Cognitum/Gait Analysisv1.3.0

Gait Analysis Overview

CSI-based walking pattern analysis for Parkinsonian gait, post-stroke asymmetry, and fall-risk scoring across the patient cohort.

Active patients
24
3 new this week
Sessions today
17
+22% vs avg
Avg gait score
72
↑ 4 pts (7d)
High fall risk
5
need follow-up
Active alerts
3
unacknowledged
Sensors online
8 / 9
ESP32-S3 fleet
Cohort gait score — 30d trend
Median, P25–P75 envelope across all active patients
Median score P25–P75 range Sessions / day
Condition distribution
Active classifications
Recent alerts
Last 24 hours
View all →

Live Monitor

Real-time CSI motion + presence stream from a paired sensor with per-stride gait parameters and live scoring.

CSI signal — channel state response
ESP32-S3 · sensor S-04 living-room · 100 Hz
streaming
Cadence
108
steps / min
Stride length
1.18
m
Gait velocity
1.06
m/s
Stride time CV
3.2
%
Asymmetry idx
0.04
0 = symmetric
Double support
22.4
% of cycle
Current scores
Computed every 10 strides
72
gait_score
30
fall_risk
Parkinsonian likelihood12%
Post-stroke asymmetry8%
Freezing-of-gait events0 / session
Event stream
Cog events emitted during this session
[ready] waiting for first stride window…

Patients

Cohort overview with latest gait score, fall risk, and primary classification.

IDPatientAgeClassification Gait scoreFall riskLast sessionSensor

Sessions

Recorded walking sessions across the cohort with computed gait parameters and classifications.

WhenPatientDurationStridesCadenceVelocityAsymmetryScoreEvents

Analytics

Aggregate trends and per-parameter distributions for clinical reporting.

Velocity distribution
m/s across cohort
Cadence distribution
steps/min
Asymmetry index
0 = symmetric
Fall risk by age band
Concentric arcs · mean fall_risk per cohort
Event counts (7d)
Cog events triggered

Alerts

Real-time gait events emitted by the cog. Acknowledge to dismiss; escalate to notify clinical staff.

Devices

ESP32-S3 sensors providing CSI motion and presence input to the gait analysis cog.

Hardware requirements
Per ADR cog spec
SoC
ESP32-S3
required
Input
CSI
motion · presence
Sample rate
100 Hz
CSI frames
Image
32 KB
compressed
Difficulty
Hard
deployment

Settings

Cog configuration, event thresholds, and integration endpoints.

Event thresholds
When to emit cog events
Classification model
Temporal gait parameters → label
Cog manifest
Source: cognitum-one/cogs/src/cogs/gait-analysis
name: gait-analysis
version: 1.3.0
category: health
hardware: ESP32-S3
inputs: [csi_motion, presence]
events:
  - gait_score
  - fall_risk
  - asymmetry_detected
  - parkinsonian_gait
size: 32 KB
difficulty: hard
description: |
  Analyzes walking patterns to detect Parkinsonian gait, post-stroke
  asymmetry, and generates fall-risk scores using temporal gait parameters.