Automated glaucoma detection is an important application of retinal image analysis. Compared with segmentation based approaches, image classification based approaches have a potential of better performance. However, it still remains a challenging problem for two reasons. Firstly, due to insufficient sample size, learning effective features is difficult. Secondly, the shape variations of optic disc introduce misalignment. To address these problem, a new classification based approach for glaucoma detection is proposed, in which deep convolutional networks derived from large-scale generic dataset is used to representing the visual appearance and holistic and local features are combined to mitigate the influence of misalignment. The proposed method achieves an area under the receiver operating characteristic curve of 0.8384 on the Origa dataset, which clearly demonstrates its effectiveness.

Download full-text PDF

Source
http://dx.doi.org/10.1109/EMBC.2016.7590952DOI Listing

Publication Analysis

Top Keywords

holistic local
8
glaucoma detection
8
based approaches
8
classification based
8
integrating holistic
4
local deep
4
deep features
4
features glaucoma
4
glaucoma classification
4
classification automated
4

Similar Publications

Leaky and structurally abnormal blood vessels and increased pressure in the tumor interstitium reduce the infiltration of CAR-T cells in solid tumors, including triple-negative breast cancer (TNBC). Furthermore, high burden of tumor cells may cause reduction of infiltrating CAR-T cells and their functional exhaustion. In this study, various effector-to-target (E:T) ratio experiments are established to model the treatment using CAR-T cells in leukemia (high E:T ratio) and solid tumor (low E:T ratio).

View Article and Find Full Text PDF

Traditional dental research paradigms often lack relevance in marginalized cultural contexts due to inherent biases and misalignment with local values. For Pacific Islanders, this issue is pronounced, as they face serious oral health challenges while remaining underrepresented in scientific discourse. In response, the authors developed the Pacific Islands Dental Research Framework (PIDRF), a culturally informed, community-driven model that directly addresses these limitations in conventional Western approaches.

View Article and Find Full Text PDF

Retrospective analysis of disease characteristics and treatment patterns among patients with esophageal cancer across 14 surgically represented centers.

Cancer Biol Med

January 2025

State Key Laboratory of Holistic Integrative Management of Gastrointestinal Cancers, Beijing Key Laboratory of Carcinogenesis and Translational Research, Department of Gastrointestinal Oncology, Peking University Cancer Hospital & Institute, Beijing 100142, China.

Objective: Esophageal cancer (EC) ranks eighth among cancers in cancer-related deaths globally, and ~44% of new cases occur in China. We sought to describe the clinical characteristics and treatment landscape of EC in China before the approval of immunotherapy in 2020.

Methods: CHANNEL was a large, retrospective study using patient-level data from 14 hospitals/cancer centers across China, including adults initiating therapy for newly diagnosed EC (January to December 2018).

View Article and Find Full Text PDF

Nurse-Led/Involved Home-Based Interventions for Older Adults With Chronic Obstructive Pulmonary Disease (COPD): A Mixed-Methods Systematic Review.

J Clin Nurs

January 2025

School of Nursing and Midwifery College of Health, Medicine, and Wellbeing Hunter and Medical Research Institute Healthcare Transformation Research Program, The Centre for Transformative Nursing, Midwifery, and Health Research, Hunter New England Local Health District, University of Newcastle, Newcastle, New South Wales, Australia.

Aims: To determine the effectiveness of nurse-led/involved home-based interventions for older people with COPD and to explore the experiences of older people and nurses with the interventions.

Design: A mixed-methods systematic review following the JBI methodology for mixed-methods systematic reviews.

Data Sources: The search included relevant and peer-reviewed studies published from January 2010 to December 2023 in CINAHL, MEDLINE, Cochrane Central Register of Controlled Trials, PsycINFO, EMBASE, JBI, EMCARE and ProQuest.

View Article and Find Full Text PDF

Rapid antiretroviral therapy (ART) refers to initiating HIV treatment within seven days of diagnosis. Multidisciplinary teams that implement rapid ART protocols can be effective in addressing the holistic needs of HIV patients. The purpose of this study was to evaluate a multidisciplinary team's implementation of a rapid ART protocol through a retrospective pre and post study design among 627 patients referred to a local HIV clinic between January 2017 and December 2022.

View Article and Find Full Text PDF

Want 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!