Healthcare 5.0 is a research trend promoting a patient-centric approach leveraging Artificial Intelligence (AI)-based solutions. It aims to enhance care and quality of life for all patients, including those with a disability. However, when applied to the health sector, AI may be perceived as not transparent: the Explainable AI (xAI) paradigm attempts to solve this issue by providing more understandable, reliable, and human-interpretable AI-based applications. In a field such as disability – characterized by various impairments, limitations in performing activities, or other kinds of restrictions – the possibility to rely on computable representations of domain knowledge in the form of ontologies can support the development of xAI for healthcare. This work proposes a systematic literature review to identify which disabilities are currently represented in domain ontologies, examining which applicative contexts the ontologies were developed for. This review also investigates how the domain ontologies are modelled, underlining several relevant aspects that may foster their adoption in xAI systems. The review process results allow for shedding light on the main disabilities represented in ontologies, tracing the research trends that were at the basis of their development. Results also enable the identification of research lines that can support semantic interoperability – thus enabling ontologies to play a significant role in explaining decision processes performed by AI-based systems in healthcare.

A review of domain ontologies for disability representation

Spoladore D.;Sacco M.;Trombetta A.
2023-01-01

Abstract

Healthcare 5.0 is a research trend promoting a patient-centric approach leveraging Artificial Intelligence (AI)-based solutions. It aims to enhance care and quality of life for all patients, including those with a disability. However, when applied to the health sector, AI may be perceived as not transparent: the Explainable AI (xAI) paradigm attempts to solve this issue by providing more understandable, reliable, and human-interpretable AI-based applications. In a field such as disability – characterized by various impairments, limitations in performing activities, or other kinds of restrictions – the possibility to rely on computable representations of domain knowledge in the form of ontologies can support the development of xAI for healthcare. This work proposes a systematic literature review to identify which disabilities are currently represented in domain ontologies, examining which applicative contexts the ontologies were developed for. This review also investigates how the domain ontologies are modelled, underlining several relevant aspects that may foster their adoption in xAI systems. The review process results allow for shedding light on the main disabilities represented in ontologies, tracing the research trends that were at the basis of their development. Results also enable the identification of research lines that can support semantic interoperability – thus enabling ontologies to play a significant role in explaining decision processes performed by AI-based systems in healthcare.
2023
Disability; Domain ontologies; Explainable Artificial Intelligence; Knowledge-based systems; PRISMA (Preferred Reporting Items for Systematic reviews and meta-Analyses statement)
Spoladore, D.; Sacco, M.; Trombetta, A.
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11383/2171592
 Attenzione

L'Ateneo sottopone a validazione solo i file PDF allegati

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 8
  • ???jsp.display-item.citation.isi??? 5
social impact