authors:
- {affiliation: Walter and Eliza Hall Institute of Medical Research, name: Pradeep Rajasekhar}
cite:
- {doi: 'https://doi.org/10.1242/jcs.261950', text: Sorensen et al., url: https://doi.org/10.1242/jcs.261950}
- {doi: 'https://doi.org/10.1038/s41592-021-01262-9', text: Gómez-de-Mariscal et al.
    Nat. Methods 2021}
- {text: Surehli et al. 2023, url: https://github.com/mukund-ks/DeepLabV3Plus-PyTorch}
config:
  deepimagej:
    allow_tiling: true
    model_keys: null
    prediction:
      preprocess:
      - {kwargs: im_preprocessing.ijm, spec: 'ij.IJ::runMacroFile'}
    pyramidal_model: false
    test_information:
      inputs:
      - name: test-input.npy
        pixel_size: {x: 0.378, y: 0.378, z: 1.0}
        size: 1024 x 1024 x 1 x 3
      memory_peak: null
      outputs:
      - {name: test-output.npy, size: 768 x 768 x 1 x 1, type: image}
      runtime: null
covers: [cover.png]
description: DL model to segment neuronal ganglia in the gut. The model expects an RGB image where neurons (Hu) are in magenta and a second marker that labels the ganglia in magenta (ChAT, nNOS, PGP9.5). Please note that im_preprocessing.ijm macro needs to be run first before running the model.
documentation: README.md
format_version: 0.4.9
inputs:
- axes: bcyx
  data_range:
  - -.inf
  - .inf
  data_type: float32
  name: input
  preprocessing:
  - kwargs:
      axes: xy
      gain: 0.00392156862
      offset: 0.0
    name: scale_linear
  - kwargs:
      axes: xy
      eps: 1e-06
      mean:
      - 0.485
      - 0.456
      - 0.406
      std:
      - 0.229
      - 0.224
      - 0.225
      mode: fixed
    name: zero_mean_unit_variance
  shape: [1, 3, 1024, 1024]
license: CC-BY-4.0
links: [deepimagej/deepimagej]
name: 2D_Ganglia_RGB_v3
outputs:
- axes: bcyx
  data_range: [0.0, 1.0]
  data_type: float32
  halo: [0, 0, 64, 64]
  name: output0
  postprocessing:
  - kwargs: {threshold: 0.6}
    name: binarize
  shape:
    offset: [0.0, -1.0, 0.0, 0.0]
    reference_tensor: input
    scale: [1.0, 1.0, 1.0, 1.0]
sample_inputs: [sample_input_0.tif]
sample_outputs: [sample_output_0.tif]
tags: [ganglia-segmentation, enteric-neuron, gut, gut wall, neuron, pytorch, ganglia,
  FPN, resnet101, binary-segmentation]
test_inputs: [test-input.npy]
test_outputs: [test-output.npy]
timestamp: '2024-11-26T13:59:58.619758'
type: model
weights:
  torchscript: {pytorch_version: 2.4.1+cpu, sha256: 1c59382b776acc2beed84bc2309e29f474aa69ff6356ec2248a04854d4feca7b,
    source: best_model_torchscript.pt}