{"title":"Identifying Autism Spectrum Disorder Using Optimization-Based Clustering","authors":"Sharifah Mousli, Sona Taheri, Jiayuan He","volume":206,"journal":"International Journal of Computer and Information Engineering","pagesStart":105,"pagesEnd":114,"ISSN":"1307-6892","URL":"https:\/\/publications.waset.org\/pdf\/10013496","abstract":"<p>Autism spectrum disorder (ASD) is a complex developmental condition involving persistent difficulties with social communication, restricted interests, and repetitive behavior. The challenges associated with ASD can interfere with an affected individual\u2019s ability to function in social, academic, and employment settings. Although there is no effective medication known to treat ASD, to our best knowledge, early intervention can significantly improve an affected individual\u2019s overall development. Hence, an accurate diagnosis of ASD at an early phase is essential. The use of machine learning approaches improves and speeds up the diagnosis of ASD. In this paper, we focus on the application of unsupervised clustering methods in ASD, as a large volume of ASD data generated through hospitals, therapy centers, and mobile applications has no pre-existing labels. We conduct a comparative analysis using seven clustering approaches, such as K-means, agglomerative hierarchical, model-based, fuzzy-C-means, affinity propagation, self organizing maps, linear vector quantisation \u2013 as well as the recently developed optimization-based clustering (COMSEP-Clust) approach. We evaluate the performances of the clustering methods extensively on real-world ASD datasets encompassing different age groups: toddlers, children, adolescents, and adults. Our experimental results suggest that the COMSEP-Clust approach outperforms the other seven methods in recognizing ASD with well-separated clusters.<\/p>","references":"[1] A. P. Association et al., \u201cAmerican psychiatric association: Diagnostic\r\nand statistical manual of mental disorders, arlington,\u201d 2013. (Online).\r\nAvailable: https:\/\/doi.org\/10.1176\/appi.books.9780890425596\r\n[2] H. S. Park, S. Y. Yi, S. A. Yoon, and S.-B. Hong, \u201cComparison\r\nof the autism diagnostic observation schedule and childhood autism\r\nrating scale in the diagnosis of autism spectrum disorder: a preliminary\r\nstudy,\u201d Journal of the Korean Academy of Child and Adolescent\r\nPsychiatry, vol. 29, no. 4, p. 172, 2018. (Online). Available:\r\nhttps:\/\/doi.org\/10.5765\/jkacap.180015\r\n[3] C. Lord, S. Risi, L. Lambrecht, E. H. Cook, B. L. Leventhal, P. C.\r\nDiLavore, A. Pickles, and M. Rutter, \u201cThe autism diagnostic observation\r\nschedule\u2014generic: A standard measure of social and communication\r\ndeficits associated with the spectrum of autism,\u201d Journal of autism\r\nand developmental disorders, vol. 30, pp. 205\u2013223, 2000. (Online).\r\nAvailable: https:\/\/doi.org\/10.1023\/A:1005592401947\r\n[4] C. Lord and C. Corsello, \u201cDiagnostic instruments in autistic\r\nspectrum disorders.\u201d Handbook of Autism and Pervasive\r\nDevelopmental Disorders, Fourth Edition, 2005. (Online). Available:\r\nhttps:\/\/doi.org\/10.1007\/978-1-4419-1698-3 908 [5] F. R. Volkmar, C. Lord, A. Bailey, R. T. Schultz, and A. Klin, \u201cAutism\r\nand pervasive developmental disorders,\u201d Journal of child psychology\r\nand psychiatry, vol. 45, no. 1, pp. 135\u2013170, 2004. (Online). Available:\r\nhttps:\/\/doi.org\/10.1046\/j.0021-9630.2003.00317.x\r\n[6] K. K. Hyde, M. N. Novack, N. LaHaye, C. Parlett-Pelleriti, R. Anden,\r\nD. R. Dixon, and E. Linstead, \u201cApplications of supervised machine\r\nlearning in autism spectrum disorder research: a review,\u201d Review\r\nJournal of Autism and Developmental Disorders, vol. 6, pp. 128\u2013146,\r\n2019. [Online]. Available: https:\/\/doi.org\/10.1007\/s40489-019-00158-x\r\n[7] M. M. Rahman, O. L. Usman, R. C. Muniyandi, S. Sahran,\r\nS. Mohamed, and R. A. Razak, \u201cA review of machine learning methods\r\nof feature selection and classification for autism spectrum disorder,\u201d\r\nBrain sciences, vol. 10, no. 12, p. 949, 2020. (Online). Available:\r\nhttps:\/\/doi.org\/10.3390\/brainsci10120949\r\n[8] E. Stevens, A. Atchison, L. Stevens, E. Hong, D. Granpeesheh,\r\nD. Dixon, and E. Linstead, \u201cA cluster analysis of challenging\r\nbehaviors in autism spectrum disorder,\u201d in 2017 16th IEEE\r\nInternational Conference on Machine Learning and Applications\r\n(ICMLA). IEEE, 2017, pp. 661\u2013666. (Online). Available:\r\nhttps:\/\/doi.org\/10.1109\/ICMLA.2017.00-85\r\n[9] M. Elbattah, R. Carette, G. Dequen, J.-L. Gu\u00b4erin, and F. Cilia,\r\n\u201cLearning clusters in autism spectrum disorder: image-based clustering\r\nof eye-tracking scanpaths with deep autoencoder,\u201d in 2019 41st Annual\r\ninternational conference of the IEEE engineering in medicine and\r\nbiology society (EMBC). IEEE, 2019, pp. 1417\u20131420. (Online).\r\nAvailable: https:\/\/doi.org\/10.1109\/EMBC.2019.8856904\r\n[10] S. Silleresi, P. Pr\u00b4evost, R. Zebib, F. Bonnet-Brilhault, D. Conte, and\r\nL. Tuller, \u201cIdentifying language and cognitive profiles in children with\r\nasd via a cluster analysis exploration: Implications for the new icd-11,\u201d\r\nAutism Research, vol. 13, no. 7, pp. 1155\u20131167, 2020. (Online).\r\nAvailable: https:\/\/doi.org\/10.1002\/aur.2268\r\n[11] A. A. Abdulrazzaq, S. S. Hamid, A. T. Al-Douri, A. Mohamad, and\r\nA. M. Ibrahim, \u201cEarly detection of autism spectrum disorders (asd) with\r\nthe help of data mining tools,\u201d BioMed Research International, vol.\r\n2022, 2022. [(Online). Available: https:\/\/doi.org\/10.1155\/2022\/1201129\r\n[12] A. Pratap and C. Kanimozhiselvi, \u201cPredictive assessment of autism\r\nusing unsupervised machine learning models,\u201d International Journal of\r\nAdvanced Intelligence Paradigms, vol. 6, no. 2, pp. 113\u2013121, 2014.\r\n[Online]. Available: https:\/\/doi.org\/10.1504\/IJAIP.2014.062174\r\n[13] C. M. Parlett-Pelleriti, E. Stevens, D. Dixon, and E. J. Linstead,\r\n\u201cApplications of unsupervised machine learning in autism spectrum\r\ndisorder research: a review,\u201d Review Journal of Autism and\r\nDevelopmental Disorders, pp. 1\u201316, 2022. (Online). Available:\r\nhttps:\/\/doi.org\/10.1007\/s40489-021-00299-y\r\n[14] S. Zheng, K. A. Hume, H. Able, S. L. Bishop, and B. A.\r\nBoyd, \u201cExploring developmental and behavioral heterogeneity among\r\npreschoolers with asd: A cluster analysis on principal components,\u201d\r\nAutism Research, vol. 13, no. 5, pp. 796\u2013809, 2020. (Online). Available:\r\nhttps:\/\/doi.org\/10.1002\/aur.2263\r\n[15] M. Uljarevi\u00b4c, A. Lane, A. Kelly, and S. Leekam, \u201cSensory subtypes\r\nand anxiety in older children and adolescents with autism spectrum\r\ndisorder,\u201d Autism Research, vol. 9, no. 10, pp. 1073\u20131078, 2016.\r\n(Online). Available: https:\/\/doi.org\/10.1002\/aur.1602\r\n[16] E. Stevens, D. R. Dixon, M. N. Novack, D. Granpeesheh, T. Smith,\r\nand E. Linstead, \u201cIdentification and analysis of behavioral phenotypes\r\nin autism spectrum disorder via unsupervised machine learning,\u201d\r\nInternational journal of medical informatics, vol. 129, pp. 29\u201336, 2019.\r\n(Online). Available: https:\/\/doi.org\/10.1016\/j.ijmedinf.2019.05.006\r\n[17] T. Obara, M. Ishikuro, G. Tamiya, M. Ueki, C. Yamanaka, S. Mizuno,\r\nM. Kikuya, H. Metoki, H. Matsubara, M. Nagai et al., \u201cPotential\r\nidentification of vitamin b6 responsiveness in autism spectrum\r\ndisorder utilizing phenotype variables and machine learning methods,\u201d\r\nScientific reports, vol. 8, no. 1, pp. 1\u20137, 2018. (Online). Available:\r\nhttps:\/\/doi.org\/10.1038\/s41598-018-33110-w\r\n[18] F. Thabtah, R. Spencer, N. Abdelhamid, F. Kamalov, C. Wentzel, Y. Ye,\r\nand T. Dayara, \u201cAutism screening: an unsupervised machine learning\r\napproach,\u201d Health Information Science and Systems, vol. 10, no. 1, p. 26,\r\n2022. (Online). Available: https:\/\/doi.org\/10.1007\/s13755-022-00191-x\r\n[19] A\u00b4 . E. Tovar, A. Rodr\u00b4\u0131guez-Granados, and N. Arias-Trejo, \u201cAtypical\r\nshape bias and categorization in autism: Evidence from children and\r\ncomputational simulations,\u201d Developmental Science, vol. 23, no. 2, p.\r\ne12885, 2020. (Online). Available: https:\/\/doi.org\/10.1111\/desc.12885\r\n[20] C. Kanimozhiselvi and A. Pratap, \u201cPossibilistic lvq neural\r\nnetwork-an application to childhood autism grading,\u201d Neural\r\nNetwork World, vol. 26, no. 3, p. 253, 2016. (Online). Available:\r\nhttps:\/\/doi.org\/10.14311\/NNW.2016.26.014\r\n[21] O. Veatch, J. Veenstra-VanderWeele, M. Potter, M. Pericak-Vance, and\r\nJ. Haines, \u201cGenetically meaningful phenotypic subgroups in autism\r\nspectrum disorders,\u201d Genes, Brain and Behavior, vol. 13, no. 3, pp.\r\n276\u2013285, 2014. (Online). Available: https:\/\/doi.org\/10.1111\/gbb.12117\r\n[22] T. Lingren, P. Chen, J. Bochenek, F. Doshi-Velez, P. Manning-Courtney,\r\nJ. Bickel, L. Wildenger Welchons, J. Reinhold, N. Bing, Y. Ni et al.,\r\n\u201cElectronic health record based algorithm to identify patients with\r\nautism spectrum disorder,\u201d PloS one, vol. 11, no. 7, p. e0159621, 2016.\r\n(Online). Available: https:\/\/doi.org\/10.1371\/journal.pone.0159621\r\n[23] T. Vargason, R. E. Frye, D. L. McGuinness, and J. Hahn, \u201cClustering\r\nof co-occurring conditions in autism spectrum disorder during early\r\nchildhood: A retrospective analysis of medical claims data,\u201d Autism\r\nResearch, vol. 12, no. 8, pp. 1272\u20131285, 2019. (Online). Available:\r\nhttps:\/\/doi.org\/10.1002\/aur.2128\r\n[24] A. Verma, L. Bang, J. E. Miller, Y. Zhang, M. T. M. Lee,\r\nY. Zhang, M. Byrska-Bishop, D. J. Carey, M. D. Ritchie, S. A.\r\nPendergrass et al., \u201cHuman-disease phenotype map derived from\r\nphewas across 38,682 individuals,\u201d The American Journal of Human\r\nGenetics, vol. 104, no. 1, pp. 55\u201364, 2019. (Online). Available:\r\nhttps:\/\/doi.org\/10.1016\/j.ajhg.2018.11.006\r\n[25] T. Obafemi-Ajayi, D. Lam, T. N. Takahashi, S. Kanne, and\r\nD. Wunsch, \u201cSorting the phenotypic heterogeneity of autism spectrum\r\ndisorders: A hierarchical clustering model,\u201d in 2015 IEEE Conference\r\non Computational Intelligence in Bioinformatics and Computational\r\nBiology (CIBCB). IEEE, 2015, pp. 1\u20137. (Online). Available:\r\nhttps:\/\/doi.org\/10.1109\/CIBCB.2015.7300337\r\n[26] A. Kushki, E. Anagnostou, C. Hammill, P. Duez, J. Brian, A. Iaboni,\r\nR. Schachar, J. Crosbie, P. Arnold, and J. P. Lerch, \u201cExamining overlap\r\nand homogeneity in asd, adhd, and ocd: a data-driven, diagnosis-agnostic\r\napproach,\u201d Translational Psychiatry, vol. 9, no. 1, p. 318, 2019. (Online).\r\nAvailable: https:\/\/doi.org\/10.1038\/s41398-019-0631-2\r\n[27] A. M. Bagirov, N. Hoseini-Monjezi, and S. Taheri, \u201cA novel\r\noptimization approach towards improving separability of clusters,\u201d\r\nComputers & Operations Research, vol. 152, pp. 106\u2013135, 2023.\r\n(Online). Available: https:\/\/doi.org\/10.1007\/978-3-030-37826-4\r\n[28] D. Dua and C. Graff, \u201cUCI Machine Learning Repository,\u201d 2019.\r\n(Online). Available: http:\/\/archive.ics.uci.edu\/ml\r\n[29] F. Thabtah, \u201cAutism spectrum disorder screening: machine learning\r\nadaptation and dsm-5 fulfillment,\u201d in Proceedings of the 1st International\r\nConference on Medical and health Informatics 2017, 2017, pp. 1\u20136.\r\n(Online). Available: https:\/\/doi.org\/10.1145\/3107514.3107515\r\n[30] J. Alwidian, A. Elhassan, and R. Ghnemat, \u201cPredicting autism spectrum\r\ndisorder using machine learning technique,\u201d International Journal of\r\nRecent Technology and Engineering, vol. 8, no. 5, pp. 4139\u20134143,\r\n2020. (Online). Available: https:\/\/doi.org\/10.35940\/ijrte.E6016.018520\r\n[31] F. Thabtah, \u201cAn accessible and efficient autism screening method\r\nfor behavioural data and predictive analyses,\u201d Health informatics\r\njournal, vol. 25, no. 4, pp. 1739\u20131755, 2019. (Online). Available:\r\nhttps:\/\/doi.org\/10.1177\/1460458218796636\r\n[32] S. Baron-Cohen, S. Wheelwright, R. Skinner, J. Martin, and\r\nE. Clubley, \u201cThe autism-spectrum quotient (aq): Evidence from\r\nasperger syndrome\/high-functioning autism, malesand females, scientists\r\nand mathematicians,\u201d Journal of autism and developmental\r\ndisorders, vol. 31, pp. 5\u201317, 2001. (Online). Available:\r\nhttps:\/\/doi.org\/10.1023\/a:1005653411471\r\n[33] C. Allison, S. Baron-Cohen, S. Wheelwright, T. Charman, J. Richler,\r\nG. Pasco, and C. Brayne, \u201cThe q-chat (quantitative checklist for autism\r\nin toddlers): a normally distributed quantitative measure of autistic traits\r\nat 18\u201324 months of age: preliminary report,\u201d Journal of autism and\r\ndevelopmental disorders, vol. 38, pp. 1414\u20131425, 2008.\r\n[34] J. MacQuuen, \u201cSome methods for classification and analysis of\r\nmultivariate observation,\u201d in Proceedings of the 5th Berkley Symposium\r\non Mathematical Statistics and Probability, 1967, pp. 281\u2013297.\r\n[35] F. Murtagh and P. Legendre, \u201cWard\u2019s hierarchical agglomerative\r\nclustering method: which algorithms implement ward\u2019s criterion?\u201d\r\nJournal of classification, vol. 31, pp. 274\u2013295, 2014. (Online).\r\nAvailable: https:\/\/doi.org\/10.1007\/s00357-014-9161-z\r\n[36] C. Fraley, A. Raftery, L. Scrucca, T. Murphy, and M. Fop, \u201cmclust:\r\nNormal mixture modeling for model-based clustering, classification, and\r\ndensity estimation,\u201d R package version, vol. 4, no. 7, 2014.\r\n[37] J. C. Bezdek, \u201cCluster validity with fuzzy sets,\u201d Journal of Cybernetics,\r\n1973. (Online). Available: https:\/\/doi.org\/10.1080\/01969727308546047\r\n[38] B. J. Frey and D. Dueck, \u201cClustering by passing messages between\r\ndata points,\u201d science, vol. 315, no. 5814, pp. 972\u2013976, 2007. (Online).\r\nAvailable: https:\/\/doi.org\/10.1126\/science.1136800\r\n[39] T. Kohonen and T. Kohonen, \u201cLearning vector quantization,\u201d Self-organizing maps, pp. 175\u2013189, 1995. [(Online). Available:\r\nhttps:\/\/doi.org\/10.1007\/978-3-642-97610-0 6\r\n[40] M. Maechler, \u201ccluster: \u201dfinding groups in data\u201d: Cluster analysis\r\nextended rousseeuw et al.\u201d 2022, last accessed 21 April 2023. (Online).\r\nAvailable: https:\/\/cran.r-project.org\/web\/packages\/cluster\/index.html\r\n[41] R Core Team, R: A Language and Environment for Statistical\r\nComputing, R Foundation for Statistical Computing, Vienna, Austria,\r\n2023. (Online). Available: http:\/\/www.R-project.org\/\r\n[42] C. Fraley, \u201cmclust: Gaussian mixture modelling for\r\nmodel-based clustering, classification, and density estimation,\u201d\r\n2022, last accessed 21 April 2023. (Online). Available:\r\nhttps:\/\/cran.r-project.org\/web\/packages\/mclust\/index.html\r\n[43] Z. Cebeci, \u201cPartitioning cluster analysis using fuzzy c-means,\u201d\r\n2017, last accessed 21 April 2023. (Online). Available:\r\nhttps:\/\/cran.r-project.org\/web\/packages\/ppclust\/vignettes\/fcm.html\r\n[44] U. Bodenhofer, \u201cPackage \u2018apcluster\u2019,\u201d 2022, last accessed 20\r\nApril 2023. (Online). Available: https:\/\/cran.r-project.org\/web\/packages\/\r\napcluster\/apcluster.pdf\r\n[45] R. Wehrens and J. Kruisselbrink, \u201cPackage \u2018kohonen\u2019,\u201d\r\n2022, last accessed 07 June 2023. (Online). Available:\r\nhttps:\/\/cran.r-project.org\/web\/packages\/kohonen\/kohonen.pdf\r\n[46] V. Nikolaidis, \u201cPackage \u2018nnlib2rcpp\u2019,\u201d 2023, last\r\naccessed 14 June 2023. (Online). Available:\r\nhttps:\/\/cran.r-project.org\/web\/packages\/nnlib2Rcpp\/index.html\r\n[47] S. Taheri, \u201cComsep-clust,\u201d 2021, last accessed 21 April 2023. (Online).\r\nAvailable: https:\/\/github.com\/SnTa2019\/Clustering-via-Nonsmooth\r\n-Optimization\r\n[48] P. J. Rousseeuw, \u201cSilhouettes: a graphical aid to the interpretation\r\nand validation of cluster analysis,\u201d Journal of computational and\r\napplied mathematics, vol. 20, pp. 53\u201365, 1987. (Online). Available:\r\nhttps:\/\/doi.org\/10.1016\/0377-0427(87)90125-7","publisher":"World Academy of Science, Engineering and Technology","index":"Open Science Index 206, 2024"}