Achim Langenbucher1, Larissa Häfner2, Timo Eppig3, Berthold Seitz2, Nóra Szentmáry4, Elias Flockerzi2. 1. Institut für Experimentelle Ophthalmologie, Universität des Saarlandes, Kirrberger Str., Gebäude 22, 66421, Homburg, Deutschland. achim.langenbucher@uks.eu. 2. Klinik für Augenheilkunde, Universitätsklinikum des Saarlandes, Kirrberger Str., Gebäude 22, 66421, Homburg, Deutschland. 3. Institut für Experimentelle Ophthalmologie, Universität des Saarlandes, Kirrberger Str., Gebäude 22, 66421, Homburg, Deutschland. 4. Dr. Rolf M. Schwiete Zentrum für Limbusstammzellforschung und kongenitale Aniridie, Universität des Saarlandes, Kirrberger Str., Gebäude 22, 66421, Homburg, Deutschland.
Abstract
BACKGROUND AND OBJECTIVE: In the last decades increasingly more systems of artificial intelligence have been established in medicine, which identify diseases or pathologies or discriminate them from complimentary diseases. Up to now the Corvis®ST (Corneal Visualization Scheimpflug Technology, Corvis®ST, Oculus, Wetzlar, Germany) yielded a binary index for classifying keratoconus but did not enable staging. The purpose of this study was to develop a prediction model, which mimics the topographic keratoconus classification index (TKC) of the Pentacam high resolution (HR, Oculus) with measurement parameters extracted from the Corvis®ST. PATIENTS AND METHODS: In this study 60 measurements from normal subjects (TKC 0) and 379 eyes with keratoconus (TKC 1-4) were recruited. After measurement with the Pentacam HR (target parameter TKC) a measurement with the Corvis®ST device was performed. From this device 6 dynamic response parameters were extracted, which were included in the Corvis biomechanical index (CBI) provided by the Corvis®ST (ARTh, SP-A1, DA ratio 1 mm, DA ratio 2 mm, A1 velocity, max. deformation amplitude). In addition to the TKC as the target, the binarized TKC (1: TKC 1-4, 0: TKC 0) was modelled. The performance of the model was validated with accuracy as an indicator for correct classification made by the algorithm. Misclassifications in the modelling were penalized by the number of stages of deviation between the modelled and measured TKC values. RESULTS: A total of 24 different models of supervised machine learning from 6 different families were tested. For modelling of the TKC stages 0-4, the algorithm based on a support vector machine (SVM) with linear kernel showed the best performance with an accuracy of 65.1% correct classifications. For modelling of binarized TKC, a decision tree with a coarse resolution showed a superior performance with an accuracy of 95.2% correct classifications followed by the SVM with linear or quadratic kernel and a nearest neighborhood classifier with cubic kernel (94.5% each). CONCLUSION: This study aimed to show the principle of supervised machine learning applied to a set-up for the modelled classification of keratoconus staging. Preprocessed measurement data extracted from the Corvis®ST device were used to mimic the TKC provided by the Pentacam device with a series of different algorithms of machine learning.
BACKGROUND AND OBJECTIVE: In the last decades increasingly more systems of artificial intelligence have been established in medicine, which identify diseases or pathologies or discriminate them from complimentary diseases. Up to now the Corvis®ST (Corneal Visualization Scheimpflug Technology, Corvis®ST, Oculus, Wetzlar, Germany) yielded a binary index for classifying keratoconus but did not enable staging. The purpose of this study was to develop a prediction model, which mimics the topographic keratoconus classification index (TKC) of the Pentacam high resolution (HR, Oculus) with measurement parameters extracted from the Corvis®ST. PATIENTS AND METHODS: In this study 60 measurements from normal subjects (TKC 0) and 379 eyes with keratoconus (TKC 1-4) were recruited. After measurement with the Pentacam HR (target parameter TKC) a measurement with the Corvis®ST device was performed. From this device 6 dynamic response parameters were extracted, which were included in the Corvis biomechanical index (CBI) provided by the Corvis®ST (ARTh, SP-A1, DA ratio 1 mm, DA ratio 2 mm, A1 velocity, max. deformation amplitude). In addition to the TKC as the target, the binarized TKC (1: TKC 1-4, 0: TKC 0) was modelled. The performance of the model was validated with accuracy as an indicator for correct classification made by the algorithm. Misclassifications in the modelling were penalized by the number of stages of deviation between the modelled and measured TKC values. RESULTS: A total of 24 different models of supervised machine learning from 6 different families were tested. For modelling of the TKC stages 0-4, the algorithm based on a support vector machine (SVM) with linear kernel showed the best performance with an accuracy of 65.1% correct classifications. For modelling of binarized TKC, a decision tree with a coarse resolution showed a superior performance with an accuracy of 95.2% correct classifications followed by the SVM with linear or quadratic kernel and a nearest neighborhood classifier with cubic kernel (94.5% each). CONCLUSION: This study aimed to show the principle of supervised machine learning applied to a set-up for the modelled classification of keratoconus staging. Preprocessed measurement data extracted from the Corvis®ST device were used to mimic the TKC provided by the Pentacam device with a series of different algorithms of machine learning.
Authors: J Bradley Randleman; William J Dupps; Marcony R Santhiago; Yaron S Rabinowitz; Doug D Koch; R Doyle Stulting; Stephen D Klyce Journal: Cornea Date: 2015-08 Impact factor: 2.651
Authors: Susanne Goebels; Timo Eppig; Stefan Wagenpfeil; Alan Cayless; Berthold Seitz; Achim Langenbucher Journal: Comput Math Methods Med Date: 2017-02-07 Impact factor: 2.238
Authors: Marcella Q Salomão; Ana Luisa Hofling-Lima; Louise Pellegrino Gomes Esporcatte; Bernardo Lopes; Riccardo Vinciguerra; Paolo Vinciguerra; Jens Bühren; Nelson Sena; Guilherme Simões Luz Hilgert; Renato Ambrósio Journal: Int J Environ Res Public Health Date: 2020-03-23 Impact factor: 3.390