Machine-Learning Points at Endoscopic, Quality of Life, and Olfactory Parameters as Outcome Criteria for Endoscopic Paranasal Sinus Surgery in Chronic Rhinosinusitis
Abstract
:1. Introduction
2. Materials and Methods
2.1. Study Design and Participants
2.2. Data Acquisition
2.2.1. Assessment of Nasal Anatomy/Pathology
Visual Scoring of Pathological Conditions in the Nose and Paranasal Sinuses
Assessment of Eosinophilia in Nasal Tissue
2.2.2. Assessment of Olfactory Function
Clinical Testing of Olfactory Functional Performance
Self-Ratings of the Perceived Olfactory Function
Further Olfaction-Related Information
2.2.3. Assessment of the Quality of Life
Assessment of Disease-Specific Quality of Life Related Parameters
Assessment of Non-Disease-Specific Quality of Life Related Parameters
2.2.4. Acquisition of Sociodemographic and Concomitant Disease-Related Information
2.3. Data Analysis
2.3.1. Data Preprocessing
2.3.2. Explorative Analyses
2.3.3. Unsupervised Identification of Data Structures Supporting Pre- to Post-Surgery Changes
2.3.4. Supervised Identification of Parameters That Carry Relevant Information about Surgery-Related Changes
2.3.5. Statistical Assessment of Changes in Relevant Parameters Related to Surgery Outcomes
3. Results
3.1. Data Structures Reflecting Changes from the Pre- to the Post-Surgery State
3.2. Parameters Containing Relevant Information about Operation-Related Changes
3.3. Changes in Relevant Parameters
4. Discussion
5. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Acknowledgments
Conflicts of Interest
References
- Orlandi, R.R.; Kingdom, T.T.; Smith, T.L.; Bleier, B.; DeConde, A.; Luong, A.U.; Poetker, D.M.; Soler, Z.; Welch, K.C.; Wise, S.K.; et al. International consensus statement on allergy and rhinology: Rhinosinusitis 2021. Int. Forum Allergy Rhinol. 2021, 11, 213–739. [Google Scholar] [CrossRef] [PubMed]
- Kohli, P.; Naik, A.N.; Harruff, E.E.; Nguyen, S.A.; Schlosser, R.J.; Soler, Z.M. The prevalence of olfactory dysfunction in chronic rhinosinusitis. Laryngoscope 2017, 127, 309–320. [Google Scholar] [CrossRef] [PubMed]
- Litvack, J.R.; Mace, J.C.; Smith, T.L. Olfactory function and disease severity in chronic rhinosinusitis. Am. J. Rhinol. Allergy 2009, 23, 139–144. [Google Scholar] [CrossRef] [Green Version]
- Pfaar, O.; Landis, B.N.; Frasnelli, J.; Hüttenbrink, K.B.; Hummel, T. Mechanical obstruction of the olfactory cleft reveals differences between orthonasal and retronasal olfactory functions. Chem. Senses 2006, 31, 27–31. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Lane, A.P.; Turner, J.; May, L.; Reed, R. A genetic model of chronic rhinosinusitis-associated olfactory inflammation reveals reversible functional impairment and dramatic neuroepithelial reorganization. J. Neurosci. 2010, 30, 2324–2329. [Google Scholar] [CrossRef] [PubMed]
- Yee, K.K.; Pribitkin, E.A.; Cowart, B.J.; Vainius, A.A.; Klock, C.T.; Rosen, D.; Feng, P.; McLean, J.; Hahn, C.G.; Rawson, N.E. Neuropathology of the olfactory mucosa in chronic rhinosinusitis. Am. J. Rhinol. Allergy 2010, 24, 110–120. [Google Scholar] [CrossRef] [Green Version]
- Banglawala, S.M.; Oyer, S.L.; Lohia, S.; Psaltis, A.J.; Soler, Z.M.; Schlosser, R.J. Olfactory outcomes in chronic rhinosinusitis with nasal polyposis after medical treatments: A systematic review and meta-analysis. Int. Forum Allergy Rhinol. 2014, 4, 986–994. [Google Scholar] [CrossRef] [PubMed]
- Le, P.T.; Soler, Z.M.; Jones, R.; Mattos, J.L.; Nguyen, S.A.; Schlosser, R.J. Systematic Review and Meta-analysis of SNOT-22 Outcomes after Surgery for Chronic Rhinosinusitis with Nasal Polyposis. Otolaryngol. Head Neck Surg. 2018, 159, 414–423. [Google Scholar] [CrossRef]
- Andrews, P.J.; Poirrier, A.L.; Lund, V.J.; Choi, D. Outcomes in endoscopic sinus surgery: Olfaction, nose scale and quality of life in a prospective cohort study. Clin. Otolaryngol. 2016, 41, 798–803. [Google Scholar] [CrossRef] [Green Version]
- Lind, H.; Joergensen, G.; Lange, B.; Svendstrup, F.; Kjeldsen, A.D. Efficacy of ESS in chronic rhinosinusitis with and without nasal polyposis: A Danish cohort study. Eur. Arch. Otorhinolaryngol. 2016, 273, 911–919. [Google Scholar] [CrossRef]
- Zhao, R.; Chen, K.; Tang, Y. Olfactory changes after endoscopic sinus surgery for chronic rhinosinusitis: A meta-analysis. Clin. Otolaryngol. 2021, 46, 41–51. [Google Scholar] [CrossRef] [PubMed]
- Zou, L.Q.; Hummel, T.; Otte, M.S.; Bitter, T.; Besser, G.; Mueller, C.A.; Welge-Lussen, A.; Bulut, O.C.; Goktas, O.; Negoias, S.; et al. Association between olfactory function and quality of life in patients with olfactory disorders: A multicenter study in over 760 participants. Rhinology 2021, 59, 164–172. [Google Scholar] [CrossRef]
- Elkholi, S.M.A.; Abdelwahab, M.K.; Abdelhafeez, M. Impact of the smell loss on the quality of life and adopted coping strategies in COVID-19 patients. Eur. Arch. Otorhinolaryngol. 2021, 19, 1–8. [Google Scholar] [CrossRef]
- Smeets, M.A.M.; Veldhuizen, M.G.; Galle, S.; Gouweloos, J.; de Haan, A.J.A.; Vernooij, J.; Visscher, F.; Kroeze, J.H.A. Sense of smell disorder and health-related quality of life. Rehabil. Psychol. 2009, 54, 404–412. [Google Scholar] [CrossRef]
- Katotomichelakis, M.; Simopoulos, E.; Tripsianis, G.; Balatsouras, D.; Danielides, G.; Kourousis, C.; Livaditis, M.; Danielides, V. Predictors of quality of life outcomes in chronic rhinosinusitis after sinus surgery. Eur. Arch. Otorhinolaryngol. 2014, 271, 733–741. [Google Scholar] [CrossRef]
- Lötsch, J. Biomedinformatics: A New Journal for the New Decade to Publish Biomedical Informatics Research. BioMedInformatics 2021, 1, 1–5. [Google Scholar] [CrossRef]
- Guyon, I.; Elisseeff, A. An introduction to variable and feature selection. J. Mach. Learn. Res. 2003, 3, 1157–1182. [Google Scholar]
- Fokkens, W.J.; Lund, V.J.; Hopkins, C.; Hellings, P.W.; Kern, R.; Reitsma, S.; Toppila-Salmi, S.; Bernal-Sprekelsen, M.; Mullol, J.; Alobid, I.; et al. European Position Paper on Rhinosinusitis and Nasal Polyps 2020. Rhinology 2020, 58, 1–464. [Google Scholar] [CrossRef] [PubMed]
- Nguyen, D.T.; Nguyen-Thi, P.L.; Jankowski, R. How does measured olfactory function correlate with self-ratings of the sense of smell in patients with nasal polyposis? Laryngoscope 2012, 122, 947–952. [Google Scholar] [CrossRef] [PubMed]
- Lund, V.J.; Kennedy, D.W. Quantification for staging sinusitis. The Staging and Therapy Group. Ann. Otol. Rhinol. Laryngol. Suppl. 1995, 167, 17–21. [Google Scholar] [CrossRef] [PubMed]
- Nathan, K.; Majhi, S.K.; Bhardwaj, R.; Gupta, A.; Ponnusamy, S.; Basu, C.; Kaushal, A. The Role of Diagnostic Nasal Endoscopy and a Computed Tomography Scan (Nose and PNS) in the Assessment of Chronic Rhinosinusitis: A Comparative Evaluation of the Two Techniques. Sinusitis 2021, 5, 59–66. [Google Scholar] [CrossRef]
- Lildholdt, T.; Rundcrantz, H.; Lindqvist, N. Efficacy of topical corticosteroid powder for nasal polyps: A double-blind, placebo-controlled study of budesonide. Clin. Otolaryngol. 1995, 20, 26–30. [Google Scholar] [CrossRef]
- Bachert, C.; Holtappels, G. Pathophysiology of chronic rhinosinusitis, pharmaceutical therapy options. GMS Curr. Top. Otorhinolaryngol. Head Neck Surg. 2015, 14, Doc09. [Google Scholar] [CrossRef] [Green Version]
- Mattos, J.L.; Soler, Z.M.; Schlosser, R.J.; Mace, J.C.; Alt, J.A.; Ramakrishnan, V.R.; Payne, S.C.; Smith, T.L.; Beswick, D.M. Olfactory Function After Surgical Treatment of CRS: A Comparison of CRS Patients to Healthy Controls. Am. J. Rhinol. Allergy 2021, 35, 391–398. [Google Scholar] [CrossRef]
- Kobal, G.; Hummel, T.; Sekinger, B.; Barz, S.; Roscher, S.; Wolf, S.R. “Sniffin’ Sticks”: Screening of olfactory performance. Rhinology 1996, 34, 222–226. [Google Scholar]
- Hummel, T.; Sekinger, B.; Wolf, S.R.; Pauli, E.; Kobal, G. ‘Sniffin’ sticks’: Olfactory performance assessed by the combined testing of odor identification, odor discrimination and olfactory threshold. Chem. Senses 1997, 22, 39–52. [Google Scholar] [CrossRef] [PubMed]
- Oleszkiewicz, A.; Schriever, V.A.; Croy, I.; Hahner, A.; Hummel, T. Updated Sniffin’ Sticks normative data based on an extended sample of 9139 subjects. Eur. Arch. Otorhinolaryngol. 2019, 276, 719–728. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Gudziol, V.; Lötsch, J.; Hahner, A.; Zahnert, T.; Hummel, T. Clinical significance of results from olfactory testing. Laryngoscope 2006, 116, 1858–1863. [Google Scholar] [CrossRef] [PubMed]
- Lötsch, J.; Hummel, T. Clinical usefulness of self-rated olfactory performance—A data science-based assessment of 6000 patients. Chem. Senses 2019, 44, 357–364. [Google Scholar] [CrossRef]
- Bogdanov, V.; Walliczek-Dworschak, U.; Whitcroft, K.L.; Landis, B.N.; Hummel, T. Response to Glucocorticosteroids Predicts Olfactory Outcome After ESS in Chronic Rhinosinusitis. Laryngoscope 2020, 130, 1616–1621. [Google Scholar] [CrossRef]
- Lötsch, J.; Hintschich, C.A.; Petridis, P.; Pade, J.; Hummel, T. Self-Ratings of Olfactory Function and Their Relation to Olfactory Test Scores. A Data Science-Based Analysis in Patients with Nasal Polyposis. Appl. Sci. 2021, 11, 7279. [Google Scholar] [CrossRef]
- Jefferson, M. Anosmia and parosmia. Practitioner 1961, 187, 715–718. [Google Scholar] [PubMed]
- Fikentscher, R.; Rasinski, C. [Parosmias--definition and clinical picture]. Laryngol. Rhinol. Otol. 1986, 65, 663–665. [Google Scholar] [CrossRef]
- Landis, B.N.; Frasnelli, J.; Croy, I.; Hummel, T. Evaluating the clinical usefulness of structured questions in parosmia assessment. Laryngoscope 2010, 120, 1707–1713. [Google Scholar] [CrossRef] [PubMed]
- Piccirillo, J.F.; Merritt, M.G., Jr.; Richards, M.L. Psychometric and clinimetric validity of the 20-Item Sino-Nasal Outcome Test (SNOT-20). Otolaryngol. Head Neck Surg. 2002, 126, 41–47. [Google Scholar] [CrossRef]
- Baumann, I.; Plinkert, P.K.; De Maddalena, H. Development of a grading scale for the Sino-Nasal Outcome Test-20 German Adapted Version (SNOT-20 GAV). Hno 2008, 56, 784–788. [Google Scholar] [CrossRef] [Green Version]
- Ellert, U.; Bellach, B.M. The SF-36 in the Federal Health Survey--description of a current normal sample. Gesundheitswesen 1999, 61 Spec No, S184–S190. [Google Scholar]
- Ware, J.E., Jr.; Sherbourne, C.D. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med. Care 1992, 30, 473–483. [Google Scholar] [CrossRef]
- Ware, J.E. The SF-36 Health Survey. In Quality of Life and Pharmaeconomics in Clinical Trials; Spilker, B., Ed.; Lipincott-Raven: Philadelphia, PA, USA, 1996; pp. 337–346. [Google Scholar]
- Ware, J.E., Jr. SF-36 health survey update. Spine 2000, 25, 3130–3139. [Google Scholar] [CrossRef]
- Brazier, J.E.; Harper, R.; Jones, N.M.; O’Cathain, A.; Thomas, K.J.; Usherwood, T.; Westlake, L. Validating the SF-36 health survey questionnaire: New outcome measure for primary care. BMJ 1992, 305, 160–164. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Hays, R.D.; Shapiro, M.F. An overview of generic health-related quality of life measures for HIV research. Qual. Life Res. 1992, 1, 91–97. [Google Scholar] [CrossRef] [PubMed]
- Ware, J.E.; Kosinski, M. SF-36 Physical & Mental Health Summary Scales: A Manual for Users of Version 1; Quality Metric Incorporated: 2001; Health Assessment Lab, New England Med. Center: Boston, MA, USA; ISBN 1-891810-00-6.
- Wilson, D.; Parsons, J.; Tucker, G. The SF-36 summary scales: Problems and solutions. Sozial und Präventivmedizin 2000, 45, 239–246. [Google Scholar] [CrossRef]
- Lötsch, J.; Ultsch, A. Random Forests Followed by Computed ABC Analysis as a Feature Selection Method for Machine Learning in Biomedical Data. In Advanced Studies in Classification and Data Science; Springer: Singapore, 2020; pp. 57–69. [Google Scholar]
- Ihaka, R.; Gentleman, R. R: A Language for Data Analysis and Graphics. J. Comput. Graph. Stat. 1996, 5, 299–314. [Google Scholar] [CrossRef]
- R Development Core Team. R: A Language and Environment for Statistical Computing; R Development Core Team: Vienna, Austria, 2008; Available online: https://www.R-project.org/ (accessed on 16 September 2021).
- Tukey, J.W. Exploratory Data Analysis; Addison-Wesley Publishing Company: Boston, MA, USA, 1977; Volume 2. [Google Scholar]
- Smirnov, N. Table for Estimating the Goodness of Fit of Empirical Distributions. Ann. Math. Stat. 1948, 19, 279–281. [Google Scholar] [CrossRef]
- Grubbs, F.E. Sample Criteria for Testing Outlying Observations. Ann. Math. Stat. 1950, 21, 27–58. [Google Scholar] [CrossRef]
- Komsta, L. Outliers: Tests for Outliers. 2011. Available online: https://CRAN.R-project.org/package=outliers (accessed on 16 September 2021).
- Ho, T.K. Random decision forests. In Proceedings of the Third International Conference on Document Analysis and Recognition, Montreal, QC, Canada, 14–16 August 1995; Volume 1, p. 278. [Google Scholar]
- Breiman, L. Random Forests. Mach. Learn. 2001, 45, 5–32. [Google Scholar] [CrossRef] [Green Version]
- van Buuren, S.; Groothuis-Oudshoorn, K. mice: Multivariate Imputation by Chained Equations in R. J. Stat. Softw. 2011, 45, 67. [Google Scholar] [CrossRef] [Green Version]
- Pearson, K. On the criterion that a given system of deviations from the probable in the case of a correlated system of variables is such that it can be reasonably supposed to have arisen from random sampling. Philos. Mag. Ser. 5 1900, 50, 157–175. [Google Scholar] [CrossRef] [Green Version]
- Mann, H.B.; Whitney, D.R. On a test of whether one of two random variables is stochastically larger than the other. Ann. Math. Stat. 1947, 18, 50–60. [Google Scholar] [CrossRef]
- Wilcoxon, F. Individual comparisons by ranking methods. Biometrics 1945, 1, 80–83. [Google Scholar] [CrossRef]
- Lötsch, J.; Ultsch, A. A non-parametric effect-size measure capturing changes in central tendency and data distribution shape. PLoS ONE 2020, 15, e0239623. [Google Scholar] [CrossRef] [PubMed]
- Cohen, J. Statistical Power Analysis for the Behavioral Sciences; Routledge: New York, NY, USA, 1988. [Google Scholar]
- Efron, B.; Tibshirani, R.J. An introduction to the Bootstrap; Chapman and Hall: San Francisco, CA, USA, 1995. [Google Scholar]
- Pagès, J. Analyse factorielle de données mixtes. Revue de Statistique Appliquée 2004, 52, 93–111. [Google Scholar]
- Hotelling, H. Analysis of a complex of statistical variables into principal components. J. Educ. Psychol. 1933, 24, 498–520. [Google Scholar] [CrossRef]
- Pearson, K. LIII. On lines and planes of closest fit to systems of points in space. Lond. Edinb. Dublin Philos. Mag. J. Sci. 1901, 2, 559–572. [Google Scholar] [CrossRef] [Green Version]
- Hirschfeld, H.O. A connection between correlation and contingency. Proc. Math. Camb. Philos. Soc. 1935, 31, 520–524. [Google Scholar] [CrossRef]
- Horst, P. Measuring complex attitudes. J. Social. Psychol. 1935, 6, 369–374. [Google Scholar] [CrossRef]
- Kaiser, H.F. The varimax criterion for analytic rotation in factor analysis. Psychometrika 1958, 23, 187–200. [Google Scholar] [CrossRef]
- Guttman, L. Some necessary conditions for common factor analysis. Psychometrika 1954, 19, 149–161. [Google Scholar] [CrossRef]
- Le, S.; Josse, J.; Husson, F.C. FactoMineR: A Package for Multivariate Analysis. J. Stat. Softw. 2008, 25, 1–18. [Google Scholar] [CrossRef] [Green Version]
- Kassambara, A. Practical Guide To Principal Component Methods in R: PCA, M(CA), FAMD, MFA, HCPC, Factoextra; CreateSpace Independent Publishing Platform, Sthda. 2017. Available online: https://CRAN.R-project.org/package=factoextra (accessed on 16 September 2021).
- Kassambara, A.; Mundt, F. Factoextra: Extract and Visualize the Results of Multivariate Data Analyses. 2020. Available online: https://cran.r-project.org/web/packages/factoextra/index.html (accessed on 16 September 2021).
- Lötsch, J.; Malkusch, S. Interpretation of cluster structures in pain-related phenotype data using explainable artificial intelligence (XAI). Eur. J. Pain 2021, 25, 442–465. [Google Scholar] [CrossRef]
- Ultsch, A.; Lötsch, J. Computed ABC Analysis for Rational Selection of Most Informative Variables in Multivariate Data. PLoS ONE 2015, 10, e0129767. [Google Scholar] [CrossRef] [Green Version]
- Juran, J.M. The non-Pareto principle; Mea culpa. Qual. Prog. 1975, 8, 8–9. [Google Scholar]
- Lee, J.S.; Paintsil, E.; Gopalakrishnan, V.; Ghebremichael, M. A comparison of machine learning techniques for classification of HIV patients with antiretroviral therapy-induced mitochondrial toxicity from those without mitochondrial toxicity. BMC Med. Res. Methodol. 2019, 19, 216. [Google Scholar] [CrossRef] [PubMed]
- Good, P.I. Resampling Methods: A Practical Guide to Data Analysis; Birkhäuser: Basel, Switzerland, 2006. [Google Scholar]
- Tillé, Y.; Matei, A. Sampling: Survey Sampling. 2016. Available online: https://CRAN.R-project.org/package=sampling(accessed on 16 September 2021).
- Altman, D.G.; Bland, J.M. Diagnostic tests. 1: Sensitivity and specificity. BMJ 1994, 308, 1552. [Google Scholar] [CrossRef] [Green Version]
- Altman, D.G.; Bland, J.M. Diagnostic tests 2: Predictive values. BMJ 1994, 309, 102. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Sørensen, T.J. A Method of Establishing Groups of Equal Amplitude in Plant Sociology Based on Similarity of Species Content and Its Application to Analyses of the Vegetation on Danish Commons; I Kommission hos E. Munksgaard: København, Denmark, 1948; Volume 5, pp. 1–34. [Google Scholar]
- Jardine, N.; van Rijsbergen, C.J. The use of hierarchic clustering in information retrieval. Inf. Storage Retr. 1971, 7, 217–240. [Google Scholar] [CrossRef]
- Brodersen, K.H.; Ong, C.S.; Stephan, K.E.; Buhmann, J.M. The Balanced Accuracy and Its Posterior Distribution. In Proceedings of the Pattern Recognition (ICPR), 2010 20th International Conference, Istanbul, Turkey, 23–26 August 2010; pp. 3121–3124. [Google Scholar]
- Kuhn, M. Caret: Classification and Regression Training. Astrophys. Source Code Libr. 2015, 53, ascl-1505. [Google Scholar]
- Robin, X.; Turck, N.; Hainard, A.; Tiberti, N.; Lisacek, F.; Sanchez, J.-C.; Müller, M. pROC: An open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinform. 2011, 12, 77. [Google Scholar] [CrossRef] [PubMed]
- Liaw, A.; Wiener, M. Classification and Regression by randomForest. R News 2002, 2, 18–22. [Google Scholar]
- Venables, W.N.; Ripley, B.D. Modern Applied Statistics with S; Springer: New York, NY, USA, 2002. [Google Scholar]
- Meyer, D.; Dimitriadou, E.; Hornik, K.; Weingessel, A.; Leisch, F. e1071: Misc Functions of the Department of Statistics, Probability Theory Group (Formerly: E1071), TU Wien. 2019. Available online: https://CRAN.R-project.org/package=e1071 (accessed on 16 September 2021).
- Cortes, C.; Vapnik, V. Support-Vector Networks. Mach. Learn. 1995, 20, 273–297. [Google Scholar] [CrossRef]
- Cover, T.; Hart, P. Nearest neighbor pattern classification. IEEE Trans. Inf. Theor. 1967, 13, 21–27. [Google Scholar] [CrossRef]
- Mouselimis, L. KernelKnn: Kernel k Nearest Neighbors. 2021. Available online: https://cran.r-project.org/web/packages/KernelKnn/KernelKnn.pdf (accessed on 16 September 2021).
- Bayes, M.; Price, M. An Essay towards Solving a Problem in the Doctrine of Chances. By the Late Rev. Mr. Bayes, F.R.S. Communicated by Mr. Price, in a Letter to John Canton, A.M.F.R.S. Philos. Trans. 1763, 53, 370–418. [Google Scholar] [CrossRef]
- Lotsch, J.; Alfredsson, L.; Lampa, J. Machine-learning based knowledge discovery in rheumatoid arthritis related registry data to identify predictors of persistent pain. Pain 2019. [Google Scholar] [CrossRef] [Green Version]
- Svetnik, V.; Liaw, A.; Tong, C.; Culberson, J.C.; Sheridan, R.P.; Feuston, B.P. Random Forest: A Classification and Regression Tool for Compound Classification and QSAR Modeling. J. Chem. Inf. Comput. Sci. 2003, 43, 1947–1958. [Google Scholar] [CrossRef] [PubMed]
- Spearman, C. The proof and measurement of association between two things. Am. J. Psychol. 1904, 15, 72–101. [Google Scholar] [CrossRef]
- Thomas, J.D.; Bradley, E. Bootstrap confidence intervals. Stat. Sci. 1996, 11, 189–228. [Google Scholar] [CrossRef]
- Rushworth, A. Inspectdf: Inspection, Comparison and Visualisation of Data Frames; 2021; Available online: https://CRAN.R-project.org/package=inspectdf (accessed on 16 September 2021).
- Wickham, H. ggplot2: Elegant Graphics for Data Analysis; Springer: Berlin/Heidelberg, Germany, 2009. [Google Scholar]
- Moore, D.S.; Notz, W.; Fligner, M.A. The Basic Practice of Statistics; W.H. Freeman and Co.: New York, NY, USA, 2013. [Google Scholar]
- Schloerke, B.; Crowley, J.; Cook, D.; Briatte, F.; Marbach, M.; Thoen, E.; Elberg, A.; Larmarange, J. GGally: Extension to ‘ggplot2’; 2018; Available online: https://cran.r-project.org/web/packages/GGally/index.html (accessed on 16 September 2021).
- Pedersen, T.L. ggforce: Accelerating ‘ggplot2’. 2020. Available online: https://CRAN.R-project.org/package=ggforce (accessed on 16 September 2021).
- Gu, Z.; Eils, R.; Schlesner, M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics 2016, 32, 2847–2849. [Google Scholar] [CrossRef] [Green Version]
- Baumann, I.; Blumenstock, G.; Praetorius, M.; Sittel, C.; Piccirillo, J.F.; Plinkert, P.K. [Patients with chronic rhinosinusitis: Disease-specific and general health-related quality of life]. Hno 2006, 54, 544–549. [Google Scholar] [CrossRef]
- Landis, B.N.; Hummel, T.; Hugentobler, M.; Giger, R.; Lacroix, J.S. Ratings of overall olfactory function. Chem. Senses 2003, 28, 691–694. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Lotsch, J.; Kringel, D.; Hummel, T. Machine learning in human olfactory research. Chem. Senses 2019, 44, 11–22. [Google Scholar] [CrossRef]
- Lötsch, J.; Hummel, T.; Ultsch, A. Machine-learned pattern identification in olfactory subtest results. Sci. Rep. 2016, 6, 35688. [Google Scholar] [CrossRef] [Green Version]
- Lötsch, J.; Reither, N.; Bogdanov, V.; Hähner, A.; Ultsch, A.; Hill, K.; Hummel, T. A brain-lesion pattern based algorithm for the diagnosis of posttraumatic olfactory loss. Rhinology 2015, 53, 365–370. [Google Scholar] [CrossRef]
- Morse, J.A.-O.X.; Shilts, M.H.; Ely, K.A.; Li, P.; Sheng, Q.; Huang, L.C.; Wannemuehler, T.J.; Chowdhury, N.I.; Chandra, R.K.; Das, S.R.; et al. Patterns of olfactory dysfunction in chronic rhinosinusitis identified by hierarchical cluster analysis and machine learning algorithms. Int. Forum Allergy Rhinol. 2019, 9, 255–264. [Google Scholar] [CrossRef]
- Walker, A.; Surda, P. Unsupervised Learning Techniques for the Investigation of Chronic Rhinosinusitis. Ann. Otol. Rhinol. Laryngol. 2019, 128, 1170–1176. [Google Scholar] [CrossRef] [PubMed]
- Ward Jr, J.H. Hierarchical grouping to optimize an objective function. J. Am. Stat. Assoc. 1963, 58, 236–244. [Google Scholar] [CrossRef]
- MacQueen, J. Some methods for classification and analysis of multivariate observations. In Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability, University of California Press, Berkeley, CA, USA, 21 June–18 July 1967; pp. 281–297. [Google Scholar]
- Ultsch, A.; Lötsch, J. Machine-learned cluster identification in high-dimensional data. J. Biomed. Inform. 2017, 66, 95–104. [Google Scholar] [CrossRef]
- Ester, M.; Kriegel, H.-P.; Sander, J.o.; Xu, X. A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise. In Proceedings of the Second International Conference on Knowledge Discovery and Data Mining, Portland, OR, USA, 2–4 August 1996; pp. 226–231. [Google Scholar]
- Yaohui, L.; Zhengming, M.; Fang, Y. Adaptive density peak clustering based on K-nearest neighbors with aggregating strategy. Knowl. Based Syst. 2017, 133, 208–220. [Google Scholar] [CrossRef]
- Lötsch, J.; Hummel, T. A machine-learned analysis suggests non-redundant diagnostic information in olfactory subtests. IBRO Rep. 2019, 6, 64–73. [Google Scholar] [CrossRef]
- Ramakrishnan, V.R.; Arbet, J.; Mace, J.C.; Suresh, K.; Shintani Smith, S.; Soler, Z.M.; Smith, T.L. Predicting Olfactory Loss In Chronic Rhinosinusitis Using Machine Learning. Chem. Senses 2021, bjab042. [Google Scholar] [CrossRef] [PubMed]
Parameter Category | Test Battery | Parameter | Mean ± SD/n Baseline | Range Baseline | Mean ± SD/n Post-Surgery | Range Post-Surgery | Wilcoxon/χ2 p-Value |
---|---|---|---|---|---|---|---|
Nasal anatomy/pathology | Lildholdt score | 3.08 ± 1.32 | 1–6 | 0.58 ± 0.96 | 0–4 | 9.115 × 10−25 | |
Lund–Kennedy score | 7.41 ± 3.22 | 1–18 | 5.76 ± 4.16 | 0–18 | 0.0002092 | ||
Eosinophilia in nasal tissue | 1.39 ± 1.11 | 0–3 | same | - | |||
Olfactory | Sniffn Sticks | TDI sum score | 17.84 ± 9.66 | 2–35.5 | 22.8 ± 8.24 | 5–41.75 | 0.0006678 |
Olfactory threshold (T) | 0.59 ± 0.74 | 0–2.3 | 0.97 ± 0.77 | 0–2.69 | 0.0003911 | ||
Odor discrimination (D) | 7.83 ± 4.15 | 0–16 | 9.48 ± 3.42 | 1–16 | 0.004951 | ||
Odor identification (I) | 7.54 ± 4.36 | 0–15 | 9.8 ± 3.68 | 1–15 | 0.0004697 | ||
Olfactory self-rating scale #1 | 1.88 ± 1.64 | 0–7 | 3.48 ± 1.72 | 0–7 | 4.367 × 10−9 | ||
Olfactory self-rating scale #2 | 3.53 ± 2.51 | 1–10 | 5.91 ± 2.77 | 1–10 | 3.669 × 10−8 | ||
Parosmia | 6 | 4 | 0.7449 | ||||
Phantosmia | 7 | 4 | 0.5337 | ||||
Quality of life | SNOT-20 | Primary nasal symptoms | 56.44 ± 18.15 | 4–92 | 26.62 ± 17.55 | 0–76 | 3.331 × 10−18 |
Secondary nasal symptoms | 31.44 ± 18.69 | 0–80 | 18.26 ± 15.04 | 0–63.33 | 9.597 × 10−7 | ||
General life quality | 33.98 ± 19.94 | 0–88.89 | 18.81 ± 15.98 | 0–62.22 | 1.838 × 10−7 | ||
SNOT-20 sum score | 38.83 ± 16.05 | 5–84 | 20.6 ± 13.66 | 1–62 | 1.516 × 10−12 | ||
SF-36 | Physical functioning (PF) | 77.61 ± 22.49 | 10–100 | 86.56 ± 19.13 | 10–100 | 0.0005252 | |
Role-physical (RP) | 70 ± 38.62 | 0–100 | 84.44 ± 31.93 | 0–100 | 0.004417 | ||
Bodily pain (BP) | 72.83 ± 24.68 | 10–100 | 82.56 ± 24.8 | 0–100 | 0.002408 | ||
General health (GH) | 56.22 ± 19.86 | 5–100 | 61.11 ± 20.67 | 15–100 | 0.1172 | ||
Vitality (VT) | 51.83 ± 19.39 | 10–100 | 60.56 ± 17.13 | 25–95 | 0.001698 | ||
Social functioning (SF) | 79.31 ± 20.28 | 12.5–100 | 88.47 ± 15.81 | 37.5–100 | 0.001161 | ||
Role-emotional (RE) | 78.89 ± 37.54 | 0–100 | 87.41 ± 29.38 | 0–100 | 0.1271 | ||
Mental health (MH) | 72.76 ± 16.26 | 36–100 | 76 ± 15.39 | 32–100 | 0.201 | ||
Physical component summary (PCS) | 64.34 ± 23.99 | 8.4–102.58 | 75.34 ± 23.42 | 4.56–108.72 | 0.0004798 | ||
Mental component summary (MCS) | 68.69 ± 21.89 | 0.95–97.86 | 73.62 ± 19.91 | −1.53–104.69 | 0.1335 | ||
Demographic/concomitant disease-related | Age | 50.5 ± 14.92 | 13.9–82.5 | 50.83 ± 14.92 | 14.6–82.83 | - | |
Body mass index (BMI) | 27.37 ± 4.88 | 20.42–44.08 | same | - | |||
Sex | 52/38 | same | - | ||||
Steroid use | 71 | same | - | ||||
Allergy to house dust | 8 | same | - | ||||
Allergy to early bloomers | 1 | same | - | ||||
Allergy to hay fever | 17 | same | - | ||||
Allergy to grasses | 3 | same | - | ||||
Allergy to cat | 2 | same | - | ||||
Asthma, pseudoallergy to acetylsalicylic acid and nasal polyps | 6 | same | - | ||||
Pseudoallergy to acetylsalicylic acid | 9 | same | - | ||||
Asthma | 26 | same | - |
Classifier | RF | logReg | ||||||
---|---|---|---|---|---|---|---|---|
Feature set | Full | Lildholdt only | Primary nasal symptoms only | Olfactory self-rating scale #2 only | Full | Lildholdt only | Primary nasal symptoms only | Olfactory self-rating scale #2 only |
Sensitivity, recall | 83.3 (69.9–96.7) | 86.7 (80–100) | 76.7 (60–90) | 76.7 (50–93.3) | 83.3 (66.7–96.7) | 86.7 (80–96.7) | 80 (66.7–93.3) | 63.3 (50–80.1) |
Specificity | 86.7 (73.3–96.7) | 83.3 (60–93.3) | 76.7 (53.3–93.3) | 53.3 (36.7–73.3) | 83.3 (66.7–96.7) | 83.3 (73.3–93.3) | 80 (66.7–90) | 70 (53.3–80) |
Positive predictive value, precision | 85.7 (74.3–96) | 83.9 (71.4–93.3) | 76.9 (64.5–91.3) | 62.8 (54.8–71.9) | 83.9 (73–95.8) | 84.4 (77.1–93.3) | 80.8 (70.3–90) | 66.7 (56.5–77.8) |
Negative predictive value | 83.9 (73.3–96) | 87.1 (79.3–100) | 76.7 (66.7–88.5) | 70.6 (55.6–88.9) | 83.9 (71.9–96.2) | 87.1 (79.3–96.3) | 80 (70.7–92.3) | 65.6 (56.3–77.4) |
F1 | 84.2 (74.6–91.8) | 85.7 (78.7–93.1) | 76.7 (66.7–85.2) | 69.4 (55.2–78.9) | 83.9 (73.1–91.8) | 86.2 (79.3–93.3) | 80 (71.2–88.5) | 65.5 (54.5–76.7) |
Accuracy | 85 (75–91.7) | 85 (78.3–93.3) | 76.7 (66.7–85) | 65 (55–75) | 83.3 (73.3–91.7) | 86.7 (78.3–93.3) | 80 (71.7–88.3) | 66.7 (56.7–76.7) |
ROC-AUC | 92.1 (85.2–97.5) | 92.8 (87–97.6) | 84.3 (75.4–92.7) | 69.8 (58.6–80.2) | 86.4 (76.8–92.9) | 93.4 (88.2–97.5) | 87.7 (79.4–94.6) | 73.3 (62.8–83.7) |
Classifier | binReg | SVM | ||||||
Feature set | Full | Lildholdt only | Primary nasal symptoms only | Olfactory self-rating scale #2 only | Full | Lildholdt only | Primary nasal symptoms only | Olfactory self-rating scale #2 only |
Sensitivity, recall | 83.3 (63.3–93.4) | 86.7 (80–96.7) | 80 (66.7–93.3) | 63.3 (50–80.1) | 86.7 (66.7–96.7) | 86.7 (56.7–100) | 80 (66.7–93.3) | 66.7 (50–83.3) |
Specificity | 83.3 (66.7–96.7) | 83.3 (73.3–93.3) | 80 (66.7–90) | 70 (53.3–80) | 86.7 (70–96.7) | 83.3 (56.7–96.7) | 80 (66.7–90) | 66.7 (50–80) |
Positive predictive value, precision | 83.3 (71.4–95) | 84.4 (77.1–93.3) | 80.8 (70.3–90) | 66.7 (56.5–77.8) | 84.8 (74.3–95.8) | 84.8 (69.8–96.3) | 80.6 (69.7–89.7) | 66.7 (56.8–77.8) |
Negative predictive value | 82.1 (69.7–93.6) | 87.1 (79.3–96.3) | 80 (70.7–92.3) | 65.6 (56.3–77.4) | 85.2 (73.5–96.3) | 87.1 (68.4–100) | 80.6 (70.3–92.3) | 66.7 (57.1–79.2) |
F1 | 82.6 (70.6–90.3) | 86.2 (79.3–93.3) | 80 (71.2–88.5) | 65.5 (54.5–76.7) | 84.8 (75–92.3) | 85.7 (69.4–93.3) | 80 (70.6–88.9) | 66.7 (54.9–76.9) |
Accuracy | 83.3 (71.7–90) | 86.7 (78.3–93.3) | 80 (71.7–88.3) | 66.7 (56.7–76.7) | 85 (75–91.7) | 85 (75–93.3) | 80 (71.7–88.3) | 66.7 (58.3–76.7) |
ROC-AUC | 86.4 (76.7–93.8) | 93.4 (88.2–97.5) | 87.7 (79.4–94.6) | 73.3 (62.8–83.7) | 85 (75–91.7) | 85 (75–93.3) | 80 (71.7–88.3) | 66.7 (58.3–76.7) |
Classifier | kNN | Naïve Bayes | ||||||
Feature set | Full | Lildholdt only | Primary nasal symptoms only | Olfactory self-rating scale #2 only | Full | Lildholdt only | Primary nasal symptoms only | Olfactory self-rating scale #2 only |
Sensitivity, recall | 76.7 (60–90) | 83.3 (56.7–100) | 76.7 (63.3–93.3) | 70 (43.3–93.3) | 76.7 (43.3–96.7) | 86.7 (80–96.7) | 80 (66.7–93.3) | 66.7 (50–86.7) |
Specificity | 73.3 (56.7–86.7) | 83.3 (56.7–100) | 80 (56.7–93.3) | 60 (40–80) | 73.3 (43.3–93.3) | 83.3 (73.3–93.3) | 80 (66.7–90) | 66.7 (50–80) |
Positive predictive value, precision | 73.5 (62.9–84.4) | 83.9 (68.2–100) | 78.6 (65.8–92) | 64 (53.3–75.9) | 74.3 (60.9–90) | 84.4 (77.1–93.3) | 81.3 (70.6–89.7) | 66.7 (56.5–76.9) |
Negative predictive value | 74.6 (63.3–88) | 84.4 (68.8–100) | 77.4 (67.7–90.3) | 67.9 (53.1–86.4) | 75.9 (59.5–95) | 87.1 (79.3–96.3) | 80 (71–92) | 66.7 (56.7–81.8) |
F1 | 74.2 (63–83.9) | 83.3 (69.4–91.8) | 77.5 (67.9–86.2) | 67.7 (49.1–78.3) | 75 (54.9–85.2) | 86.2 (78.7–93.3) | 80 (71.4–88.9) | 66.7 (54.5–77.6) |
Accuracy | 73.3 (63.3–83.3) | 83.3 (71.7–91.7) | 78.3 (68.3–86.7) | 65 (53.3–75) | 75 (61.7–85) | 86.7 (78.3–93.3) | 80 (71.7–88.3) | 66.7 (56.7–76.7) |
ROC-AUC | 80.1 (70.3–89.4) | 90.3 (79.7–96.3) | 84.5 (75.6–92.5) | 67.5 (55.4–78.9) | 82.6 (70.4–91.3) | 93.4 (88.2–97.5) | 87.7 (79.4–94.6) | 73.3 (62.8–83.7) |
Classifier | CART | C5.0 | ||||||
Feature set | Full | Lildholdt only | Primary nasal symptoms only | Olfactory self-rating scale #2 only | Full | Lildholdt only | Primary nasal symptoms only | Olfactory self-rating scale #2 only |
Sensitivity, recall | 83.3 (66.7–96.7) | 86.7 (76.7–100) | 76.7 (59.9–93.3) | 83.3 (46.7–93.3) | 85 (70–96.7) | 90 (63.3–100) | 76.7 (56.7–96.7) | 86.7 (49.9–96.7) |
Specificity | 83.3 (70–93.3) | 83.3 (56.7–93.3) | 80 (53.3–96.7) | 53.3 (36.7–73.3) | 83.3 (66.7–96.7) | 80 (53.3–93.3) | 76.7 (50–96.7) | 53.3 (36.7–73.3) |
Positive predictive value, precision | 82.8 (72.4–93.3) | 83.9 (69.8–93.1) | 79.3 (65.8–95) | 63 (55.5–72) | 83.3 (72.2–95.7) | 80.6 (68.1–92.3) | 78.1 (64.1–94.7) | 63.4 (55.8–72.2) |
Negative predictive value | 83.9 (71.4–95.8) | 87.1 (78.8–100) | 76.7 (67.5–90.9) | 75 (55.6–90.5) | 84.6 (73–96) | 89.3 (71.8–100) | 77.1 (67.5–93.8) | 77.8 (56.4–92.9) |
F1 | 83.3 (72.7–90.3) | 85.2 (78.7–92.1) | 76.9 (67.8–86.2) | 71.2 (52.6–80) | 83.6 (74.2–91.2) | 84.5 (73.5–90) | 76.9 (67.9–84.7) | 72.9 (54.5–80) |
Accuracy | 83.3 (73.3–90) | 85 (76.7–91.7) | 76.7 (68.3–86.7) | 66.7 (55–76.7) | 83.3 (75–91.7) | 83.3 (75–90) | 76.7 (68.3–85) | 68.3 (56.7–76.7) |
ROC-AUC | 85 (73.2–93.4) | 89 (76.7–96.9) | 80 (70–91.6) | 68.3 (58.3–77.8) | 91.6 (83.2–96.8) | 83.3 (75–95.2) | 76.7 (68.3–85) | 68.3 (58.3–76.7) |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2021 by the authors. 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/).
Share and Cite
Lötsch, J.; Hintschich, C.A.; Petridis, P.; Pade, J.; Hummel, T. Machine-Learning Points at Endoscopic, Quality of Life, and Olfactory Parameters as Outcome Criteria for Endoscopic Paranasal Sinus Surgery in Chronic Rhinosinusitis. J. Clin. Med. 2021, 10, 4245. https://doi.org/10.3390/jcm10184245
Lötsch J, Hintschich CA, Petridis P, Pade J, Hummel T. Machine-Learning Points at Endoscopic, Quality of Life, and Olfactory Parameters as Outcome Criteria for Endoscopic Paranasal Sinus Surgery in Chronic Rhinosinusitis. Journal of Clinical Medicine. 2021; 10(18):4245. https://doi.org/10.3390/jcm10184245
Chicago/Turabian StyleLötsch, Jörn, Constantin A. Hintschich, Petros Petridis, Jürgen Pade, and Thomas Hummel. 2021. "Machine-Learning Points at Endoscopic, Quality of Life, and Olfactory Parameters as Outcome Criteria for Endoscopic Paranasal Sinus Surgery in Chronic Rhinosinusitis" Journal of Clinical Medicine 10, no. 18: 4245. https://doi.org/10.3390/jcm10184245
APA StyleLötsch, J., Hintschich, C. A., Petridis, P., Pade, J., & Hummel, T. (2021). Machine-Learning Points at Endoscopic, Quality of Life, and Olfactory Parameters as Outcome Criteria for Endoscopic Paranasal Sinus Surgery in Chronic Rhinosinusitis. Journal of Clinical Medicine, 10(18), 4245. https://doi.org/10.3390/jcm10184245