|----------|
| **Brain Tumor** | 95.2% | 94.8% | 96.1% |
| **Breast Tumor** | 93.5% | 92.7% | 94.3% |
| **Lung Tumor** | 96.1% | 95.9% | 96.8% |
📌 Trained using **TensorFlow/Keras** on NVIDIA GPUs.
---
## **Limitations & Future Work**
🔹 Limited dataset coverage—may not generalize to all MRI variations.
🔹 Possible false positives/negatives in real-world cases.
🔹 Can be improved with **transfer learning** on hospital-specific datasets.
---
## **Citation**
If you use this model, please cite:
```bibtex
@misc{MRI_LLM,
author = {Vijayendher Gatla},
title = {MRI-Based Tumor Detection Models},
year = {2025},
url = {https://huggingface.co/wizaye/MRI_LLM}
}
``` |