Examples

This page provides complete examples of SMINT workflows, demonstrating how to use the package for different spatial omics applications.

Basic Workflow

The following example demonstrates a complete workflow from preprocessing through segmentation to visualization:

```python import os from smint.preprocessing import preprocess_ome_tiff from smint.segmentation import process_large_image from smint.visualization import visualize_cell_outlines import pandas as pd import matplotlib.pyplot as plt

Set up directories

image_path = "path/to/image.ome.tif" output_dir = "results" os.makedirs(output_dir, exist_ok=True)

Step 1: Preprocess the image

preprocessed = preprocess_ome_tiff( image_path=image_path, sigma=1.5, normalize=True )

Combine channels if needed

if len(preprocessed) >= 2: from smint.preprocessing import combine_channels

channel_names = list(preprocessed.keys())
combined = combine_channels(
    preprocessed[channel_names[0]],
    preprocessed[channel_names[1]]
)

# Save combined image
import tifffile
combined_path = os.path.join(output_dir, "combined.tif")
tifffile.imwrite(combined_path, combined)
print(f"Saved combined image to {combined_path}")

# Use combined image for segmentation
segmentation_input = combined_path

else: # Use first channel for segmentation import tifffile channel_name = list(preprocessed.keys())[0] segmentation_input = os.path.join(output_dir, f"{channel_name}.tif") tifffile.imwrite(segmentation_input, preprocessed[channel_name])

Step 2: Segment cells and nuclei

results = process_large_image( image_path=segmentation_input, csv_base_path=os.path.join(output_dir, "cell_outlines"), chunk_size=(2048, 2048), # Cell Model parameters cell_model_path="cyto", cells_diameter=120.0, cells_flow_threshold=0.4, cells_cellprob_threshold=-1.5, cells_channels=[0, 0], # Nuclei Model parameters nuclei_model_path="nuclei", nuclei_diameter=40.0, nuclei_flow_threshold=0.4, nuclei_cellprob_threshold=-1.2, nuclei_channels=[0, 0], # Visualization parameters visualize=True, visualize_output_dir=os.path.join(output_dir, "visualizations") )

print(f"Found {results['total_cells']} cells and {results['total_nuclei']} nuclei")

Step 3: Load and visualize the results

if results.get('cells_csv_path'): cells_df = pd.read_csv(results['cells_csv_path']) print(f"Loaded {cells_df['global_cell_id'].nunique()} cells")

# Create visualization
fig = visualize_cell_outlines(cells_df, color_by='global_cell_id')
plt.savefig(os.path.join(output_dir, "cell_outlines.png"), dpi=300)
plt.close(fig)