Cardiovascular diseases are a group of heart and blood vessels disorders that are one of the main causes of dead and invalidity. Prevention is fundamental to diagnose the condition in its early stage. Machine learning techniques have been proven to be useful tools to support clinicians in their daily tasks. Particularly the big availability of digital clinical information has made possible to model decision support tools that simulate the human reasoning and knowledge. However, medical data, and clinicians' reasoning are inherently uncertain and vague. Fuzzy logic inference systems (FIS) have been proven to be effective in representing medical knowledge and reasoning. However they suffer by the curse-of-dimensionality. To overcome this problem hierarchical fuzzy inference system (HFIS) are used. In this paper, we propose a HFIS for cardiovascular risk level prediction. Vital signs, collected through non-invasive technologies, are used to derive the fuzzy rules. Comparison between plain FIS and hierarchical FIS show an improvement on the classification performance, together with a significant model simplification, that is rules more easily interpretable.

A Hierarchical Fuzzy System for Risk Assessment of Cardiovascular Disease

Gabriella Casalino;Vincenzo Pasquadibisceglie;Gianluca Zaza
2020-01-01

Abstract

Cardiovascular diseases are a group of heart and blood vessels disorders that are one of the main causes of dead and invalidity. Prevention is fundamental to diagnose the condition in its early stage. Machine learning techniques have been proven to be useful tools to support clinicians in their daily tasks. Particularly the big availability of digital clinical information has made possible to model decision support tools that simulate the human reasoning and knowledge. However, medical data, and clinicians' reasoning are inherently uncertain and vague. Fuzzy logic inference systems (FIS) have been proven to be effective in representing medical knowledge and reasoning. However they suffer by the curse-of-dimensionality. To overcome this problem hierarchical fuzzy inference system (HFIS) are used. In this paper, we propose a HFIS for cardiovascular risk level prediction. Vital signs, collected through non-invasive technologies, are used to derive the fuzzy rules. Comparison between plain FIS and hierarchical FIS show an improvement on the classification performance, together with a significant model simplification, that is rules more easily interpretable.
2020
978-1-7281-4384-2
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/11586/314608
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 13
  • ???jsp.display-item.citation.isi??? 8
social impact