giuseppericcio commited on
Commit
2c4154f
·
verified ·
1 Parent(s): 955a8be

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +18 -19
README.md CHANGED
@@ -1,15 +1,15 @@
1
- ---
2
- title: BrainGemma3D
3
- emoji: 🚀
4
- colorFrom: gray
5
- colorTo: blue
6
- sdk: gradio
7
- sdk_version: 6.6.0
8
- app_file: app.py
9
- pinned: false
10
- license: cc-by-4.0
11
- short_description: 🧠Brain Report Automation via Inflated ViT in 3D
12
- ---
13
 
14
  # 🧠 BrainGemma3D Dashboard
15
 
@@ -159,11 +159,11 @@ Click **"📄 Download PDF Report"** to generate a professional radiology report
159
  You can test the dashboard with publicly available brain MRI datasets:
160
 
161
  ### BraTS (Brain Tumors):
162
- - Download from: [BraTS Challenge](https://www.med.upenn.edu/cbica/brats2020/data.html)
163
  - Example: `BraTS20_Training_001_flair.nii.gz`
164
 
165
- ### ADNI (Alzheimer's Disease):
166
- - Download from: [ADNI Database](http://adni.loni.usc.edu/)
167
  - Requires registration (free for researchers)
168
 
169
  ---
@@ -243,7 +243,7 @@ AI-generated reports may contain errors, hallucinations, or incomplete informati
243
 
244
  ## 📊 Performance Benchmarks
245
 
246
- ### Accuracy Metrics (BraTS 2020 Validation)
247
 
248
  | Metric | BrainGemma3D | Baseline |
249
  |--------|--------------|----------|
@@ -277,7 +277,6 @@ This project was developed by:
277
  ### Built With
278
  - [Google MedGemma](https://huggingface.co/google/medgemma-1.5-4b-it) — Medical domain language model
279
  - [Google MedSigLIP](https://huggingface.co/google/medsiglip-base-patch16-448) — Medical vision encoder
280
- - [BraTS 2021](https://www.med.upenn.edu/cbica/brats2021/) — Brain tumor segmentation dataset
281
  - [Hugging Face Transformers](https://huggingface.co/docs/transformers) — Model framework
282
 
283
  ---
@@ -297,8 +296,8 @@ This project was developed by:
297
  - **MedGemma Model Family**: https://huggingface.co/google/medgemma
298
 
299
  ### Datasets
300
- - **BraTS Challenge**: https://www.med.upenn.edu/cbica/brats2020
301
- - **ADNI Database**: http://adni.loni.usc.edu
302
 
303
  ---
304
 
 
1
+ ---
2
+ title: BrainGemma3D
3
+ emoji: 🚀
4
+ colorFrom: gray
5
+ colorTo: blue
6
+ sdk: gradio
7
+ sdk_version: 6.6.0
8
+ app_file: app.py
9
+ pinned: false
10
+ license: cc-by-4.0
11
+ short_description: 🧠Brain Report Automation via Inflated ViT in 3D
12
+ ---
13
 
14
  # 🧠 BrainGemma3D Dashboard
15
 
 
159
  You can test the dashboard with publicly available brain MRI datasets:
160
 
161
  ### BraTS (Brain Tumors):
162
+ - Download from: [BraTS 2020](https://www.kaggle.com/datasets/awsaf49/brats20-dataset-training-validation)
163
  - Example: `BraTS20_Training_001_flair.nii.gz`
164
 
165
+ ### MPI-Leipzig Mind-Brain-Body (Healthy Controls):
166
+ - Download from: [MPI-Leipzig Mind-Brain-Body](https://openneuro.org/datasets/ds000221/versions/00002)
167
  - Requires registration (free for researchers)
168
 
169
  ---
 
243
 
244
  ## 📊 Performance Benchmarks
245
 
246
+ ### Accuracy Metrics (BraTS 2020 dataset)
247
 
248
  | Metric | BrainGemma3D | Baseline |
249
  |--------|--------------|----------|
 
277
  ### Built With
278
  - [Google MedGemma](https://huggingface.co/google/medgemma-1.5-4b-it) — Medical domain language model
279
  - [Google MedSigLIP](https://huggingface.co/google/medsiglip-base-patch16-448) — Medical vision encoder
 
280
  - [Hugging Face Transformers](https://huggingface.co/docs/transformers) — Model framework
281
 
282
  ---
 
296
  - **MedGemma Model Family**: https://huggingface.co/google/medgemma
297
 
298
  ### Datasets
299
+ - **[BraTS 2020](https://www.kaggle.com/datasets/awsaf49/brats20-dataset-training-validation)**: 369 brain tumor MRI cases with clinical annotations from [TextBraTS 2021](https://github.com/Jupitern52/TextBraTS)
300
+ - **Healthy Controls**: 99 preprocessed healthy brain scans (skull-stripped, normalized) from [MPI-Leipzig Mind-Brain-Body](https://openneuro.org/datasets/ds000221/versions/00002)
301
 
302
  ---
303