CaliAli parameter settings

The CaliAli pipeline uses a structured approach to initialize, configure, and manage parameters across different processing stages. The function CaliAli_parameters() acts as the central hub for handling parameters, ensuring modularity and flexibility.

📌 Key Features of CaliAli Parameter Management

1️⃣ CaliAli submodules:

CaliAli organizes its processing pipeline into separate submodules, each with its own set of parameters:

Downsampling
Preprocessing (detrending, background processing)
Motion Correction (rigid and non-rigid)
Inter-Session Alignment
CNMF-E (Calcium signal extraction and demixing)

The parameters utilized by each submodule is determine on a CaliAli_options structure:

◼ CaliAli_options
├─ downsampling
│  ├─ BVsize
│  │  Value: [1.5 2.25]
│  ├─ file_extension
│  │  Value: 'avi'
│  └─ ...
├─ preprocessing
│  ├─ dendrite_filter_size
│  │  Value: [0.5 0.6 0.7 0.8]
│  ├─ dendrite_theta
│  │  Value: 30
│  └─ ...
└─ ...

The default set of parameters can be obtained with CaliAli_demo_parameters(); function.

2️⃣ Flexible Input Handling

CaliAli parameters can be set in Three ways:

Default Initialization:
If no inputs are provided, default parameters are loaded:
```
CaliAli_options = CaliAli_parameters();
```

Custom Parameter Structure:
Users can provide an existing structure to modify specific parameters:

param.sf=20 % custom sample frequency
param.gSig=3  %custom neuron filter size
CaliAli_options= CaliAli_parameters(param);

Custom name-value pairs :
Users can provide name-value pairs

CaliAli_options = CaliAli_parameters('sf',20); % set sampling frequency to 20

The function CaliAli_demo_parameters() demonstrates how to create this structure. The recommended approach for analyzing your own data is to modify and duplicate this code to suit your needs: open CaliAli_demo_parameters.m

Adjusting CaliAli Parameters.

CaliAli requires setting 33 parameters. However, in practice you only need to strictly focus on three:

Frame rate: sf
Neuron filtering size: gSig which correspond to 1/5 of the average neuron size in pixels. If left empty, CaliAli now assigns gSig = 5 / spatial_ds, which matches the demo data and works well for non-downsampled videos.
Blood vessel size: : BVsize which correspond to the size of blood vessels in pixels [min, max]

Follow the Demo Parameter_selection_demo.mlx to learn how to define these parameters interactively using the BV_app and NeuronSize_app:

    open Parameter_selection_demo.mlx

For advanced users, a detailed description of other parameters and methods for setting them can be found in: Parameter Index