webfm

A web-based functional mapping utility.

...

Prerequisites

In order to run WebFM, you'll need:

1. A working Node installation
2. ( ... BCI2000 / BCI2000Web dependencies ... )

...

Installation

Installing WebFM requires three steps:

1. Cloning the WebFM git repository
2. Installing the Node dependencies via npm for the server and for the client
3. Building the client application scripts

Git

To clone the WebFM repo, navigate to wherever you'd like WebFM to be installed and run

git clone https://github.com/cronelab/webfm.git

Node

Let {webfm} denote the path to the root of the WebFM repository. First, navigate to {webfm} and run

npm install

This will install the dependencies for the WebFM server.

For the client, navigate to {webfm}/app and once again run

npm install

Build

To build the client application scripts, navigate to {webfm}/app and run

./build

To test that everything has installed properly, navigate to {webfm} and run

node webfm

You should see

Serving {webfm}/public on 54321:tcp

If you encounter issues starting the server on the default port, you can specify a port that works with the -p option:

node webfm -p 8080

Usage

...

Data formats

WebFM has two distinct ways of packaging data: datasets, which encapsulate one viewable record—one spatial map, one event-related activity/connectivity map, etc.—and bundles, which wrap together multiple related datasets into a unified unit.

Dataset

Datasets, which have the .fm file extension, are simply JSON files with a designated structure. Datasets contain two kinds of information: metadata, which are extra bits of information that give the data needed (or useful) context; and contents, which are the actual data points of interest.

Metadata

The fields in a dataset's metadata—stored in the dataset's metadata field—provide the information necessary for the WebFM viewer to properly render the enclosed data. All fields of the metadata are optional; however, for certain kinds of data, certain combinations of omitted metadata may make it so that WebFM does not have sufficient information to render the data properly.

Metadata may also exist in standalone files (usually named .metadata) for use, e.g., with the _import field described below. For example, bundles have a unified .metadata file _imported by all of the bundle members, to prevent duplication of large chunks of common metadata. As another example, WebFM stores a .metadata file for each subject, which contains the subject-specific metadata necessary to do online mapping (brainImage, sensorGeometry, etc.).

The following are the (presently) meaningful metadata fields, as well as example values:

####_import

Examples: "./.metadata", ["/path/number/1", "/path/number/2", ... ]

A string (or array of strings) specifying files to import and incorporate as additional metadata. If an array, files will be loaded in the specified order. The server performs these imports when the dataset is requested by a client; as such, this field is never present in the final dataset object served to the client.

####_export

Example:

{
    "brainImage": "./PYXXNXXX-Brain.png",
    "sensorGeometry": "./PYXXNXXX-Sensors.csv"
}

The _export field is used to denote operations that should be performed by the export utility to produce a final dataset—for example, base64-encoding an image, or converting a spreadsheet to a u-v map. This field should not be present in a fully formed dataset.

####subject

Example: "PYXXNXXX"

The identifier of the subject from whom this data originated

####brainImage

Example: "data:image/.png;base64,iVBORw0KG..."

A string containing the base64-encoded binary image data to be used when displaying the data on a spatial map

####sensorGeometry

Example:

{
    "CH1": {
        "u": 0.1,
        "v": 0.4
    },
    ...
}

A mapping from channel names to u-v coordinates (ranging from 0 to 1 on each axis) for placing electrodes on the brainImage

####montage

Example: ["CH01", "CH02", "CH03", ... ]

A list of channel names, specifying both which channels should be displayed and, when relevant, in what order. The WebFM frontend automatically populates this from whatever other details are available if it is not specified.

####setting

Example:

{
    "task": "PictureNaming",
    "stimulusType": "animals"
}

An object providing details on the context in which the data were collected

####kind

Example: "event related potential"

A human-readable description of what kind of data is in this dataset

####labels

Example: ["timeseries", "potential", "bipolar"]

An array of strings providing information about how the data should be displayed or interpreted

####baselineWindow

Example:

{
    "start": -0.8,
    "end": -0.2
}

For timeseries data, tart and end, in seconds, of the epoch used as the statistical "null" period; should correspond to contents.times

Contents

...

####values

Example:

{
    "CH01": [0.0, 0.01, ... ],
    ...
}

Raw values to be plotted as-is, formatted as an object mapping channel names to values. When this field is present, it is implied that no statistical computations should be performed before displaying the data.

####times

Example: [0.0, 0.001, 0.002, ...]

For timeseries data, an array containing the time (in seconds) corresponding to each sample in the arrays of values or stats.estimators.{estimator}.

####stats

An object describing the statistical properties of the data, allowing the WebFM frontend to perform its own statistical work (e.g., multiple comparison correction). The structure of this object is described in more detail below.

####trials

An array of objects, each depicting one realization of the data encapsulated by the dataset. The structure of these objects is described in more detail below.

Stats

...

####distribution

Meaningful values: "gaussian", "p-value"

The kind of distribution described by the entry in the estimators and baseline fields

####estimators

Example:

{
    "mean": {
        "CH01": [ ... ],
        ...
    },
    "variance": {
        "CH01": [ ... ],
        ...
    },
    ...
}

...

Each distribution implies its own set of fields that should be present in estimators. The following distributions are supported:

distribution value	Meaning	estimators fields
"gaussian"	Normal distributions	"mean", "variance", "count"
"p-value"	Raw p-values from some unspecified hypothesis test	"value" (to be displayed), "p"

####baseline

Example:

{
    "mean": {
        "CH01": 0.58,
        ...
    },
    "variance": {
        "CH01": 1.60,
        ...
    },
    ...
}

..

Trials

...

Each trial has the following fields:

####setting

Example:

{
    "stimulusCode": 0
}

...

####values

Example:

{
    "CH01": [0.0, 0.01, ... ],
    ...
}

...

Bundle

Bundles, which have the .fmbundle extension, are simply directories with a designated structure to allow grouping of related .fm files.

Bundles are currently under development, and will be available in a future release.

Common data layouts

...

Event-related band power

...

Resting state connectivity

...

Event-related connectivity

...

Event-related potential

...

License

...