{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,1,30]],"date-time":"2026-01-30T08:47:55Z","timestamp":1769762875389,"version":"3.49.0"},"reference-count":47,"publisher":"Springer Science and Business Media LLC","issue":"S5","license":[{"start":{"date-parts":[[2020,8,1]],"date-time":"2020-08-01T00:00:00Z","timestamp":1596240000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2020,8,20]],"date-time":"2020-08-20T00:00:00Z","timestamp":1597881600000},"content-version":"vor","delay-in-days":19,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["BMC Med Inform Decis Mak"],"published-print":{"date-parts":[[2020,8]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:sec>\n<jats:title>Background<\/jats:title>\n<jats:p>As of today, cancer is still one of the most prevalent and high-mortality diseases, summing more than 9 million deaths in 2018. This has motivated researchers to study the application of machine learning-based solutions for cancer detection to accelerate its diagnosis and help its prevention. Among several approaches, one is to automatically classify tumor samples through their gene expression analysis.<\/jats:p>\n<\/jats:sec><jats:sec>\n<jats:title>Methods<\/jats:title>\n<jats:p>In this work, we aim to distinguish five different types of cancer through RNA-Seq datasets: thyroid, skin, stomach, breast, and lung. To do so, we have adopted a previously described methodology, with which we compare the performance of 3 different autoencoders (AEs) used as a deep neural network weight initialization technique. Our experiments consist in assessing two different approaches when training the classification model \u2014 fixing the weights after pre-training the AEs, or allowing fine-tuning of the entire network \u2014 and two different strategies for embedding the AEs into the classification network, namely by only importing the encoding layers, or by inserting the complete AE. We then study how varying the number of layers in the first strategy, the AEs latent vector dimension, and the imputation technique in the data preprocessing step impacts the network\u2019s overall classification performance. Finally, with the goal of assessing how well does this pipeline generalize, we apply the same methodology to two additional datasets that include features extracted from images of malaria thin blood smears, and breast masses cell nuclei. We also discard the possibility of overfitting by using held-out test sets in the images datasets.<\/jats:p>\n<\/jats:sec><jats:sec>\n<jats:title>Results<\/jats:title>\n<jats:p>The methodology attained good overall results for both RNA-Seq and image extracted data. We outperformed the established baseline for all the considered datasets, achieving an average F<jats:sub>1<\/jats:sub> score of 99.03, 89.95, and 98.84 and an MCC of 0.99, 0.84, and 0.98, for the RNA-Seq (when detecting thyroid cancer), the Malaria, and the Wisconsin Breast Cancer data, respectively.<\/jats:p>\n<\/jats:sec><jats:sec>\n<jats:title>Conclusions<\/jats:title>\n<jats:p>We observed that the approach of fine-tuning the weights of the top layers imported from the AE reached higher results, for all the presented experiences, and all the considered datasets. We outperformed all the previous reported results when comparing to the established baselines.<\/jats:p>\n<\/jats:sec>","DOI":"10.1186\/s12911-020-01150-w","type":"journal-article","created":{"date-parts":[[2020,8,20]],"date-time":"2020-08-20T06:02:53Z","timestamp":1597903373000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":24,"title":["Using autoencoders as a weight initialization method on deep neural networks for disease detection"],"prefix":"10.1186","volume":"20","author":[{"given":"Mafalda Falc\u00e3o","family":"Ferreira","sequence":"first","affiliation":[]},{"given":"Rui","family":"Camacho","sequence":"additional","affiliation":[]},{"given":"Lu\u00eds F.","family":"Teixeira","sequence":"additional","affiliation":[]}],"member":"297","published-online":{"date-parts":[[2020,8,20]]},"reference":[{"key":"1150_CR1","unstructured":"World Health Organization (WHO). Cancer. 2018. https:\/\/www.who.int\/health-topics\/cancer. Accessed on 22 Nov 2019."},{"key":"1150_CR2","unstructured":"World Health Organization (WHO). Cancer Fact Sheet. 2018. https:\/\/www.who.int\/en\/news-room\/fact-sheets\/detail\/cancer. Accessed on 22 Nov 22 2019."},{"key":"1150_CR3","unstructured":"BC Cancer. Change in 5-year survival rates by cancer type for adults in BC, 1997 - 2016. 2019. http:\/\/www.bccancer.bc.ca\/statistics-and-reports-site\/Documents\/Five_Year_Survival_Change_Report_2016. _20190321.pdf. Accessed on 22 Nov 2019."},{"key":"1150_CR4","doi-asserted-by":"publisher","first-page":"8","DOI":"10.1016\/j.csbj.2014.11.005","volume":"13","author":"K Kourou","year":"2015","unstructured":"Kourou K, Exarchos TP, Exarchos KP, Karamouzis MV, Fotiadis DI. Machine learning applications in cancer prognosis and prediction. Comput Struct Biotechnol J. 2015; 13:8\u201317.","journal-title":"Comput Struct Biotechnol J"},{"key":"1150_CR5","unstructured":"National Center for Biotechnology Information (NCBI). Gene Expression. 2017. https:\/\/www.ncbi.nlm.nih.gov\/probe\/docs\/applexpression\/. Accessed on 25 Nov 2019."},{"key":"1150_CR6","unstructured":"The Cancer Genome Atlas (TCGA). The Cancer Genome Atlas. https:\/\/tcga-data.nci.nih.gov\/. Accessed on 25 Nov 2019."},{"key":"1150_CR7","unstructured":"The International Cancer Genome Consortium (ICGC). The International Cancer Genome Consortium. https:\/\/icgc.org. Accessed on 25 Nov 2019."},{"issue":"11","key":"1150_CR8","doi-asserted-by":"publisher","first-page":"951","DOI":"10.1101\/pdb.top084970","volume":"2015","author":"KR Kukurba","year":"2015","unstructured":"Kukurba KR, Montgomery SB. RNA sequencing and analysis. Cold Spring Harbor Protocol. 2015; 2015(11):951\u201369.","journal-title":"Cold Spring Harbor Protocol"},{"issue":"9","key":"1150_CR9","doi-asserted-by":"publisher","first-page":"44","DOI":"10.1038\/s41389-019-0157-8","volume":"8","author":"F Gao","year":"2019","unstructured":"Gao F, Wang W, Tan M, Zhu L, Zhang Y, Fessler E, Vermeulen L, Wang X. Deepcc: a novel deep learning-based framework for cancer molecular subtype classification. Oncogenesis. 2019; 8(9):44.","journal-title":"Oncogenesis"},{"issue":"1","key":"1150_CR10","doi-asserted-by":"publisher","first-page":"17256","DOI":"10.1038\/s41598-019-53989-3","volume":"9","author":"Y Sun","year":"2019","unstructured":"Sun Y, Zhu S, Ma K, Liu W, Yue Y, Hu G, Lu H, Chen W. Identification of 12 cancer types through genome deep learning. Sci Rep. 2019; 9(1):17256.","journal-title":"Sci Rep"},{"key":"1150_CR11","doi-asserted-by":"publisher","unstructured":"Kim B-H, Yu K, Lee PCW. Cancer classification of single-cell gene expression data by neural network. Bioinformatics. 2019. https:\/\/doi.org\/10.1093\/bioinformatics\/btz772.","DOI":"10.1093\/bioinformatics\/btz772"},{"key":"1150_CR12","volume-title":"Advances in Neural Information Processing Systems 28","author":"RK Srivastava","year":"2015","unstructured":"Srivastava RK, Greff K, Schmidhuber J. Training very deep networks In: Cortes C, Lawrence ND, Lee DD, Sugiyama M, Garnett R, editors. Advances in Neural Information Processing Systems 28. New York: Curran Associates, Inc.: 2015. p. 2377\u201385."},{"issue":"Feb","key":"1150_CR13","first-page":"625","volume":"11","author":"D Erhan","year":"2010","unstructured":"Erhan D, Bengio Y, Courville A, Manzagol P-A, Vincent P, Bengio S. Why does unsupervised pre-training help deep learning?J Mach Learn Res. 2010; 11(Feb):625\u201360.","journal-title":"J Mach Learn Res"},{"issue":"4","key":"1150_CR14","doi-asserted-by":"publisher","first-page":"247","DOI":"10.1007\/s40484-019-0189-2","volume":"7","author":"J Zheng","year":"2019","unstructured":"Zheng J, Wang K. Emerging deep learning methods for single-cell RNA-Seq data analysis. Quant Biol. 2019; 7(4):247\u201354.","journal-title":"Quant Biol"},{"issue":"9","key":"1150_CR15","doi-asserted-by":"publisher","first-page":"845","DOI":"10.1186\/s12864-017-4226-0","volume":"18","author":"R Xie","year":"2017","unstructured":"Xie R, Wen J, Quitadamo A, Cheng J, Shi X. A deep auto-encoder model for gene expression prediction. BMC Genomics. 2017; 18(9):845.","journal-title":"BMC Genomics"},{"key":"1150_CR16","doi-asserted-by":"publisher","unstructured":"Teixeira V, Camacho R, Ferreira PG. Learning influential genes on cancer gene expression data with stacked denoising autoencoders. In: IEEE International Conference on Bioinformatics and Biomedicine (BIBM): 2017. p. 1201\u20135. https:\/\/doi.org\/10.1109\/bibm.2017.8217828.","DOI":"10.1109\/bibm.2017.8217828"},{"issue":"1","key":"1150_CR17","doi-asserted-by":"publisher","first-page":"321","DOI":"10.1613\/jair.953","volume":"16","author":"NV Chawla","year":"2002","unstructured":"Chawla NV, Bowyer KW, Hall LO, Kegelmeyer WP. Smote: Synthetic minority over-sampling technique. J Artif Intell Res. 2002; 16(1):321\u201357.","journal-title":"J Artif Intell Res"},{"key":"1150_CR18","doi-asserted-by":"publisher","first-page":"99","DOI":"10.1016\/j.cmpb.2018.10.004","volume":"166","author":"Y Xiao","year":"2018","unstructured":"Xiao Y, Wu J, Lin Z, Zhao X. A semi-supervised deep learning method based on stacked sparse auto-encoder for cancer prediction using RNA-Seq data. Comput Methods Programs Biomed. 2018; 166:99\u2013105.","journal-title":"Comput Methods Programs Biomed"},{"key":"1150_CR19","doi-asserted-by":"publisher","unstructured":"Ferreira MF, Camacho R, Teixeira LF. Autoencoders as weight initialization of deep classification networks applied to papillary thyroid carcinoma. In: Proceedings of the 2018 IEEE International Conference on Bioinformatics and Biomedicine (BIBM): 2018. p. 629\u201332. https:\/\/doi.org\/10.1109\/bibm.2018.8621356.","DOI":"10.1109\/bibm.2018.8621356"},{"key":"1150_CR20","doi-asserted-by":"crossref","unstructured":"L\u00f3pez-Garc\u00eda G, Jerez JM, Franco L, Veredas FJ. A transfer-learning approach to feature extraction from cancer transcriptomes with deep autoencoders. In: Proceedings of the International Work-Conference on Artificial Neural Networks. Springer: 2019. p. 912\u201324.","DOI":"10.1007\/978-3-030-20521-8_74"},{"key":"1150_CR21","unstructured":"Ferreira MF, Camacho R, Teixeira LF. Autoencoders as weight initialization of deep classification networks for cancer versus cancer studies. CoRR. 2020; abs\/2001.05253. http:\/\/arxiv.org\/abs\/2001.05253."},{"key":"1150_CR22","unstructured":"NumPy. NumPy. https:\/\/numpy.org. Accessed on 23 Mar 2020."},{"key":"1150_CR23","unstructured":"Scikit-Learn. Scikit-Learn: Machine Learning in Python. https:\/\/scikit-learn.org\/stable\/. Accessed on 23 Mar 2020."},{"key":"1150_CR24","unstructured":"Keras. Keras: The Python Deep Learning library. https:\/\/keras.io. Accessed on 23 Mar 2020."},{"key":"1150_CR25","unstructured":"Pandas. Pandas Documentation. https:\/\/pandas.pydata.org\/docs\/. Accessed on 23 Mar 2020."},{"key":"1150_CR26","unstructured":"Matplotlib. Matplotlib - Version 3.2.1. https:\/\/matplotlib.org. Accessed on 23 Mar 2020."},{"key":"1150_CR27","unstructured":"cBioPortal. cBioPortal For Cancer Genomics - Datasets. https:\/\/www.cbioportal.org\/datasets. Accessed on 13 Jan 2020."},{"key":"1150_CR28","unstructured":"cBioPortal. The cBioPortal Z-Score calculation method. https:\/\/github.com\/cBioPortal\/cbioportal\/blob\/master\/docs\/Z-Score-normalization-script.md. Accessed on 19 Mar 2020."},{"issue":"3","key":"1150_CR29","doi-asserted-by":"publisher","first-page":"581","DOI":"10.1093\/biomet\/63.3.581","volume":"63","author":"DB Rubin","year":"1976","unstructured":"Rubin DB. Inference and missing data. Biometrika. 1976; 63(3):581\u201392.","journal-title":"Biometrika"},{"key":"1150_CR30","unstructured":"Fraunhofer AICOS Portugal. DEMalariaScope - Automatic detection of malaria in blood smears using smartphones. https:\/\/www.aicos.fraunhofer.pt\/en\/our_work\/projects\/malariascope.html. Accessed on 18 Mar 2020."},{"issue":"10","key":"1150_CR31","doi-asserted-by":"publisher","first-page":"2167","DOI":"10.3390\/s17102167","volume":"17","author":"L Rosado","year":"2017","unstructured":"Rosado L, Da Costa JMC, Elias D, Cardoso JS. Mobile-based analysis of malaria-infected thin blood smears: automated species and life cycle stage determination. Sensors. 2017; 17(10):2167.","journal-title":"Sensors"},{"key":"1150_CR32","unstructured":"Dua D, Graff C. University of California Irvine Machine Learning Repository. 2017. http:\/\/archive.ics.uci.edu\/ml. Accessed on 6 Feb 2020."},{"key":"1150_CR33","doi-asserted-by":"publisher","DOI":"10.7551\/mitpress\/5236.001.0001","volume-title":"Parallel distributed processing: Explorations in the microstructure of cognition, vol. 1","author":"DE Rumelhart","year":"1986","unstructured":"Rumelhart DE, Hinton GE, Williams RJ. Parallel distributed processing: Explorations in the microstructure of cognition, vol. 1. Cambridge: Press, MIT; 1986, pp. 318\u201362. Chap. Learning Internal Representations by Error Propagation."},{"key":"1150_CR34","volume-title":"Proceedings of the 27th International Conference on International Conference on Machine Learning, ICML\u201910","author":"V Nair","year":"2010","unstructured":"Nair V, Hinton GE. Rectified linear units improve restricted boltzmann machines. In: Proceedings of the 27th International Conference on International Conference on Machine Learning, ICML\u201910. USA: Omnipress: 2010. p. 807\u201314."},{"key":"1150_CR35","unstructured":"Goodfellow I, Bengio Y, Courville A. Deep Learning.The MIT Press; 2016. ISBN: 0262035618."},{"key":"1150_CR36","unstructured":"Theis L, Shi W, Cunningham A, Husz\u00e1r F. Lossy image compression with compressive autoencoders. CoRR. 2017; abs\/1703.00395."},{"issue":"1","key":"1150_CR37","doi-asserted-by":"publisher","first-page":"390","DOI":"10.1038\/s41467-018-07931-2","volume":"10","author":"G Eraslan","year":"2019","unstructured":"Eraslan G, Simon LM, Mircea M, Mueller NS, Theis FJ. Single-cell RNA-Seq denoising using a deep count autoencoder. Nat Commun. 2019; 10(1):390.","journal-title":"Nat Commun"},{"key":"1150_CR38","volume-title":"Advances in Neural Information Processing Systems 19","author":"Y Bengio","year":"2007","unstructured":"Bengio Y, Lamblin P, Popovici D, Larochelle H. Greedy layer-wise training of deep networks In: Sch\u00f6lkopf B, Platt JC, Hoffman T, editors. Advances in Neural Information Processing Systems 19. Cambridge: MIT Press: 2007. p. 153\u201360."},{"key":"1150_CR39","volume-title":"Proceedings of the 25th International Conference on Machine Learning, ICML \u201908","author":"P Vincent","year":"2008","unstructured":"Vincent P, Larochelle H, Bengio Y, Manzagol P-A. Extracting and composing robust features with denoising autoencoders. In: Proceedings of the 25th International Conference on Machine Learning, ICML \u201908. New York: ACM: 2008. p. 1096\u2013103."},{"issue":"1","key":"1150_CR40","first-page":"1929","volume":"15","author":"N Srivastava","year":"2014","unstructured":"Srivastava N, Hinton G, Krizhevsky A, Sutskever I, Salakhutdinov R. Dropout: A simple way to prevent neural networks from overfitting. J Mach Learn Res. 2014; 15(1):1929\u201358.","journal-title":"J Mach Learn Res"},{"key":"1150_CR41","unstructured":"Ng A. CS294A Lecture notes - Sparse autoencoder: Standford University. https:\/\/web.stanford.edu\/class\/cs294a\/sparseAutoencoder.pdf. Accessed on 18 Nov 2019."},{"key":"1150_CR42","unstructured":"Ioffe S, Szegedy C. Batch normalization: Accelerating deep network training by reducing internal covariate shift. arXiv preprint. 2015. arXiv:1502.03167."},{"issue":"4","key":"1150_CR43","doi-asserted-by":"publisher","first-page":"303","DOI":"10.1007\/BF02551274","volume":"2","author":"G Cybenko","year":"1989","unstructured":"Cybenko G. Approximation by superpositions of a sigmoidal function. Math Control Signals Syst. 1989; 2(4):303\u201314.","journal-title":"Math Control Signals Syst"},{"key":"1150_CR44","unstructured":"Kingma DP, Ba J. Adam: A method for stochastic optimization. CoRR. 2014; abs\/1412.6980."},{"issue":"2","key":"1150_CR45","doi-asserted-by":"publisher","first-page":"442","DOI":"10.1016\/0005-2795(75)90109-9","volume":"405","author":"BW Matthews","year":"1975","unstructured":"Matthews BW. Comparison of the predicted and observed secondary structure of t4 phage lysozyme. Biochim Biophys Acta Protein Struct. 1975; 405(2):442\u201351.","journal-title":"Biochim Biophys Acta Protein Struct"},{"key":"1150_CR46","unstructured":"Sampaio AFT. DL4Malaria: Deep Learning Approaches for the Automated Detection and Characterisation of Malaria Parasites on Thin Blood Smear Images. Master\u2019s thesis: Faculty of Engineering, University of Porto; 2019."},{"key":"1150_CR47","doi-asserted-by":"publisher","unstructured":"Mushtaq Z, Yaqub A, Hassan A, Su SF. Performance analysis of supervised classifiers using pca based techniques on breast cancer. In: Proceedings of the 2019 International Conference on Engineering and Emerging Technologies (ICEET): 2019. p. 1\u20136. https:\/\/doi.org\/10.1109\/ceet1.2019.8711868.","DOI":"10.1109\/ceet1.2019.8711868"}],"container-title":["BMC Medical Informatics and Decision Making"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s12911-020-01150-w.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1186\/s12911-020-01150-w\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s12911-020-01150-w.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2021,8,19]],"date-time":"2021-08-19T23:07:14Z","timestamp":1629414434000},"score":1,"resource":{"primary":{"URL":"https:\/\/bmcmedinformdecismak.biomedcentral.com\/articles\/10.1186\/s12911-020-01150-w"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2020,8]]},"references-count":47,"journal-issue":{"issue":"S5","published-print":{"date-parts":[[2020,8]]}},"alternative-id":["1150"],"URL":"https:\/\/doi.org\/10.1186\/s12911-020-01150-w","relation":{},"ISSN":["1472-6947"],"issn-type":[{"value":"1472-6947","type":"electronic"}],"subject":[],"published":{"date-parts":[[2020,8]]},"assertion":[{"value":"22 May 2020","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"8 June 2020","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"20 August 2020","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"Not applicable.","order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethics approval and consent to participate"}},{"value":"Not applicable.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Consent for publication"}},{"value":"The authors declare that they have no competing interests.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}],"article-number":"141"}}