LIVR-VUB/eHeps-Bodipy-models

eHeps Cellpose Models

This repository contains custom Cellpose models trained for segmenting eHeps (engineered hepatocytes) in microscopy images.

These models were developed by the LIVR-VUB (Liver Cell Biology Research Group) at Vrije Universiteit Brussel.

Models Available

Model Name	Description	Target
cpsam_eHeps_nuclei	Segmenting nuclei in eHeps cultures.	Nuclei (DAPI/Hoechst)
cpsam_eHeps_v3	General cell segmentation for eHeps (v3).	Cells / Cytoplasm

Usage

You can use these models with the Cellpose Python library.

Installation

pip install cellpose

Python Example

from cellpose import models, io

# Load the model
# Choose 'cpsam_eHeps_nuclei' or 'cpsam_eHeps_v3'
model_path = "cpsam_eHeps_v3" 
model = models.CellposeModel(gpu=True, pretrained_model=model_path)

# Load your image
# image = io.imread('your_image.tif')

# Run segmentation
# For nuclei model, use channels=[0,0] if grayscale
# For cell model, adjust channels as needed (e.g., [1,2] for Cytoplasm, Nuclei)
masks, flows, styles = model.eval(image, diameter=None, channels=[0,0])

Training

These models were fine-tuned using the Cellpose "human-in-the-loop" approach on the CP-SAM architecture (Cellpose-SAM).

Citation

If you use these models in your research, please cite:

LIVR-VUB / IA-Yuwei-Bodipy-Pipeline

---

Uploaded by LIVR-VUB