This repository contains the trained deep learning model weights used in the study:
"AI-Assisted Cytological Detection and Classification of Canine Round Cell Tumors"
-
YOLOv8x Object Detection Model
File: yolov8x_best.pt
Purpose: Localization of tumor regions in cytological images.
-
ConvNeXt-Tiny Classification Model
File: convnext_tiny_final_earlystop.pth
Purpose: Multi-class classification of cropped tumor regions into:
- Histiocytoma
- Lymphoma
- Mast Cell Tumor
- Transmissible Venereal Tumor (TVT)
- Negative (non-round cell tumor)
These models are provided for research and educational purposes only and are not intended to replace professional veterinary diagnostic evaluation.
