Publicação

Retinal OCT speckle as a biomarker for glaucoma diagnosis and staging

Detalhes bibliográficos
Resumo:	This paper presents a novel image analysis strategy that increases the potential of macular Optical Coherence Tomography (OCT) by using speckle features as biomarkers in different stages of glaucoma. A large pool of features (480) were computed for a subset of macular OCT volumes of the Leuven eye study cohort. The dataset contained 258 subjects that were divided into four groups based on their glaucoma severity: Healthy (56), Mild (94), Moderate (48), and Severe (60). The OCT speckle features were categorized as statistical properties, statistical distributions, contrast, spatial gray-level dependence matrices, and frequency domain features. The averaged thicknesses of ten retinal layers were also collected. Kruskal-Wallis H test and multivariable regression models were used to infer the most significant features related to glaucoma severity classification and to the correlation with visual field mean deviation. Four features were selected as being the most relevant: the ganglion cell layer (GCL) and the inner plexiform layer (IPL) thicknesses, and two OCT speckle features, the data skewness computed on the retinal nerve fiber layer (RNFL) and the scale parameter (a) of the generalized gamma distribution fitted to the GCL data. Based on a significance level of 0.05, the regression models revealed that RNFL skewness exhibited the highest significance among the features considered for glaucoma severity staging (p-values of 8.6×10-6 for the logistic model and 2.8×10-7 for the linear model). Furthermore, it demonstrated a strong negative correlation with the visual field mean deviation (ρ=-0.64). The post hoc analysis revealed that, when distinguishing healthy controls from glaucoma subjects, GCL thickness is the most relevant feature (p-value of 8.7×10-5). Conversely, when comparing the Mild versus Moderate stages of glaucoma, RNFL skewness emerged as the only feature exhibiting statistical significance (p-value = 0.001). This work shows that macular OCT speckle contains information that is currently not used in clinical practice, and not only complements structural measurements (thickness) but also has a potential for glaucoma staging.
Autores principais:	Vaz, Pedro G.
Outros Autores:	Vaz, Pedro G.; Vaz, Pedro G.; LIBPhys - Laboratory for Instrumentation, Biomedical Engineering and Radiation Physics; LIBPhys - Laboratory for Instrumentation, Biomedical Engineering and Radiation Physics; José Paulo Santos; José Paulo Santos; University of Coimbra; University of Coimbra; Falcão, Amílcar; 01-03-1290; Coimbra; Largo da Porta Férrea; PT; research; Coimbra; PT; 0000-0003-3490-7789; staff; Brea, Luisa Sanchez; Silva, Vânia Bastos; Silva, Vânia Bastos; Silva, Vânia B.; LIBPhys - Laboratory for Instrumentation, Biomedical Engineering and Radiation Physics; LIBPhys - Laboratory for Instrumentation, Biomedical Engineering and Radiation Physics; José Paulo Santos; José Paulo Santos; University of Coimbra; University of Coimbra; Falcão, Amílcar; 01-03-1290; Coimbra; Largo da Porta Férrea; PT; research; Coimbra; PT; staff; van Eijgen, Jan; Stalmans, Ingeborg; Cardoso, João; Cardoso, J. M. R.; Cardoso, J. M. R.; Cardoso, João Manuel; Cardoso, J.; LIP – Laboratory of Instrumentation and Experimental Particle Physics; LIP – Laboratory of Instrumentation and Experimental Particle Physics; research; jmrcardoso@uc.pt; 0000-0002-8832-8208; DA17-1FA7-E78F; 35766426900; J-4486-2013; staff; J. Cardoso background is in Nuclear Electronics Instrumentation and Processing for X and gamma ray spectroscopy. His PhD (2006), demonstrated that optimal signal analysis of X-ray semiconductor detector signals can be achieved for specific noise profiles of the front-end acquisition electronics. It was applied in the data processing of the Charge Radius Experiment with Muonic Atoms at Paul Scherrer Institut (PSI), in Switzerland, (the CREMA collaboration) with very high impact (Nature in 2010, Science in 2013 and high rank journals). In 2007 became involved in the XENON collaboration (Dark Matter Direct Search Experiment) which aims at directly detect the fraction of dark matter in the Universe by means of a xenon dual-phase ultra-pure time projection chamber installed in Laboratori Nazionali del Gran Sasso of INFN, Italy. His successful participation yielded the role of workgroup co-leader of the Slow Control System for the XENON1T phase, operating with 3.5-ton of liquid xenon. Currently is involved in the conception/commissioning of control system for 8-to-10 ton fiducial volume up-scale (XENONnT) and projecting the control for the ultimate detector for dark matter search with the DARWIN project. Since 2009, J Cardoso is deeply interested in data acquisition and processing, to the development of innovative instrumentation and methodologies in biomedical applications. He started by developing innovative non-invasive hemodynamics instruments for individual health condition assessment and population characterization, leading to the supervision of 4 PhD thesis concluded in 2013 (1), 2014 (2) and 2016 (1), and more than 29 MSc thesis. He also proposed a new method for reducing the rehabilitation time of cochlear implanted young patients by using eye-tracking techniques in open-field (as PI of a FCT funded project) and awarded with a University Ignite fund. Other ongoing works under his supervision include sleep apnea diagnosis with machine learning techniques using vectorized human snore sound analysis and optoelectronic single-pixel instrumentation for bio-marker distribution assessment. He is co-author of 112 papers (82 in Q1 journals) in international peer-reviewed journals, having published 41 since 2018. Considering WoS metrics, his works have over 10k citations with an h-index of 40 (i-10 of 46) (June 2023). Amongst the scientific production accomplished after his PhD works, J Cardoso co-authored 5 papers in major multidisciplinary journals, Nature (2010, 2019) and Science (2013, 2015 and 2016), while 30 papers were published in high impact factor journals in the area of Physics (Phys.Rev.Lett/Phys.Rev.D) and 8 in the area of Biomedical Engineering. Since 2013, J Cardoso concluded the supervision of 1 post-doctoral researcher, 4 PhDs (finished and approved with highest mark) and currently is co-supervisor of 2 PhD students (in biomedical engineering and computer science). He also supervised 44 MSc thesis (14 in the last 5 years, several with companies co-supervision, plus 3 currently underway). Considering thesis juries by invitation, J Cardoso participated in 6 PhD and 51 MSc juries. He currently holds 1 patents in co-authorship with one his former PhD students, under the scope of non-invasive continuous blood pressure assessment. He was the PI of a project that uses single-pixel techniques for imaging biomarkers (240k) and he is involved in 2 other projects that were assigned with similar funding: XENON1T/nT project (SAICT 28557 @20% time) and biomedical optical coherence elastography for imaging retina mechanical properties (SAICT 32162 @20% time). He collaborations with CITEUC (instrumentation for MAG-GIC PTDC/CTA-GEO/31744/2017), with Chemistry-Physics (VIBSonCancer development of instrumentation for Raman spectroscopy). He is Associate Professor from the Univ of Coimbra and is group leader for the biomedical engineering strand of LIBPhys.; van Walsum, Theo; Klein, Stefan; Barbosa Breda, João; Andrade De Jesus, Danilo
Assunto:	OCT Speckle Glaucoma Staging
Ano:	2023
País:	Portugal
Tipo de documento:	artigo
Tipo de acesso:	acesso aberto
Instituição associada:	Universidade de Coimbra
Idioma:	inglês
Origem:	Estudo Geral - Universidade de Coimbra

Descrição
Resumo:	This paper presents a novel image analysis strategy that increases the potential of macular Optical Coherence Tomography (OCT) by using speckle features as biomarkers in different stages of glaucoma. A large pool of features (480) were computed for a subset of macular OCT volumes of the Leuven eye study cohort. The dataset contained 258 subjects that were divided into four groups based on their glaucoma severity: Healthy (56), Mild (94), Moderate (48), and Severe (60). The OCT speckle features were categorized as statistical properties, statistical distributions, contrast, spatial gray-level dependence matrices, and frequency domain features. The averaged thicknesses of ten retinal layers were also collected. Kruskal-Wallis H test and multivariable regression models were used to infer the most significant features related to glaucoma severity classification and to the correlation with visual field mean deviation. Four features were selected as being the most relevant: the ganglion cell layer (GCL) and the inner plexiform layer (IPL) thicknesses, and two OCT speckle features, the data skewness computed on the retinal nerve fiber layer (RNFL) and the scale parameter (a) of the generalized gamma distribution fitted to the GCL data. Based on a significance level of 0.05, the regression models revealed that RNFL skewness exhibited the highest significance among the features considered for glaucoma severity staging (p-values of 8.6×10-6 for the logistic model and 2.8×10-7 for the linear model). Furthermore, it demonstrated a strong negative correlation with the visual field mean deviation (ρ=-0.64). The post hoc analysis revealed that, when distinguishing healthy controls from glaucoma subjects, GCL thickness is the most relevant feature (p-value of 8.7×10-5). Conversely, when comparing the Mild versus Moderate stages of glaucoma, RNFL skewness emerged as the only feature exhibiting statistical significance (p-value = 0.001). This work shows that macular OCT speckle contains information that is currently not used in clinical practice, and not only complements structural measurements (thickness) but also has a potential for glaucoma staging.