A hybrid neural network approach to gait analysis for remote patient care

Shahzaib Hamid; Ikramullah Khosa; Muhammad Aksam Iftikhar; Muhammad Azfar Yaqub; Dongkyun Kim; Muhammad Rehan Usman

doi:10.1016/j.icte.2026.01.012

A hybrid neural network approach to gait analysis for remote patient care

Shahzaib Hamid
, Ikramullah Khosa
, Muhammad Aksam Iftikhar
, Muhammad Azfar Yaqub
, Dongkyun Kim
, Muhammad Rehan Usman

COMSATS University Islamabad
Free University of Bozen-Bolzano
Kyungpook National University

Research output: Contribution to journal › Article › peer-review

Abstract

Computer Vision based Gait Analysis (GA) has evolved Remote Patient Monitoring (RPM), by estimating key gait parameters (GPs) from live videos. Estimation of GPs is evaluated through four different algorithms, including state-of-the-art hybrid LSTM-Transformer. Three accuracy parameters are being utilized i.e., correlation, mean absolute error (MAE) and mean absolute percentage error (MAPE). Hybrid LSTM-Transformer stands out with MAEs of 0.1397 m/s for speed, 0.090 steps/s for cadence, 5.532° for knee flexion, and 6.25 for GDI; MAPE values of 20.81 % (speed), 11.00 % (cadence), 3.50 % (knee flexion), and 8.00 % (GDI); and correlations of 0.791, 0.790, 0.851, and 0.753 for the GPs estimation.

Original language	English
Number of pages	6
Journal	ICT Express
DOIs	https://doi.org/10.1016/j.icte.2026.01.012
Publication status	E-pub ahead of print - 28 Jan 2026

Keywords

Computer vision (CV)
Gait analysis (GA)
Multimodal hybrid models
Remote Patient monitoring (RPM)

Access to Document

10.1016/j.icte.2026.01.012Licence: CC BY

Cite this

@article{43324161d816442abb1d715f3dddf850,

title = "A hybrid neural network approach to gait analysis for remote patient care",

abstract = "Computer Vision based Gait Analysis (GA) has evolved Remote Patient Monitoring (RPM), by estimating key gait parameters (GPs) from live videos. Estimation of GPs is evaluated through four different algorithms, including state-of-the-art hybrid LSTM-Transformer. Three accuracy parameters are being utilized i.e., correlation, mean absolute error (MAE) and mean absolute percentage error (MAPE). Hybrid LSTM-Transformer stands out with MAEs of 0.1397 m/s for speed, 0.090 steps/s for cadence, 5.532° for knee flexion, and 6.25 for GDI; MAPE values of 20.81 \% (speed), 11.00 \% (cadence), 3.50 \% (knee flexion), and 8.00 \% (GDI); and correlations of 0.791, 0.790, 0.851, and 0.753 for the GPs estimation.",

keywords = "Computer vision (CV), Gait analysis (GA), Multimodal hybrid models, Remote Patient monitoring (RPM)",

author = "Shahzaib Hamid and Ikramullah Khosa and Iftikhar, \{Muhammad Aksam\} and Yaqub, \{Muhammad Azfar\} and Dongkyun Kim and Usman, \{Muhammad Rehan\}",

year = "2026",

month = jan,

day = "28",

doi = "10.1016/j.icte.2026.01.012",

language = "English",

journal = "ICT Express",

issn = "2405-9595",

publisher = "Korean Institute of Communications and Information Sciences",

}

TY - JOUR

T1 - A hybrid neural network approach to gait analysis for remote patient care

AU - Hamid, Shahzaib

AU - Khosa, Ikramullah

AU - Iftikhar, Muhammad Aksam

AU - Yaqub, Muhammad Azfar

AU - Kim, Dongkyun

AU - Usman, Muhammad Rehan

PY - 2026/1/28

Y1 - 2026/1/28

N2 - Computer Vision based Gait Analysis (GA) has evolved Remote Patient Monitoring (RPM), by estimating key gait parameters (GPs) from live videos. Estimation of GPs is evaluated through four different algorithms, including state-of-the-art hybrid LSTM-Transformer. Three accuracy parameters are being utilized i.e., correlation, mean absolute error (MAE) and mean absolute percentage error (MAPE). Hybrid LSTM-Transformer stands out with MAEs of 0.1397 m/s for speed, 0.090 steps/s for cadence, 5.532° for knee flexion, and 6.25 for GDI; MAPE values of 20.81 % (speed), 11.00 % (cadence), 3.50 % (knee flexion), and 8.00 % (GDI); and correlations of 0.791, 0.790, 0.851, and 0.753 for the GPs estimation.

AB - Computer Vision based Gait Analysis (GA) has evolved Remote Patient Monitoring (RPM), by estimating key gait parameters (GPs) from live videos. Estimation of GPs is evaluated through four different algorithms, including state-of-the-art hybrid LSTM-Transformer. Three accuracy parameters are being utilized i.e., correlation, mean absolute error (MAE) and mean absolute percentage error (MAPE). Hybrid LSTM-Transformer stands out with MAEs of 0.1397 m/s for speed, 0.090 steps/s for cadence, 5.532° for knee flexion, and 6.25 for GDI; MAPE values of 20.81 % (speed), 11.00 % (cadence), 3.50 % (knee flexion), and 8.00 % (GDI); and correlations of 0.791, 0.790, 0.851, and 0.753 for the GPs estimation.

KW - Computer vision (CV)

KW - Gait analysis (GA)

KW - Multimodal hybrid models

KW - Remote Patient monitoring (RPM)

U2 - 10.1016/j.icte.2026.01.012

DO - 10.1016/j.icte.2026.01.012

M3 - Article

AN - SCOPUS:105029129838

SN - 2405-9595

JO - ICT Express

JF - ICT Express

ER -

A hybrid neural network approach to gait analysis for remote patient care

Abstract

Keywords

Access to Document

Fingerprint

Cite this