cortex.machinemedicine.com
Navigation
Sign out
My Flows
+ Publish new Flow
Flow Settings
Select a Flow to edit.
General
Inputs
Outputs
Control things such as Flow display name and description.
Average Tapping Velocity
Custom analysis for finger tapping velocity.
Save
Triggered Flows
Filter by Flow Name...
Filter by Tag...
Filter by Assessment ID...
All Statuses
StatusFlow NameVersionAssessment IDCreated At (UTC)Updated At (UTC)
No triggered flows found.
Source Files
question4
00:07 / 00:19
ROI selection
start time
MM:SS.xxx
end time
MM:SS.xxx
Cortical Flows
NameDescriptionVersion
Activity Recognition Identifies selected patient activities in a video. default
De-Identification Anonymises faces of subjects detected in the video. default
Facemesh Estimation Identifies facial landmarks & head rotation angles. default
PD Model Inference Severity score estimates for Finger Tapping – Left. FT_L
Pose Estimation Detects body poses and groups into unique subjects. default
Average Tapping Velocity Custom analysis for finger tapping velocity. v1.0
Process
StatusFlow NameSource FileProgress
Running Pose Estimation question4
0%Est. time remaining: 00:04
Running Average Tapping Velocity question4
0%Est. time remaining: 00:03
Available Outputs
NameDescriptionFlow versionROICreated (UTC)
Pose Estimation Pose landmarks of all subjects. v1.0:defaultFullJul 02, 2026
Average Tapping Velocity Tapping velocity metrics. v1.0:v1.0FullJul 03, 2026
Overlay Outputs
Clear Overlay
Publish new Flow
To publish your Flow, you'll need to install our helper Python package.
Install
In order to communicate with the publishing endpoint you will need an API token. You can create one below. Once generated, copy it to your local environment. The token will expire after 24 hours.
Generate API Token
Terminal — zsh
$
Cortex Loom installed
average_tapping_velocity.py — cortex-loom
Explorer
CORTEX-LOOM
average_tapping_velocity.py
average_tapping_velocity.py
1
import numpy as np
2
from cortex_loom import FlowBuilder
3
from cortex_loom.types import PoseObject
4
def average_tapping_velocity(pose: PoseObject):
5
"""Compute the velocity [pixels/s] of left hand finger tapping."""
6
xy_index = pose.pose_array["LeftHandIndexFingerTip"][:, :2]
7
xy_thumb = pose.pose_array["LeftHandThumbTip"][:, :2]
8
sec_time = pose.timestamps / 1000
9
distance = np.sqrt(((xy_index - xy_thumb) ** 2).sum(axis=-1))
10
velocity = np.abs(np.diff(distance)) / np.diff(sec_time)
11
return np.mean(velocity)
12
builder = FlowBuilder(api_token=API_TOKEN)
13
builder.publish(func=average_tapping_velocity)
main
Ln 13, Col 47 UTF-8 Python 3.11.4
Flow published successfully
Average Tapping Velocity is now available in Cortex.