Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.3946/kjme.2009.21.4.333 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Predicting branch retinal vein occlusion development using multimodal deep learning and pre-onset fundus hemisection images.
Sci Rep
January 2025
Department of Ophthalmology, Gangnam Severance Hospital, Institute of Vision Research, Yonsei University College of Medicine, 211, Eonjuro, Gangnam-gu, Seoul, 06273, Republic of Korea.
Branch retinal vein occlusion (BRVO) is a leading cause of visual impairment in working-age individuals, though predicting its occurrence from retinal vascular features alone remains challenging. We developed a deep learning model to predict BRVO based on pre-onset, metadata-matched fundus hemisection images. This retrospective cohort study included patients diagnosed with unilateral BRVO from two Korean tertiary centers (2005-2023), using hemisection fundus images from 27 BRVO-affected eyes paired with 81 unaffected hemisections (27 counter and 54 contralateral) for training.
View Article and Find Full Text PDFCXR-LLaVA: a multimodal large language model for interpreting chest X-ray images.
Eur Radiol
January 2025
Department of Radiology, Seoul National University College of Medicine, Seoul National University Hospital, Seoul, Republic of Korea.
Objective: This study aimed to develop an open-source multimodal large language model (CXR-LLaVA) for interpreting chest X-ray images (CXRs), leveraging recent advances in large language models (LLMs) to potentially replicate the image interpretation skills of human radiologists.
Materials And Methods: For training, we collected 592,580 publicly available CXRs, of which 374,881 had labels for certain radiographic abnormalities (Dataset 1) and 217,699 provided free-text radiology reports (Dataset 2). After pre-training a vision transformer with Dataset 1, we integrated it with an LLM influenced by the LLaVA network.
Comparison of Toric Intraocular Lens Axis Accuracy between Optical Biometry and Dual Scheimpflug Topography.
Korean J Ophthalmol
January 2025
Department of Ophthalmology, Seoul National University Bundang Hospital, Seongnam, Korea.
Purpose: To evaluate the accuracy of toric intraocular lens (IOL) axis prediction between two preoperative measurement devices: the optical biometry (IOLMaster 500 or 700) and the dual Scheimpflug topography (Galilei G4).
Methods: Medical records of 64 eyes from 44 patients who underwent phacoemulsification and posterior chamber toric IOL (Zeiss AT TORBI 709M) implantation between July 2017 and January 2022 were reviewed. All patients underwent preoperative evaluation by optical biometry (IOLMaster 500 or IOLMaster 700) and Galilei G4.
Risk of depression in glaucoma patients with vision impairment: A nationwide cohort study.
Heliyon
January 2025
Department of Ophthalmology, Yeouido St. Mary's Hospital, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea.
Purpose: We aimed to investigate the risk of developing depression in individuals with primary open-angle glaucoma with associated vision impairment.
Methods: We conducted a nationwide, population-based cohort study using data from the Korean National Health Information Database and National Disability Registry. We assessed baseline characteristics such as age, sex, income level, lifestyle factors, anthropometric data, lab results, and Charlson Comorbidity Index scores through diagnostic codes and health screening data.
Generation of super-resolution images from barcode-based spatial transcriptomics by deep image prior.
Cell Rep Methods
December 2024
Portrai, Inc., Dongsullagil, 78-18 Jongrogu, Seoul, Republic of Korea; Department of Nuclear Medicine, Seoul National University Hospital, 03080 Seoul, Republic of Korea; Department of Nuclear Medicine, Seoul National University College of Medicine, 03080 Seoul, Republic of Korea. Electronic address:
Spatially resolved transcriptomics (ST) has revolutionized the field of biology by providing a powerful tool for analyzing gene expression in situ. However, current ST methods, particularly barcode-based methods, have limitations in reconstructing high-resolution images from barcodes sparsely distributed in slides. Here, we present SuperST, an algorithm that enables the reconstruction of dense matrices (higher-resolution and non-zero-inflated matrices) from low-resolution ST libraries.
View Article and Find Full Text PDFWant AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!