Skip to content

Processing a Study

Run Congifuration

Opening one of the "accordion" panels will reveal the following view:

RunEASL_StudyConfig

As suggested by the headings, the panel is subdivided into 3 sections:

  • Study Settings, where you can specify which study to process, how many resources (CPU cores) to allocate, and which ExploreASL module to run in the first place.
  • Run Control, where you can initiate the processing of the study, pause it, resume from pause, or cancel it. This section also displays the current progress completion of the pipeline.
  • Progam Feedback, much like in running the Import Module, feedback is given to the user as to the progress of the pipeline. Note that the feedback is more succinct, as the pipeline generates a lot of log statements that would otherwise clutter the program feedback panel. In addition, when certain quality-control steps complete, axial and coronal slices of the processed data are displayed in the panel as images.

Study Settings

The fields in this section are as follows:

  • Study Root Path: The root path of the study to process. This is the folder that contains the dataPar.json file generate in the previous module.
  • Number of Cores Allocated for this Study: Self-explanatory. Note that certain numbers may be disabled if the program detects that another study already has a certain number of cores allocated to it.
  • Modules to Run: The particular ExploreASL module to run. Options include:
  • Strucural Module
  • ASL Module
  • Structural & ASL Module
  • Population Module ❗ NOTE ❗ If selecting the lattermost, you may notice the run button is greyed out. This is probably due to the fact that only a single core can be allocated to running a Population Module.

Running State

Once you have specified the study root path, the ExploreASL module to run, and have clicked Run Study, the panel will change to the following view:

RunEASL_StartedRun

At this point in time, you cannot alter the study configuration or change the number of studies to process. This is, again, for the sake of preventing over-extension of your workstations resources and to prevent the pipeline from crashing.

As the pipeline progresses, you will be able to observe that under the derivatives/ExploreASL/lock folder, there are folders and files being generated that correspond to the current state of the pipeline. As checkpoints are passed in the pipeline, certain .status files are created to indicate the passing of that checkpoint.

RunEASL_StatusFileAndImageFeedback

⚠ Warning ⚠

If you manipulate this lock structure or the .status files, you may cause errors in the pipeline. This is purely for your awareness and is not a feature that is intended to be used by the user.

What if something went wrong during an earlier step?

Before running, the program re-validates the dataPar.json file and the rawdata folder in order to minimize the chance of a user-based error propagating into the analysis pipeline. If you attempt to run a study with an invalid setup, you will be greeted with an error message explaining the problem. For example, if we attempt to run the tampered dataPar.json file from the [DataPar Module](../3_DataPar/4_Saving_And_Loading.md#loading, we will be greeted with the following error message:

RunEASL_WhatIfDataParIsWrong

Completing a Run

If a run is completed successfully, you will be greeted with the following message:

RunEASL_FinishedWithSuccess

And you will be able to either start a new run on another module, or move onto Data Visualization.

Alternatively, you may encounter an error during the run. If this is the case, you will be greeted with the following message:

RunEASL_FinishedWithError

As you can see, the error message features an overview of the number of steps that failed and a summary of the earlier step that failed for which subject/visit/session. From here, it is up to the user to review the log files found under derivatives/ExploreASL/logs detailing exactly what went wrong or to review their dataset to see if there are corrupt files or other issues that may have caused the pipeline to fail.

Completing a Study Analysis

A study is considered analyzed when all 3 modules (Structural, ASL, Population) have been successfully completed. It is not possible to perform the next module, Data Visualization, for a particular study until all 3 modules have been completed.