SIM/WFM Segmentation
====================

Structured Illumination Microscopy (SIM) and Widefield Microscopy (WFM) segmentation module provides comprehensive tools for super-resolution and conventional fluorescence microscopy data analysis.

Core Functions
--------------

Core Segmentation API
----------------------

SIM/WFM segmentation now uses the unified Segmenter API for consistent interface.

**Recommended Usage** (Unified API):

.. code-block:: python

    from ipa.processing.segmentation import SIMERSegmenter, SIMMitoSegmenter, WFMSegmenter
    
    # ER Network Segmentation
    er_seg = SIMERSegmenter(device='cuda')
    er_seg.load_model()
    results = er_seg.predict(volume)
    
    # Mitochondria Segmentation
    mito_seg = SIMMitoSegmenter(device='cuda')
    mito_seg.load_model()
    results = mito_seg.predict(volume)
    
    # General WFM Segmentation
    wfm_seg = WFMSegmenter(device='cuda')
    wfm_seg.load_model()
    results = wfm_seg.predict(volume)

For detailed API documentation, see :doc:`segmentation`.

Example Usage
-------------

**ER Network Segmentation**

.. code-block:: python

    from ipa.processing.segmentation import SIMERSegmenter
    from ipa.data_loader import UniversalDataLoader
    
    # Load SIM ER data
    er_data = UniversalDataLoader.load_data("SIM_ER.tif")
    
    # Segment ER network
    segmenter = SIMERSegmenter(device='cuda')
    segmenter.load_model()
    results = segmenter.predict(er_data)
    er_mask = results['er_mask']

**ISG Sphere Detection**

.. code-block:: python

    from ipa.processing.segmentation import SIMISGSegmenter
    
    # Load ISG immunofluorescence data
    isg_data = UniversalDataLoader.load_data("SIM_ISG.tif")
    
    # Detect spherical granules
    segmenter = SIMISGSegmenter(device='cuda')
    segmenter.load_model()
    results = segmenter.predict(isg_data)
    isg_mask = results['isg_mask']
    print(f"Detected {num_spheres} ISG spheres")

**Nucleus and Cell Segmentation**

.. code-block:: python

    from ipa.processing.segmentation import segment_cell_shape, segment_nucleus
    
    # Cell shape segmentation using actin
    cell_mask = segment_cell_shape(
        actin_data,
        threshold=None,  # Automatic threshold
        min_size=1000
    )
    
    # Nuclear segmentation
    nucleus_data = UniversalDataLoader.load_data("SIM_raw_N.tif")
    labeled_nuclei = segment_nucleus(nucleus_data)
    num_nuclei = np.max(labeled_nuclei)
    print(f"Detected {num_nuclei} nuclei")