Marked changes occur in the brain during people's lives, and individual rates of aging have revealed pronounced differences, giving rise to subject-specific brainprints that are the signature of the brain. These are shaped by a great variety of factors, both endogenous and exogenous. Accurate predictions of brain age (BA) can be derived from neuroimaging endophenotypes by using machine and deep learning (DL) techniques. Predictive models leading to accurate estimates while revealing which features contribute the most to final predictions are key to unveiling the mechanisms underlying the evolution of brain aging patterns. Explainable artificial intelligence (XAI) methods are emerging as enabling technology in different fields, and biomedicine is no exception. Within this framework, this article examines BA and presents a comprehensive review of recent advances in the exploitation of explainable machine learning (ML)/DL methods, highlighting the main open issues and providing hints for future directions.

Explainable Artificial Intelligence for Magnetic Resonance Imaging Aging Brainprints: Grounds and challenges

Boscolo Galazzo, Ilaria
;
Cruciani, Federica;Brusini, Lorenza;Salih, Ahmed;Storti, Silvia Francesca;Menegaz, Gloria
2022

Abstract

Marked changes occur in the brain during people's lives, and individual rates of aging have revealed pronounced differences, giving rise to subject-specific brainprints that are the signature of the brain. These are shaped by a great variety of factors, both endogenous and exogenous. Accurate predictions of brain age (BA) can be derived from neuroimaging endophenotypes by using machine and deep learning (DL) techniques. Predictive models leading to accurate estimates while revealing which features contribute the most to final predictions are key to unveiling the mechanisms underlying the evolution of brain aging patterns. Explainable artificial intelligence (XAI) methods are emerging as enabling technology in different fields, and biomedicine is no exception. Within this framework, this article examines BA and presents a comprehensive review of recent advances in the exploitation of explainable machine learning (ML)/DL methods, highlighting the main open issues and providing hints for future directions.
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: http://hdl.handle.net/11562/1060402
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? 1
social impact