Cargando…
A hybrid unsupervised—Deep learning tandem for electrooculography time series analysis
Medical data are often tricky to get mined for patterns even by the generally demonstrated successful modern methodologies of deep learning. This paper puts forward such a medical classification task, where patient registers of two of the categories are sometimes hard to be distinguished because of...
Autores principales: | , , , , , , , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
Public Library of Science
2020
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7373280/ https://www.ncbi.nlm.nih.gov/pubmed/32692779 http://dx.doi.org/10.1371/journal.pone.0236401 |
_version_ | 1783561472691929088 |
---|---|
author | Stoean, Ruxandra Stoean, Catalin Becerra-García, Roberto García-Bermúdez, Rodolfo Atencia, Miguel García-Lagos, Francisco Velázquez-Pérez, Luis Joya, Gonzalo |
author_facet | Stoean, Ruxandra Stoean, Catalin Becerra-García, Roberto García-Bermúdez, Rodolfo Atencia, Miguel García-Lagos, Francisco Velázquez-Pérez, Luis Joya, Gonzalo |
author_sort | Stoean, Ruxandra |
collection | PubMed |
description | Medical data are often tricky to get mined for patterns even by the generally demonstrated successful modern methodologies of deep learning. This paper puts forward such a medical classification task, where patient registers of two of the categories are sometimes hard to be distinguished because of samples showing characteristics of both labels in turn in several repetitions of the screening procedure. To this end, the current research appoints a pre-processing clustering step (through self-organizing maps) to group the data based on shape similarity and relabel it accordingly. Subsequently, a deep learning approach (a tandem of convolutional and long short-term memory networks) performs the training classification phase on the ‘cleaned’ samples. The dual methodology was applied for the computational diagnosis of electrooculography tests within spino-cerebral ataxia of type 2. The accuracy obtained for the discrimination into three classes was of 78.24%. The improvement that this duo brings over the deep learner alone does not stem from significantly higher accuracy results when the performance is considered for all classes. The major finding of this combination is that half of the presymptomatic cases were correctly found, in opposition to the single deep model, where this category was sacrificed by the learner in favor of a good accuracy overall. A high accuracy in general is desirable for any medical task, however the correct identification of cases before the symptoms become evident is more important. |
format | Online Article Text |
id | pubmed-7373280 |
institution | National Center for Biotechnology Information |
language | English |
publishDate | 2020 |
publisher | Public Library of Science |
record_format | MEDLINE/PubMed |
spelling | pubmed-73732802020-08-13 A hybrid unsupervised—Deep learning tandem for electrooculography time series analysis Stoean, Ruxandra Stoean, Catalin Becerra-García, Roberto García-Bermúdez, Rodolfo Atencia, Miguel García-Lagos, Francisco Velázquez-Pérez, Luis Joya, Gonzalo PLoS One Research Article Medical data are often tricky to get mined for patterns even by the generally demonstrated successful modern methodologies of deep learning. This paper puts forward such a medical classification task, where patient registers of two of the categories are sometimes hard to be distinguished because of samples showing characteristics of both labels in turn in several repetitions of the screening procedure. To this end, the current research appoints a pre-processing clustering step (through self-organizing maps) to group the data based on shape similarity and relabel it accordingly. Subsequently, a deep learning approach (a tandem of convolutional and long short-term memory networks) performs the training classification phase on the ‘cleaned’ samples. The dual methodology was applied for the computational diagnosis of electrooculography tests within spino-cerebral ataxia of type 2. The accuracy obtained for the discrimination into three classes was of 78.24%. The improvement that this duo brings over the deep learner alone does not stem from significantly higher accuracy results when the performance is considered for all classes. The major finding of this combination is that half of the presymptomatic cases were correctly found, in opposition to the single deep model, where this category was sacrificed by the learner in favor of a good accuracy overall. A high accuracy in general is desirable for any medical task, however the correct identification of cases before the symptoms become evident is more important. Public Library of Science 2020-07-21 /pmc/articles/PMC7373280/ /pubmed/32692779 http://dx.doi.org/10.1371/journal.pone.0236401 Text en © 2020 Stoean et al http://creativecommons.org/licenses/by/4.0/ This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/) , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. |
spellingShingle | Research Article Stoean, Ruxandra Stoean, Catalin Becerra-García, Roberto García-Bermúdez, Rodolfo Atencia, Miguel García-Lagos, Francisco Velázquez-Pérez, Luis Joya, Gonzalo A hybrid unsupervised—Deep learning tandem for electrooculography time series analysis |
title | A hybrid unsupervised—Deep learning tandem for electrooculography time series analysis |
title_full | A hybrid unsupervised—Deep learning tandem for electrooculography time series analysis |
title_fullStr | A hybrid unsupervised—Deep learning tandem for electrooculography time series analysis |
title_full_unstemmed | A hybrid unsupervised—Deep learning tandem for electrooculography time series analysis |
title_short | A hybrid unsupervised—Deep learning tandem for electrooculography time series analysis |
title_sort | hybrid unsupervised—deep learning tandem for electrooculography time series analysis |
topic | Research Article |
url | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7373280/ https://www.ncbi.nlm.nih.gov/pubmed/32692779 http://dx.doi.org/10.1371/journal.pone.0236401 |
work_keys_str_mv | AT stoeanruxandra ahybridunsuperviseddeeplearningtandemforelectrooculographytimeseriesanalysis AT stoeancatalin ahybridunsuperviseddeeplearningtandemforelectrooculographytimeseriesanalysis AT becerragarciaroberto ahybridunsuperviseddeeplearningtandemforelectrooculographytimeseriesanalysis AT garciabermudezrodolfo ahybridunsuperviseddeeplearningtandemforelectrooculographytimeseriesanalysis AT atenciamiguel ahybridunsuperviseddeeplearningtandemforelectrooculographytimeseriesanalysis AT garcialagosfrancisco ahybridunsuperviseddeeplearningtandemforelectrooculographytimeseriesanalysis AT velazquezperezluis ahybridunsuperviseddeeplearningtandemforelectrooculographytimeseriesanalysis AT joyagonzalo ahybridunsuperviseddeeplearningtandemforelectrooculographytimeseriesanalysis AT stoeanruxandra hybridunsuperviseddeeplearningtandemforelectrooculographytimeseriesanalysis AT stoeancatalin hybridunsuperviseddeeplearningtandemforelectrooculographytimeseriesanalysis AT becerragarciaroberto hybridunsuperviseddeeplearningtandemforelectrooculographytimeseriesanalysis AT garciabermudezrodolfo hybridunsuperviseddeeplearningtandemforelectrooculographytimeseriesanalysis AT atenciamiguel hybridunsuperviseddeeplearningtandemforelectrooculographytimeseriesanalysis AT garcialagosfrancisco hybridunsuperviseddeeplearningtandemforelectrooculographytimeseriesanalysis AT velazquezperezluis hybridunsuperviseddeeplearningtandemforelectrooculographytimeseriesanalysis AT joyagonzalo hybridunsuperviseddeeplearningtandemforelectrooculographytimeseriesanalysis |