PURPOSE: The volume and complexity of data produced during videokeratography examinations present a challenge of interpretation. As a consequence, results are often analyzed qualitatively by subjective pattern recognition or reduced to comparisons of summary indices. We describe the application of decision tree induction, an automated machine learning classification method, to discriminate between normal and keratoconic corneal shapes in an objective and quantitative way. We then compared this method with other known classification methods. METHODS: The corneal surface was modeled with a seventh-order Zernike polynomial for 132 normal eyes of 92 subjects and 112 eyes of 71 subjects diagnosed with keratoconus. A decision tree classifier was induced using the C4.5 algorithm, and its classification performance was compared with the modified Rabinowitz-McDonnell index, Schwiegerling's Z3 index (Z3), Keratoconus Prediction Index (KPI), KISA%, and Cone Location and Magnitude Index using recommended classification thresholds for each method. We also evaluated the area under the receiver operator characteristic (ROC) curve for each classification method. RESULTS: Our decision tree classifier performed equal to or better than the other classifiers tested: accuracy was 92% and the area under the ROC curve was 0.97. Our decision tree classifier reduced the information needed to distinguish between normal and keratoconus eyes using four of 36 Zernike polynomial coefficients. The four surface features selected as classification attributes by the decision tree method were inferior elevation, greater sagittal depth, oblique toricity, and trefoil. CONCLUSION: Automated decision tree classification of corneal shape through Zernike polynomials is an accurate quantitative method of classification that is interpretable and can be generated from any instrument platform capable of raw elevation data output. This method of pattern classification is extendable to other classification problems.
PURPOSE: The volume and complexity of data produced during videokeratography examinations present a challenge of interpretation. As a consequence, results are often analyzed qualitatively by subjective pattern recognition or reduced to comparisons of summary indices. We describe the application of decision tree induction, an automated machine learning classification method, to discriminate between normal and keratoconic corneal shapes in an objective and quantitative way. We then compared this method with other known classification methods. METHODS: The corneal surface was modeled with a seventh-order Zernike polynomial for 132 normal eyes of 92 subjects and 112 eyes of 71 subjects diagnosed with keratoconus. A decision tree classifier was induced using the C4.5 algorithm, and its classification performance was compared with the modified Rabinowitz-McDonnell index, Schwiegerling's Z3 index (Z3), Keratoconus Prediction Index (KPI), KISA%, and Cone Location and Magnitude Index using recommended classification thresholds for each method. We also evaluated the area under the receiver operator characteristic (ROC) curve for each classification method. RESULTS: Our decision tree classifier performed equal to or better than the other classifiers tested: accuracy was 92% and the area under the ROC curve was 0.97. Our decision tree classifier reduced the information needed to distinguish between normal and keratoconus eyes using four of 36 Zernike polynomial coefficients. The four surface features selected as classification attributes by the decision tree method were inferior elevation, greater sagittal depth, oblique toricity, and trefoil. CONCLUSION: Automated decision tree classification of corneal shape through Zernike polynomials is an accurate quantitative method of classification that is interpretable and can be generated from any instrument platform capable of raw elevation data output. This method of pattern classification is extendable to other classification problems.
Authors: N Nagy; C Decaestecker; R Kiss; F Rypens; D Van Gansbeke; J Mockel; P Rocmans; I Salmon Journal: Int J Mol Med Date: 1999-09 Impact factor: 4.101
Authors: K Zadnik; D O Mutti; N E Friedman; P A Qualley; L A Jones; P Qui; H S Kim; J C Hsu; M L Moeschberger Journal: Invest Ophthalmol Vis Sci Date: 1999-08 Impact factor: 4.799
Authors: Karla Zadnik; Karen Steger-May; Barbara A Fink; Charlotte E Joslin; Jason J Nichols; Carol E Rosenstiel; Julie A Tyler; Julie A Yu; Thomas W Raasch; Kenneth B Schechtman Journal: Cornea Date: 2002-10 Impact factor: 2.651
Authors: Ashraf M Mahmoud; Cynthia J Roberts; Richard G Lembach; Michael D Twa; Edward E Herderick; Timothy T McMahon Journal: Cornea Date: 2008-05 Impact factor: 2.651
Authors: Alicia Galindo-Ferreiro; Julita De Miguel-Gutierrez; Manuel González-Sagrado; Alberto Galvez-Ruiz; Rajiv Khandekar; Silvana Schellini; Julio Galindo-Alonso Journal: Int J Ophthalmol Date: 2017-09-18 Impact factor: 1.779
Authors: Jason D Marsack; Jos J Rozema; Carina Koppen; Marie-Jose Tassignon; Raymond A Applegate Journal: Optom Vis Sci Date: 2013-04 Impact factor: 1.973
Authors: Yan Li; David M Meisler; Maolong Tang; Ake T H Lu; Vishakha Thakrar; Bibiana J Reiser; David Huang Journal: Ophthalmology Date: 2008-11-05 Impact factor: 12.079