Early glaucoma diagnosis and progression analysis based on machine learning hybrid classifiers

• Grant number: 07/51281-9
• Support Opportunities: Regular Research Grants
• Start date: May 01, 2007
• End date: April 30, 2009
• Field of knowledge: Health Sciences - Medicine - Medical Clinics
• Principal Investigator: Vital Paulino Costa
• Grantee: Vital Paulino Costa
• Host Institution: Faculdade de Ciências Médicas (FCM). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil

Abstract

Glaucoma is a progressive optic neuropathy characterized by nerve fiber loss with associated visual field defects. Its diagnosis is based on the observation of classic changes of the nerve fiber layer and optic disc head that lead to visual field defects. Achromatic perimetry is still considered the gold standard psicophysical exam to diagnose and monitor the progression of glaucoma. However, recent studies have suggested that a great amount of ganglionar cells are lost before a visual field defect is detected in achromatic perimetry. New technologies are being developed in order to detect the glaucomatous damage earlier, including short wavelength perimetry and frequency doubling perimetry (functional evaluation), GDx®, HRT II® and OCT® (for nerve fiber layer and or optic disc analysis). These new technologies are associated with low sensitivities and specificities for early glaucoma diagnosis" when used in separate, but its association may increase the diagnosis capacity. The present study aims: 1) to develop an artificial intelligence system to: a) integrate anatomical, epidemiological and functional data to increase the sensitivity and specificity of glaucoma diagnosis and b) integrate the same data on a longitudinal study to detect early glaucoma progression in suspect and glaucomatous patients; 2) to build a normative database with Brazilian individuals for each instrument. Normal individuals, patients with glaucoma or ocular hypertension will be recruited at the Glaucoma Service of University of Campinas for a complete ophthalmic examination, including all the above cited exams. A normative database will be created for each instrument based on the exams of the normal individuals. Results obtained from normal and glaucomatous patients will be employed to develop a hybrid classification system capable of increasing the sensitivity and specificity of glaucoma diagnosis. Finally, longitudinal studies that will follow patients with ocular hypertension and glaucoma will allow the development of hybrid classification systems capable of enhancing early glaucoma diagnosis and detecting glaucoma progression, respectively. (AU)

Scientific publications

(11)

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

BARELLA, KLEYTON A.; CREMASCO, FERNANDA; ZANGALLI, CAMILA; COSTA, VITAL P.. Effects of Misalignments in the Retinal Nerve Fiber Layer Thickness Measurements with Spectral Domain Optical Coherence Tomography. JOURNAL OF OPHTHALMOLOGY, 2014. (07/51281-9)

BARELLA, KLEYTON ARLINDO; COSTA, VITAL PAULINO; VIDOTTI, VANESSA GONCALVES; SILVA, FABRICIO REIS; DIAS, MARCELO; GOMI, EDSON SATOSHI. Glaucoma Diagnostic Accuracy of Machine Learning Classifiers Using Retinal Nerve Fiber Layer and Optic Nerve Data from SD-OCT. JOURNAL OF OPHTHALMOLOGY, 2013. (07/51281-9)

CREMASCO, FERNANDA; MASSA, GRAZIELA; VIDOTTI, VANESSA GONCALVES; DE CARVALHO LUPINACCI, ALVARO PEDROSO; COSTA, VITAL PAULINO. Intrasession, intersession, and interexaminer variabilities of retinal nerve fiber layer measurements with spectral-domain OCT. EUROPEAN JOURNAL OF OPHTHALMOLOGY, v. 21, n. 3, p. 264-270, MAY-JUN 2011. (07/51281-9)

SILVA, FABRICIO R.; VIDOTTI, VANESSA G.; CREMASCO, FERNANDA; DIAS, MARCELO; GOMI, EDSON S.; COSTA, VITAL P.. Sensitivity and specificity of machine learning classifiers for glaucoma diagnosis using Spectral Domain OCT and standard automated perimetry. Arquivos Brasileiros de Oftalmologia, v. 76, n. 3, p. 170-174, MAR-JUN 2013. (07/51281-9)

MASSA, G. C.; VIDOTTI, V. G.; CREMASCO, F.; LUPINACCI, A. P. C.; COSTA, V. P.. Influence of pupil dilation on retinal nerve fibre layer measurements with spectral domain OCT. EYE, v. 24, n. 9, p. 1498-1502, SEP 2010. (07/51281-9)

DIAS, MARCELO; VIDOTTI, VANESSA; COSTA, VITAL PAULINO; GOMI, EDSON SATOSHI; EVANS, BM. High Definition Optical Coherence Tomography and Standard Automated Perimetry Dataset Generator for Glaucoma Diagnosis. 2009 FIRST ANNUAL ORNL BIOMEDICAL SCIENCE & ENGINEERING CONFERENCE: EXPLORING THE INTERSECTIONS OF INTERDISCIPLINARY BIOMEDICAL RESEARCH, v. N/A, p. 2-pg., 2009-01-01. (07/51281-9)

SHIGUEOKA, LEONARDO SEIDI; CABRAL DE VASCONCELLOS, JOSE PAULO; SCHIMITI, RUI BARROSO; CASTRO REIS, ALEXANDRE SOARES; DE OLIVEIRA, GABRIEL OZEAS; GOMI, EDSON SATOSHI; ROCHA VIANNA, JAYME AUGUSTO; DOS REIS LISBOA, RENATO DICHETTI; MEDEIROS, FELIPE ANDRADE; COSTA, VITAL PAULINO. Automated algorithms combining structure and function outperform general ophthalmologists in diagnosing glaucoma. PLoS One, v. 13, n. 12, DEC 5 2018. (07/51281-9)

VIDOTTI, VANESSA G.; COSTA, VITAL P.; SILVA, FABRICIO R.; RESENDE, GRAZIELA M.; CREMASCO, FERNANDA; DIAS, MARCELO; GOMI, EDSON S.. Sensitivity and specificity of machine learning classifiers and spectral domain OCT for the diagnosis of glaucoma. EUROPEAN JOURNAL OF OPHTHALMOLOGY, v. 23, n. 1, p. 61-69, JAN-FEB 2013. (07/51281-9)

BARELLA, KLEYTON ARLINDO; COSTA, VITAL PAULINO; VIDOTTI, VANESSA GONCALVES; SILVA, FABRICIO REIS; DIAS, MARCELO; GOMI, EDSON SATOSHI. Glaucoma Diagnostic Accuracy of Machine Learning Classifiers Using Retinal Nerve Fiber Layer and Optic Nerve Data from SD-OCT. JOURNAL OF OPHTHALMOLOGY, v. 2013, p. 7-pg., 2013-01-01. (07/51281-9)

BARELLA, KLEYTON A.; CREMASCO, FERNANDA; ZANGALLI, CAMILA; COSTA, VITAL P.. Effects of Misalignments in the Retinal Nerve Fiber Layer Thickness Measurements with Spectral Domain Optical Coherence Tomography. JOURNAL OF OPHTHALMOLOGY, v. 2014, p. 7-pg., 2014-01-01. (07/51281-9)

COSTA, V. P.; VIDOTTI, V. G.; RESENDE, G. M.; SILVA, F. R.; CREMASCO, F.; DIAS, M.; GOMI, E.. Sensitivity and Specificity of Machine Learning Classifiers and Spectral Domain OCT for the Diagnosis of Glaucoma. INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE, v. 51, n. 13, p. 2-pg., 2010-04-01. (07/51281-9)