Using AI to Detect Sarcopenia and Cognitive Decline Through Gait Analysis
Author Information
Author(s): Aznar-Gimeno Rocío, Perez-Lasierra Jose Luis, Pérez-Lázaro Pablo, Bosque-López Irene, Azpíroz-Puente Marina, Salvo-Ibáñez Pilar, Morita-Hernandez Martin, Hernández-Ruiz Ana Caren, Gómez-Bernal Antonio, Rodrigalvarez-Chamarro María de la Vega, Alfaro-Santafé José-Víctor, del Hoyo-Alonso Rafael, Alfaro-Santafé Javier
Primary Institution: Instituto Tecnológico de Aragón (ITA)
Hypothesis
Can artificial intelligence algorithms based on gait analysis effectively detect sarcopenia and cognitive decline in older adults?
Conclusion
The study shows that combining sensor and computer vision data for gait analysis can effectively screen for sarcopenia and cognitive decline.
Supporting Evidence
- Models based on sensor data and computer vision achieved high predictive accuracy for cognitive decline.
- The best model for cognitive decline had an F1-score of 0.914, with 95% sensitivity.
- The combined technologies model for sarcopenia showed an F1-score of 0.748 with 100% sensitivity.
- Participants were classified based on established criteria for sarcopenia and cognitive decline.
Takeaway
Researchers used technology to watch how older people walk to find out if they have muscle loss or memory problems.
Methodology
A cross-sectional case-control study was conducted with gait analysis using sensors and computer vision to classify participants based on sarcopenia and cognitive decline criteria.
Potential Biases
Potential bias due to convenience sampling and the presence of multimorbidity in participants.
Limitations
The sample size for sarcopenia was small and imbalanced, and the cross-sectional design limits causal inferences.
Participant Demographics
Participants were aged 60 years or older, with 57.14% being female and a median age of 70.5 years.
Statistical Information
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website