Fidgety Philip and the Suggested Clinical Immobilization Test: Annotation data for developing a machine learning algorithm

The cartoon Fidgety Philip, the banner of Western-ADHD diagnosis, depicts a ‘restless’ child exhibiting hyperactive-behaviors with hyper-arousability and/or hypermotor-restlessness (H-behaviors) during sitting. To overcome the gaps between differential diagnostic considerations and modern computing methodologies, we have developed a non-interpretative, neutral pictogram-guided phenotyping language (PG-PL) for describing body-segment movements during sitting (Journal of Psychiatric Research). To develop the PG-PL, seven research assistants annotated three original Fidgety Philip cartoons. Their annotations were analyzed with descriptive statistics. To review the PG-PL's performance, the same seven research assistants annotated 12 snapshots with free hand annotations, followed by using the PG-PL, each time in randomized sequence and on two separate occasions. After achieving satisfactory inter-observer agreements, the PG-PL annotation software was used for reviewing videos where the same seven research assistants annotated 12 one-minute long video clips. The video clip annotations were finally used to develop a machine learning algorithm for automated movement detection (Journal of Psychiatric Research). These data together demonstrate the value of the PG-PL for manually annotating human movement patterns. Researchers are able to reuse the data and the first version of the machine learning algorithm to further develop and refine the algorithm for differentiating movement patterns.


GPU:
OpenPose recommends any GPU with at least 10GB VRAM.
Note: This application cannot be used on a Mac.

Video Upload Requirements
• The video must only have one person in it • The video must be in MP4 format and be ~30 frames per second.
• The video must be at least five seconds long. When you upload your patient video, the software will ask you to select an empty folder to save your results to. This is the folder you should select. set-up application before you upload a patient video into the application, there is some prep work to be done.
Here is a list of files that will be created inside this folder.

Note:
There are a few more items that could be added to this folder if the deidentification feature is used.
The deidentification feature will be covered later.

Review the two video upload options (See
• Upload New Video to Process. This option is intended for new video content.
• Upload Existing File. This option is intended for videos you have previously uploaded to the application.
Make sure to follow the Video Upload Requirements outlined on page 3.

Select an upload option.
If you're Uploading a New Video to Process: • Select the folder you wish to save the data to. (Review set-up instructions on page 4) • Press continue and let the application process. While the video is processing, your computer will divert most of its computing power to the application. It is recommended to run the software overnight. There will be a loading screen indicating the percentage of completion.
Note: It will take a long time to process depending on the specification of the computer. Computers without a dedicated GPU that has at least 8GB of memory will take hours just to process a one-minute video. The reason is because OpenPose requires an exorbitant amount of processing power to run.
If you're Uploading an Existing File: • Select the folder where you saved the video previously and the application will load it into the main screen. This option can be found at the top left of the screen labeled 'De-identification'. When the button is pressed, a dialog box will pop-up detailing the de-identification process as well as displaying an option to proceed with the de-identification.
After the de-identification completes, the newly blurred video will be loaded into the video player. (The skeleton version would also be blurred as well).
The user can find the blurred videos in the saved folder. The video and skeleton path in the settings will also be changed to the paths of the blurred videos. If you want to load the original video, you can change the video path to the original path.
Note: It will take some time (around eleven minutes for a one-minute video, but this will vary depending on the specifications of your hardware) for the de-identification process to complete depending on the length of the video.

Warning:
The algorithm is not perfect.
There might be frames where the faces are not blurred. We recommend users view the processed video to ensure full de-identification before dissemination.
If there are gaps in blurring, use other software (e.g., Adobe Premiere, Camtasia) to manually add blurring.
de-identify video