|
--- |
|
license: mit |
|
--- |
|
|
|
--- |
|
license: mit |
|
--- |
|
|
|
📄 **Model Card for Hugging Face (ProViCNet Weights)** |
|
|
|
## **ProViCNet: Prostate-Specific Foundation Models with Patch-Level Contrast for Cancer Detection** |
|
|
|
|
|
### **📌 Overview** |
|
ProViCNet is an **organ-specific foundation model** designed for **prostate cancer detection** using **multi-modal medical imaging (mpMRI & TRUS)**. The model leverages **Vision Transformers (ViTs) with patch-level contrastive learning** to improve **cancer localization and classification**. |
|
|
|
These pre-trained weights are provided for **research and clinical AI development** and can be used for **inference (feature extraction and cancer detection)** on prostate imaging datasets. |
|
|
|
📌 **For usage examples and detailed documentation, visit:** |
|
🔗 **[ProViCNet GitHub Repository](https://github.com/pimed/ProViCNet/)** |
|
|
|
📄 **Reference Paper:** |
|
🔗 **[ProViCNet: Organ-Specific Foundation Model for Prostate Cancer Detection](https://arxiv.org/abs/2502.00366)** |
|
|
|
|
