Disability classification using machine learning on functional assessment data

Résumé

Abstract The World Health Organization (WHO) estimates that over 1.3 billion people globally experience significant disabilities, a figure projected to exceed two billion by 2050. This increasing burden highlights the need for efficient tools to assist in disability assessment. This study leverages artificial intelligence (AI) to support healthcare professionals in conducting functional assessments and accurately classifying disability types. We used machine learning models, including Random Forest, Artificial Neural Networks (ANN), Support Vector Machines (SVM), Naïve Bayes, and k-Nearest Neighbor (KNN), to develop predictive models based on functional assessment data collected from 597,202 records across four regions. The best results were obtained using Random Forest with an accuracy of 92%, specificity of 98.6%, and an AUC of 96%. The ANN also performed well with an accuracy of 91.9%. This study demonstrates how AI can enhance disability classification by improving accuracy and reducing human error in medical assessments.

Citer ce document

Accès au document

Texte intégral en lecture en ligne, réservé aux abonnés SPHAERO et aux membres de l'institution. Se connecter

Auteur(s)

Statistiques

Consultations : 1

Téléchargements : 0