Hugging Face's logo

larrikin-coder
/
brain-tumor-cnn

Eval Results
Model card Files Community
brain-tumor-cnn
File size: 2,811 Bytes 
7845d6e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ebf0a45
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
 
---
license: mit
datasets:
- brain-mri-dataset
metrics:
- accuracy
- auc
model-index:
- name: CNN Brain Tumor Classifier
  results:
    - task:
        type: image-classification
      dataset:
        name: Brain MRI Dataset
        type: brain-mri-dataset
      metrics:
        - name: Accuracy
          type: accuracy
          value: 0.90
        - name: AUC
          type: auc
          value: 0.90
---

# 🧠 CNN Brain Tumor Classifier

## Model Description
This repository contains a Convolutional Neural Network (CNN) built with **TensorFlow/Keras** for classifying brain MRI scans.  
The model can distinguish between three types of brain tumors and healthy scans.  

⚠️ **Disclaimer**: This model is provided for **educational and research purposes only**.  
It is **not a medical diagnostic tool** and should not be used in clinical practice.  

---

## Classes
The model predicts one of the following four categories:
- **Glioma**
- **Meningioma**
- **Pituitary tumor**
- **No tumor** (healthy)

---

## Training Details
- **Framework**: TensorFlow / Keras  
- **Architecture**: Custom CNN  
- **Input size**: 224 × 224 RGB MRI images  
- **Loss function**: `categorical_crossentropy`  
- **Optimizer**: `Adam`  
- **Epochs**: 10  
- **Metrics**: Accuracy, AUC  

---

## Dataset
- **Source**: Public brain MRI dataset (glioma, meningioma, pituitary, no tumor)  
- **Preprocessing**:
  - Images resized to 224 × 224  
  - Normalized to [0, 1] range  
  - Augmentation (rotation, flipping, zoom) applied during training  

---

## Evaluation Results
| Metric              | Value     |
|---------------------|-----------|
| Training Accuracy   | ~96%      |
| Validation Accuracy | ~85–90%   |
| AUC (training)      | ~0.90     |

*(values may vary depending on train/validation split)*  

---

## Usage

### Installation
```bash
pip install tensorflow huggingface_hub
```


```python
from tensorflow.keras.models import load_model
from huggingface_hub import hf_hub_download
import numpy as np
from tensorflow.keras.preprocessing import image

# Download model file from Hugging Face Hub
model_path = hf_hub_download(
    repo_id="larrikin-coder/brain-tumor-cnn",  # replace with your repo
    filename="cnn_model.h5"
)

# Load model
model = load_model(model_path)

# Preprocess an image
img = image.load_img("test_mri.jpg", target_size=(224, 224))
img_array = image.img_to_array(img) / 255.0
img_array = np.expand_dims(img_array, axis=0)

# Predict
pred = model.predict(img_array)
class_names = ["glioma", "meningioma", "pituitary", "no_tumor"]
print("Prediction:", class_names[np.argmax(pred)])
```

@misc{larrikin-coder2025,
  title={CNN Brain Tumor Classifier},
  author={Larrikin Coder},
  year={2025},
  howpublished={\url{https://huggingface.co/larrikin-coder/brain-tumor-cnn}}
}