Cargando…

Occlusion-Based Explanations in Deep Recurrent Models for Biomedical Signals

The biomedical field is characterized by an ever-increasing production of sequential data, which often come in the form of biosignals capturing the time-evolution of physiological processes, such as blood pressure and brain activity. This has motivated a large body of research dealing with the devel...

Descripción completa

Detalles Bibliográficos
Autores principales: Resta, Michele, Monreale, Anna, Bacciu, Davide
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2021
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8394492/
https://www.ncbi.nlm.nih.gov/pubmed/34441204
http://dx.doi.org/10.3390/e23081064
_version_ 1783743960938708992
author Resta, Michele
Monreale, Anna
Bacciu, Davide
author_facet Resta, Michele
Monreale, Anna
Bacciu, Davide
author_sort Resta, Michele
collection PubMed
description The biomedical field is characterized by an ever-increasing production of sequential data, which often come in the form of biosignals capturing the time-evolution of physiological processes, such as blood pressure and brain activity. This has motivated a large body of research dealing with the development of machine learning techniques for the predictive analysis of such biosignals. Unfortunately, in high-stakes decision making, such as clinical diagnosis, the opacity of machine learning models becomes a crucial aspect to be addressed in order to increase the trust and adoption of AI technology. In this paper, we propose a model agnostic explanation method, based on occlusion, that enables the learning of the input’s influence on the model predictions. We specifically target problems involving the predictive analysis of time-series data and the models that are typically used to deal with data of such nature, i.e., recurrent neural networks. Our approach is able to provide two different kinds of explanations: one suitable for technical experts, who need to verify the quality and correctness of machine learning models, and one suited to physicians, who need to understand the rationale underlying the prediction to make aware decisions. A wide experimentation on different physiological data demonstrates the effectiveness of our approach both in classification and regression tasks.
format Online
Article
Text
id pubmed-8394492
institution National Center for Biotechnology Information
language English
publishDate 2021
publisher MDPI
record_format MEDLINE/PubMed
spelling pubmed-83944922021-08-28 Occlusion-Based Explanations in Deep Recurrent Models for Biomedical Signals Resta, Michele Monreale, Anna Bacciu, Davide Entropy (Basel) Article The biomedical field is characterized by an ever-increasing production of sequential data, which often come in the form of biosignals capturing the time-evolution of physiological processes, such as blood pressure and brain activity. This has motivated a large body of research dealing with the development of machine learning techniques for the predictive analysis of such biosignals. Unfortunately, in high-stakes decision making, such as clinical diagnosis, the opacity of machine learning models becomes a crucial aspect to be addressed in order to increase the trust and adoption of AI technology. In this paper, we propose a model agnostic explanation method, based on occlusion, that enables the learning of the input’s influence on the model predictions. We specifically target problems involving the predictive analysis of time-series data and the models that are typically used to deal with data of such nature, i.e., recurrent neural networks. Our approach is able to provide two different kinds of explanations: one suitable for technical experts, who need to verify the quality and correctness of machine learning models, and one suited to physicians, who need to understand the rationale underlying the prediction to make aware decisions. A wide experimentation on different physiological data demonstrates the effectiveness of our approach both in classification and regression tasks. MDPI 2021-08-17 /pmc/articles/PMC8394492/ /pubmed/34441204 http://dx.doi.org/10.3390/e23081064 Text en © 2021 by the authors. https://creativecommons.org/licenses/by/4.0/Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
spellingShingle Article
Resta, Michele
Monreale, Anna
Bacciu, Davide
Occlusion-Based Explanations in Deep Recurrent Models for Biomedical Signals
title Occlusion-Based Explanations in Deep Recurrent Models for Biomedical Signals
title_full Occlusion-Based Explanations in Deep Recurrent Models for Biomedical Signals
title_fullStr Occlusion-Based Explanations in Deep Recurrent Models for Biomedical Signals
title_full_unstemmed Occlusion-Based Explanations in Deep Recurrent Models for Biomedical Signals
title_short Occlusion-Based Explanations in Deep Recurrent Models for Biomedical Signals
title_sort occlusion-based explanations in deep recurrent models for biomedical signals
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8394492/
https://www.ncbi.nlm.nih.gov/pubmed/34441204
http://dx.doi.org/10.3390/e23081064
work_keys_str_mv AT restamichele occlusionbasedexplanationsindeeprecurrentmodelsforbiomedicalsignals
AT monrealeanna occlusionbasedexplanationsindeeprecurrentmodelsforbiomedicalsignals
AT bacciudavide occlusionbasedexplanationsindeeprecurrentmodelsforbiomedicalsignals