isoview 0.1.2

Multi-view light sheet microscopy image processing pipeline

IsoView Light Sheet Microscopy Pipeline

Acquisition Modes

Raw input is always flat — SPC##_TM#####_ANG###_CM#_CHN##_PH#.stack files in a single directory. Mode is auto-detected from SPC/TM counts:

Condition	Mode	Description
Multiple TM values	timelapse	time series, any number of specimens
Single TM + multiple SPC	tiled	spatial tiles, one timepoint
Single TM + single SPC	single	treated as timelapse with 1 timepoint

Filename Tags

Tag	Meaning	Raw	Corrected
SPC## / SPM##	specimen / tile	SPC00	SPM00
TM#####	timepoint	TM00000	TM000000
CM#	camera	CM0	CM00
CHN##	channel (raw)	CHN00, CHN01	—
VW##	view (corrected)	—	VW00 (z-scan), VW90 (y-scan)
ANG###	illumination angle	ANG000	— (dropped)
PH#	phase	PH0	— (dropped)

Output Layout

Output directory: {input_dir.name}{corrected_suffix} as sibling of input.

All below result path's use the default corrected_suffix .corrected.

Correction (correct_stack.py)

Mode	Path	Filename
Timelapse	root.corrected/SPM00/TM000000/	SPM00_TM000000_CM00_VW00.ome.tif
Tiled	root.corrected/SPM00/	SPM00_CM00_VW00.ome.tif

Each tile will have a SPM0N folder, where N is the tile index.

Fusion (multi_fuse)

Mode	Path	Filename
Timelapse	root.corrected/Results/MultiFused_adaptive/SPM00/TM000000/	SPM00_TM000000_CM00_CM01_VW00.ome.tif
Tiled	root.corrected/Results/MultiFused_adaptive/SPM00/	SPM00_CM00_CM01_VW00.ome.tif

Default pairs: [(0, 1), (2, 3)] — cameras 0,1 share CHN00/VW00; cameras 2,3 share CHN01/VW90. Only the second camera in each pair gets rotation/flip transforms.

Supported Output Formats

Format	Extension	Notes
OME-TIFF	.ome.tif	with metadata, optional resolution pyramids
Zarr v3	ome.zarr	OME-NGFF metadata
KLB	.klb	Keller Lab Block (bzip2)

Quickstart

1. isoview init <path/to/raw> — scaffolds a scripts/ folder next to the raw data with correct_stack.py, multi_fuse.py, pipeline.ipynb, and parameters.md.
2. Open scripts/ in VSCode and run correct_stack.py then multi_fuse.py, or open pipeline.ipynb and run both cells.
3. Each step appends its config + isoview_version to isoview_config.json next to the data, so every run is reproducible.

See examples/compression_demo.py for a full end-to-end run on 10 timepoints. Full parameter reference: parameters.md.

Usage

Entrypoints: pipeline/correct_stack.py (correction) and pipeline/multi_fuse.py (fusion).

from pathlib import Path
from isoview import ProcessingConfig, correct_stack, multi_fuse

config = ProcessingConfig(
    input_dir=Path(r"E:\isoview\dataset"),
    # specimens=None,                       # auto-detect from SPC## in filenames
    # timepoints=None,                      # auto-detect from TM## in filenames
    corrected_suffix=".corrected",          # output folder suffix
    specimen=0,                             # default specimen index
    camera_pairs=[(0, 1), (2, 3)],          # ortho camera pairs to fuse

    # output
    output_format="tif",                    # tif, zarr, or klb
    compression="zstd",                     # zstd, lzw, deflate, or None
    compression_level=3,                    # 1-22 for zstd, 1-9 for others

    # transforms (applied to second camera in each pair)
    rotation=0,                             # 0=none, 1=90cw, -1=90ccw
    flip_horizontal=False,
    flip_vertical=False,

    # correction
    median_kernel=(3, 3),                   # dead pixel filter, None to disable
    background_percentile=5.0,
    mask_percentile=1.0,
    segment_mode=1,                         # 0=none, 1=segment+mask, 2=masks, 3=global

    # fusion
    blending_method="adaptive",             # adaptive, geometric, average, wavelet
    blending_range=20,                      # transition zone width (z-planes)

    # per-specimen overrides (tiled mode)
    # tile_crops={"SPM00": {"crop_depth": {0: 450}}}
    # view_orientation={"SPM00": {"flip_axis": 1}, "SPM01": {"flip_axis": 0}}
)

correct_stack(config)
multi_fuse(config, estimate_params=True, apply_fusion=True)