OtherDisability and rehabilitation. Assistive technology2018

Detecting destabilizing wheelchair conditions for maintaining seated posture.

Anna Crawford, Kiley Armstrong, Kenneth Loparo and 2 others

PMID 28366027

WHAT IT FOUND

A chest-worn sensor classified wheelchair collisions and rough terrain with 95.8% accuracy in one user, but missed a bump.

Wheelchair sensors alone failed badly, misclassifying level ground as collisions. This is a single-case technical test, not clinical evidence.

Key findings

01Chest accelerations alone classified sudden stops, rough terrain and level ground correctly 100% of the time, with one bump misclassified as level ground.

02Wheelchair accelerations alone misclassified four of six rough terrain and six of six level ground trials as sudden stops, resulting in 58.3% overall accuracy.

03Sudden stops and collisions were correctly classified 100% of the time across all three sensor configurations.

STILL TO COME

How it was doneWhat they found

Read the rest of this summary

You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.

Already have one?

What it does not show

The study involved only one subject with a specific injury level (T11), so results cannot be generalized to other wheelchair users with different physical abilities or seating configurations. Only six trials per condition were collected, which is a very small sample size for training and validating a machine learning classifier. The conditions were simulated in a controlled laboratory environment using dowels and a barrier, which may not accurately reflect the complexity and unpredictability of real-world outdoor surfaces. The study tested detection algorithms, not actual safety interventions or assistive devices, so it does not demonstrate that falls can be prevented.

Declared interests

The authors report no conflicts of interest. The study was supported by the US Department of Veterans Affairs.

The easy way to misread this

Do not assume this technology is ready for clinical use or that it prevents falls. This was a single-subject laboratory test of a detection algorithm. It did not test an actual safety device, and the high accuracy for collisions does not mean the system works reliably for subtle hazards like bumps, which it missed in the best-performing configuration.

Read it on PubMed →


The study

Participants
1
Certainty of evidence
Low

Browse

    Cite

    Anna Crawford, Kiley Armstrong, Kenneth Loparo, et al. Detecting destabilizing wheelchair conditions for maintaining seated posture. Disability and rehabilitation. Assistive technology. 2018.

    Read the original