Deep phenotype unsupervised machine learning revealed the significance of pachychoroid features in etiology and visual prognosis of age-related macular degeneration

Hosoda, Yasuhiro; Miyake, Masahiro; Yamashiro, Kenji; Ooto, Sotaro; Takahashi, Ayako; Oishi, Akio; Miyata, Manabu; Uji, Akihito; Muraoka, Yuki; Tsujikawa, Akitaka

doi:10.1038/s41598-020-75451-5

Cited by 36 publications

(29 citation statements)

References 67 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…pachychoroid diseases are common. 14,41 Subsequent studies also support this idea, and the criteria for a thick choroid currently are the subject of debate. 37,42,43 However, despite the increasing clinical importance of the choroid, only a few large-scale cohort studies have provided normative data for choroidal thickness.…”

Section: Discussionmentioning

confidence: 97%

Distribution of Choroidal Thickness and Choroidal Vessel Dilation in Healthy Japanese Individuals

Mori

Miyake

Hosoda

et al. 2021

Ophthalmology Science

Self Cite

View full text Add to dashboard Cite

Purpose: To report fundamental epidemiologic data for choroidal parameters such as choroidal thickness and index of choroidal vascularity in Japanese individuals and to evaluate their correlations with age, sex, systemic parameters, and other ocular parameters.Design: Population-based cohort study.Participants: A total of 9850 individuals participated in the first follow-up of the Nagahama Prospective Cohort for Comprehensive Human Bioscience (the Nagahama Study) conducted between 2013 and 2016.Methods: All participants underwent standardized ophthalmic examinations, including OCT with enhanced depth imaging (EDI; RS-3000 Advance; Nidek). We manually segmented the choroidoscleral interface to measure subfoveal choroidal thickness (SFCT) and calculated the normalized choroidal intensity obtained with EDI (NCI EDI ) and choroidal vascularity index (CVI). These are indices of choroidal brightness in OCT and reportedly represent the dilation of choroidal vessels. After summarizing the age-sex stratified distributions of SFCT, NCI EDI , and CVI, their associations with age, sex, axial length (AL), and spherical equivalent (SE) were evaluated using linear regression analysis with adjustments for possible confounders.Main Outcome Measures: Distribution of SFCT, NCI EDI , and CVI in the healthy Japanese population and their characteristics.Results: Age-sex standardized SFCT, NCI EDI , and CVI were 291.2 mm, 0.653, and 66.88%, respectively. In both men and women, SFCT was associated negatively with age (P < 0.001) and NCI EDI was associated positively with age (P < 0.001). Although both SFCT and NCI EDI did not differ significantly between men and women overall (P ¼ 0.87 and P ¼ 0.21, respectively), among younger participants (35e50 years of age), men showed significantly greater SFCT than women (P < 0.001). Only in men was CVI associated positively with age (P < 0.001). In the multivariable analysis, SFCT was associated significantly with age, sex, AL, SE, and the interaction term of age and sex (P < 0.001). Independent of SFCT, NCI EDI and CVI were associated significantly with age (P < 0.001).Conclusions: We report normative Japanese SFCT, NCI EDI , and CVI data using a large general Japanese cohort. The association analysis of SFCT with NCI EDI and CVI suggested that younger individuals have a more lumen-rich choroid for their choroidal thickness than older individuals.

show abstract

Section: Discussionmentioning

confidence: 97%

Distribution of Choroidal Thickness and Choroidal Vessel Dilation in Healthy Japanese Individuals

Mori

Miyake

Hosoda

et al. 2021

Ophthalmology Science

Self Cite

View full text Add to dashboard Cite

show abstract

“…Nonetheless, this can be a clear limitation of the current study since there can be overlap of SFCT between pachychoroid and nonpachychoroid eyes. 32 Third, there was a significant difference in age and sex distribution between pachychoroid and nonpachychoroid eyes both in the dataset and test set. However, we believe this will not significantly affect clinical application of the models since younger age and male prominence are well-established features of pachychoroid and also should be observed in real life.…”

Section: Discussionmentioning

confidence: 88%

Classification of Pachychoroid on Optical Coherence Tomographic En Face Images Using Deep Convolutional Neural Networks

Lee

et al. 2021

Trans. Vis. Sci. Tech.

View full text Add to dashboard Cite

To study the efficacy of deep convolutional neural networks (DCNNs) to differentiate pachychoroid from nonpachychoroid on en face optical coherence tomography (OCT) images at the large choroidal vessel.Methods: En face OCT images were collected from eyes with neovascular age-related macular degeneration, polypoidal choroidal vasculopathy, and central serous chorioretinopathy. All images were prelabeled pachychoroid or nonpachychoroid based on quantitative and qualitative criteria for choroidal morphology on multimodal imaging by two retina specialists. In total, 1188 nonpachychoroid and 884 pachychoroid images were used for training (80%) and validation (20%). Accuracy for identification of pachychoroid by DCNN models was analyzed. Trained models were tested on a test set containing 79 nonpachychoroid and 93 pachychoroid images. Results:The accuracy on the validation set was 94.1%, 93.2%, 94.7%, and 94.4% in DenseNet, GoogLeNet, ResNet50, and Inception-v3, respectively. On a test set, each model demonstrated accuracy of 80.2%, 83.1%, 89.5%, and 90.1% and an F1 score of 0.782, 0.824, 0.904, and 0.901, respectively.Conclusions: DCNN models could classify pachychoroid and nonpachychoroid with good performance on OCT en face images. Automated classification of pachychoroid will be useful for tailored treatment of individual patients with exudative maculopathy.Translational Relevance: En face OCT images can be used by DCNN for classification of pachychoroid.

show abstract

“…Recently, Hosoda et al used machine learning to review Japanese patients with AMD and suggested that 46% of cases were reclassi ed as PNV. 22 However, the diagnosis criteria for PNV have not been clearly determined. Thus, we employed the de nition criteria of previous reports 18, 23 and found a strong correlation between CCT and CVD.…”

Section: Discussionmentioning

confidence: 99%

“…According to the recent reports on PNV analysis, 18,19,22,23,25,31 we diagnosed a patient as PNV when all following criteria were satis ed: Type1 CNV (including PCV), CVD ≥ 180 µm, CCT ≥ 220 µm, absence of drusen in the fundus photographs, and increased choroidal vascular permeability in the late phase of the ICGA images (Fig. 3).…”

Section: Methodsmentioning

confidence: 99%

Analysis Focusing on Plasma Von Willebrand Factor in Pachychoroid Neovasculopathy and Age-related Macular Degeneration

Hirai

Yamashita

Matsumoto

et al. 2021

Preprint

View full text Add to dashboard Cite

Pachychoroid neovasculopathy (PNV) is a new disease concept. Some cases diagnosed as age-related macular degeneration (AMD) have been re-diagnosed as PNV. However, the biological features of PNV are still uncertain. The purpose of this study was to compare PNV and AMD by an analysis focusing on von Willebrand factor (VWF) and complement factor H (CFH). Ninety-seven patients who were previously diagnosed with treatment naïve AMD were enrolled in this study. They were re-classified PNV and AMD based on the clinical criteria and 33 patients were classified as PNV and 64 patients as AMD. We examined the clinical data and performed VWF multimer analysis and genetic polymorphisms in the CFH. PNV group was significantly younger than AMD group. The presence of unusually large VWF multimers (UL-VWFMs) and subretinal hemorrhage was significantly higher in PNV than in AMD. In the CFH I62V, allele A tended to be more frequent in PNV group and recessive homozygous (AA) was found only in PNV group. Thus, the genetic alternations in CFH may result in the residual UL-VWFMs and platelet thrombosis in the choriocapillaris of PNV. In conclusion, our results suggest the biological differences between PNV and AMD.

show abstract

Deep phenotype unsupervised machine learning revealed the significance of pachychoroid features in etiology and visual prognosis of age-related macular degeneration

Cited by 36 publications

References 67 publications

Distribution of Choroidal Thickness and Choroidal Vessel Dilation in Healthy Japanese Individuals

Distribution of Choroidal Thickness and Choroidal Vessel Dilation in Healthy Japanese Individuals

Classification of Pachychoroid on Optical Coherence Tomographic En Face Images Using Deep Convolutional Neural Networks

Analysis Focusing on Plasma Von Willebrand Factor in Pachychoroid Neovasculopathy and Age-related Macular Degeneration

Contact Info

Product

Resources

About