Update README.md
Browse files
README.md
CHANGED
|
@@ -22,7 +22,6 @@ tags:
|
|
| 22 |
---
|
| 23 |
|
| 24 |
# Vision Transformer for Brain Tumor Multiclass Classification (v2)
|
| 25 |
-
|
| 26 |
This model is a fine-tuned Vision Transformer (ViT) for multiclass brain tumor MRI classification. It predicts one of the following five classes:
|
| 27 |
|
| 28 |
• Glioma
|
|
@@ -34,7 +33,6 @@ This model is a fine-tuned Vision Transformer (ViT) for multiclass brain tumor M
|
|
| 34 |
The model is integrated into an online diagnostic support platform at https://www.medscanai.net/.
|
| 35 |
|
| 36 |
---
|
| 37 |
-
|
| 38 |
## Model Details
|
| 39 |
|
| 40 |
- **Base Model:** google/vit-base-patch16-224-in21k
|
|
@@ -46,14 +44,12 @@ The model is integrated into an online diagnostic support platform at https://ww
|
|
| 46 |
- **Not for clinical diagnosis**
|
| 47 |
|
| 48 |
---
|
| 49 |
-
|
| 50 |
## Dataset
|
| 51 |
|
| 52 |
The primary dataset is sourced from the Kaggle "Brain Tumor MRI Dataset" by Masoud Nickparvar.
|
| 53 |
Additional "unknown" category samples were collected from publicly available online sources to evaluate robustness.
|
| 54 |
|
| 55 |
---
|
| 56 |
-
|
| 57 |
## Training
|
| 58 |
|
| 59 |
- 20 epochs using Hugging Face Trainer
|
|
@@ -63,9 +59,6 @@ Additional "unknown" category samples were collected from publicly available onl
|
|
| 63 |
- Evaluation metric: Accuracy
|
| 64 |
|
| 65 |
---
|
| 66 |
-
|
| 67 |
-
## Performance
|
| 68 |
-
|
| 69 |
## Performance
|
| 70 |
|
| 71 |
| Class | Precision | Recall | F1-score |
|
|
@@ -79,7 +72,6 @@ Additional "unknown" category samples were collected from publicly available onl
|
|
| 79 |
**Overall accuracy: 98 percent**
|
| 80 |
|
| 81 |
---
|
| 82 |
-
|
| 83 |
## How to Use
|
| 84 |
|
| 85 |
```python
|
|
|
|
| 22 |
---
|
| 23 |
|
| 24 |
# Vision Transformer for Brain Tumor Multiclass Classification (v2)
|
|
|
|
| 25 |
This model is a fine-tuned Vision Transformer (ViT) for multiclass brain tumor MRI classification. It predicts one of the following five classes:
|
| 26 |
|
| 27 |
• Glioma
|
|
|
|
| 33 |
The model is integrated into an online diagnostic support platform at https://www.medscanai.net/.
|
| 34 |
|
| 35 |
---
|
|
|
|
| 36 |
## Model Details
|
| 37 |
|
| 38 |
- **Base Model:** google/vit-base-patch16-224-in21k
|
|
|
|
| 44 |
- **Not for clinical diagnosis**
|
| 45 |
|
| 46 |
---
|
|
|
|
| 47 |
## Dataset
|
| 48 |
|
| 49 |
The primary dataset is sourced from the Kaggle "Brain Tumor MRI Dataset" by Masoud Nickparvar.
|
| 50 |
Additional "unknown" category samples were collected from publicly available online sources to evaluate robustness.
|
| 51 |
|
| 52 |
---
|
|
|
|
| 53 |
## Training
|
| 54 |
|
| 55 |
- 20 epochs using Hugging Face Trainer
|
|
|
|
| 59 |
- Evaluation metric: Accuracy
|
| 60 |
|
| 61 |
---
|
|
|
|
|
|
|
|
|
|
| 62 |
## Performance
|
| 63 |
|
| 64 |
| Class | Precision | Recall | F1-score |
|
|
|
|
| 72 |
**Overall accuracy: 98 percent**
|
| 73 |
|
| 74 |
---
|
|
|
|
| 75 |
## How to Use
|
| 76 |
|
| 77 |
```python
|