The relevance of feature selection methods to the classification of obsessive compulsive disorder based on volumetric measures
Carregando...
Citações na Scopus
14
Tipo de produção
article
Data de publicação
2017
Editora
ELSEVIER SCIENCE BV
Indexadores
Título da Revista
ISSN da Revista
Título do Volume
Autores
TRAMBAIOLLI, Lucas R.
BIAZOLI JR., Claudinei E.
BALARDIN, Joana B.
SATO, Joao R.
Autor de Grupo de pesquisa
Editores
Coordenadores
Organizadores
Citação
JOURNAL OF AFFECTIVE DISORDERS, v.222, p.49-56, 2017
Resumo
Background: Magnetic resonance images (MRI) show detectable anatomical and functional differences between individuals with obsessive-compulsive disorder (OCD) and healthy subjects. Moreover, machine learning techniques have been proposed as tools to identify potential biomarkers and, ultimately, to support clinical diagnosis. However, few studies to date have investigated feature selection (FS) influences in OCD MRI-based classification. Methods: Volumes of cortical and subcortical structures, from MRI data of 38 OCD patients (split into two groups according symptoms severity) and 36 controls, were submitted to seven feature selection algorithms. FS aims to select the most relevant and less redundant features which discriminate between two classes. Then, a classification step was applied, from which the classification performances before and after different FS were compared. For the performance evaluation, leave-one-subject-out accuracies of Support Vector Machine classifiers were considered. Results: Using different FS algorithms, performance improvement was achieved for Controls vs. All OCD discrimination (19.08% of improvement reducing by 80% the amount of features), Controls vs. Low OCD (20.10%, 75%), Controls vs. High OCD (17.32%, 85%) and Low OCD vs. High OCD (10.53%, 75%). Furthermore, all algorithms pointed out classical cortico-striato-thalamo-cortical circuitry structures as relevant features for OCD classification. Limitations: Limitations include the sample size and using only filter approaches for FS. Conclusions: Our results suggest that FS positively impacts. OCD classification using machine-learning techniques. Complementarily, FS algorithms were able to select biologically plausible features automatically.
Palavras-chave
Magnetic resonance imaging, Machine learning, Obsessive-compulsive disorder, Feature selection
Referências
- Abramovitch A, 2013, CLIN PSYCHOL REV, V33, P1163, DOI 10.1016/j.cpr.2013.09.004
- Bari A, 2013, PROG NEUROBIOL, V108, P44, DOI 10.1016/j.pneurobio.2013.06.005
- Bishop CM, 2006, PATTERN RECOGNITION
- Carmona S, 2007, NEUROSCI LETT, V421, P218, DOI 10.1016/j.neulet.2007.05.047
- Chen Y, 2006, FEATURE EXTRACTION, P315, DOI 10.1007/978-3-540-35488-8_13
- Cherian AV, 2014, J AFFECT DISORDERS, V152, P387, DOI 10.1016/j.jad.2013.09.042
- Chu C, 2012, NEUROIMAGE, V60, P59, DOI 10.1016/j.neuroimage.2011.11.066
- Cristianini N., 2000, INTRO SUPPORT VECTOR
- Dale AM, 1999, NEUROIMAGE, V9, P179, DOI 10.1006/nimg.1998.0395
- De Martino F, 2008, NEUROIMAGE, V43, P44, DOI 10.1016/j.neuroimage.2008.06.037
- Esterman M, 2010, NEUROIMAGE, V50, P572, DOI 10.1016/j.neuroimage.2009.10.092
- Fischl B, 2004, CEREB CORTEX, V14, P11, DOI 10.1093/cercor/bhg087
- Fischl B, 1999, NEUROIMAGE, V9, P195, DOI 10.1006/nimg.1998.0396
- Fischl B, 2002, NEURON, V33, P341, DOI 10.1016/S0896-6273(02)00569-X
- Fischl B, 2000, P NATL ACAD SCI USA, V97, P11050, DOI 10.1073/pnas.200033797
- Friedlander L, 2006, CLIN PSYCHOL REV, V26, P32, DOI 10.1016/j.cpr.2005.06.010
- Gilbert AR, 2008, J AFFECT DISORDERS, V109, P117, DOI 10.1016/j.jad.2007.12.223
- Gilbert AR, 2008, NEUROSCI LETT, V435, P45, DOI 10.1016/j.neulet.2008.02.011
- Goncalves OF, 2015, NEUROREPORT, V26, P495, DOI 10.1097/WNR.0000000000000377
- Habeck C, 2010, CELL BIOCHEM BIOPHYS, V58, P53, DOI 10.1007/s12013-010-9093-0
- Hall M., 2009, SIGKDD EXPLORATIONS, V11, P10, DOI 10.1145/1656274.1656278
- Hall MA, 1998, AUST COMP S, V20, P181
- Haynes JD, 2006, NAT REV NEUROSCI, V7, P523, DOI 10.1038/nrn1931
- Hoexter MQ, 2013, J AFFECT DISORDERS, V150, P1213, DOI 10.1016/j.jad.2013.05.041
- Hu XY, 2016, EUR NEUROPSYCHOPHARM, V26, P246, DOI 10.1016/j.euroneuro.2015.12.014
- JENKINS IH, 1994, J NEUROSCI, V14, P3775
- Kim J. J., 2001, BR J PSYCHIAT, V179
- Kloppel S, 2012, NEUROIMAGE, V61, P457, DOI 10.1016/j.neuroimage.2011.11.002
- Kohavi R, 1997, ARTIF INTELL, V97, P273, DOI 10.1016/S0004-3702(97)00043-X
- Koprivova J., 2009, NEUROSCI LETT, V464, P62
- Lazaro L, 2011, PROG NEURO-PSYCHOPH, V35, P1863, DOI 10.1016/j.pnpbp.2011.07.016
- Lazaro L, 2009, PSYCHIAT RES-NEUROIM, V172, P140, DOI 10.1016/j.pscychresns.2008.12.007
- Lee H. D., 2006, IBEROAM J ARTIF INTE, V32
- Li BJ, 2016, FRONT PSYCHIATRY, V7, DOI 10.3389/fpsyt.2016.00078
- Li F, 2014, HUM BRAIN MAPP, V35, P2643, DOI 10.1002/hbm.22357
- Liu H, 1995, PROC INT C TOOLS ART, P388, DOI 10.1109/TAI.1995.479783
- Liu H, 2008, CH CRC DATA MIN KNOW, P3
- Luts J, 2007, ARTIF INTELL MED, V40, P87, DOI 10.1016/j.artmed.2007.02.002
- Mancebo MC, 2014, COMPR PSYCHIAT, V55, P1498, DOI 10.1016/j.comppsych.2014.04.010
- Marchand WR, 2008, NEUROREPORT, V19, P957, DOI 10.1097/WNR.0b013e328302c873
- Menzies L, 2008, NEUROSCI BIOBEHAV R, V32, P525, DOI 10.1016/j.neubiorev.2007.09.005
- Mitchell TM, 2004, MACH LEARN, V57, P145, DOI 10.1023/B:MACH.0000035475.85309.1b
- Mourao-Miranda J, 2011, NEUROIMAGE, V58, P793, DOI 10.1016/j.neuroimage.2011.06.042
- Nakamae T, 2012, PROG NEURO-PSYCHOPH, V37, P90, DOI 10.1016/j.pnpbp.2012.01.001
- Orru G, 2012, NEUROSCI BIOBEHAV R, V36, P1140, DOI 10.1016/j.neubiorev.2012.01.004
- Parrado-Hernandez E, 2014, MED IMAGE ANAL, V18, P435, DOI 10.1016/j.media.2014.01.006
- Pauls DL, 2014, NAT REV NEUROSCI, V15, P410, DOI 10.1038/nrn3746
- Penades R, 2007, EUR PSYCHIAT, V22, P404, DOI 10.1016/j.eurpsy.2006.05.001
- Pereira F, 2009, NEUROIMAGE, V45, pS199, DOI 10.1016/j.neuroimage.2008.11.007
- Piras F, 2015, CORTEX, V62, P89, DOI 10.1016/j.cortex.2013.01.016
- R Core Team, 2011, R LANG ENV STAT COMP
- Robnik-Sikonja M, 2003, MACH LEARN, V53, P23, DOI 10.1023/A:1025667309714
- Sidtis JJ, 2003, NEUROIMAGE, V20, P615, DOI 10.1016/S1053-8119(03)00349-5
- Sokolova M, 2009, INFORM PROCESS MANAG, V45, P427, DOI 10.1016/j.ipm.2009.03.002
- Soriano-Mas C, 2007, NEUROIMAGE, V35, P1028, DOI 10.1016/j.neuroimage.2007.01.011
- Szeszko PR, 2008, AM J PSYCHIAT, V165, P1299, DOI 10.1176/appi.ajp.2008.08010033
- Tang WJ, 2015, BEHAV BRAIN RES, V294, P72, DOI 10.1016/j.bbr.2015.07.061
- Togao O., 2010, PSYCHIAT RES, V184
- Tohka J, 2016, NEUROINFORMATICS, V14, P279, DOI 10.1007/s12021-015-9292-3
- van den Heuvel OA, 2009, BRAIN, V132, P853, DOI 10.1093/brain/awn267
- Venkatasubramanian G, 2012, J PSYCHIATR RES, V46, P1161, DOI 10.1016/j.jpsychires.2012.06.003
- Witten IH, 2011, MOR KAUF D, P1
- Yoo SY, 2008, J KOREAN MED SCI, V23, P24, DOI 10.3346/jkms.2008.23.1.24
- Yu L., 2003, ICML, V3, P856
- Zarei M, 2011, BIOL PSYCHIAT, V70, P1083, DOI 10.1016/j.biopsych.2011.06.032