Skip to content

Analysis of ocular pathological specimen images using image analysis software

Analysis of ocular pathological specimen images using image analysis software - Analysis of ocular pathological specimen images using image analysis software

Status
Recruiting
Phases
Unknown
Study type
Observational
Source
JPRN
Registry ID
JPRN-UMIN000047465
Enrollment
100
Registered
2022-05-01
Start date
2022-05-01
Completion date
Unknown
Last updated
2026-06-29

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

Orbital disease

Interventions

None listed

Sponsors

Osaka Metropolitan University
Lead Sponsor

Eligibility

Sex/Gender
All

Inclusion criteria

Inclusion criteria: 1. From April 1, 2008 to December 31, 2021, a part of the eye was removed by surgery at Osaka City University Visual Pathology or Kobe Kaisei Hospital for the purpose of diagnosis and treatment of the disease, and patients who have received pathologic1. diagnosis who can obtain pathological sample images are eligible. 2.For comparison, pathological sample images of eye tissue in non-lesions are also subject to analysis. 3.Those who are 20 years of age or older at the time of diagnosis

Exclusion criteria

Exclusion criteria: 1.Patients with complications that affect the evaluation 2. Patients who have offered not to participate in this study from the published information

Design outcomes

Primary

MeasureTime frame
Quantitative evaluation results by topological geometric method. Specifically, the ratio (b1 / b0) of the 0-dimensional Vetch number, the one-dimensional Vetch number, and the Betch number obtained from each pathological sample image, and the threshold value of the bivaluation when it is obtained. Basis for setting the main evaluation items: A method for calculating quantitative evaluation of the main outcome 1 is described. In this study, we analyze images using homology. In the analysis using homology, the vetch number of the image is an important concept. Hereinafter, the number of veches based on the definition limited to the two-dimensional image will be outlined. All pathological images in this study are two-dimensional images.

Secondary

MeasureTime frame
1, How does the threshold of bivalanization change when the vetch number is obtained by normal structure, inflammation and neoplastic changes? 2,The correct diagnosis rate when the identification model which distinguishes the above structure, inflammation, tumor, emphysema change from the logistics regression model and the machine learning model for multiclass classification is created.

Countries

Japan

Contacts

Public ContactMizuki Tagami

Osaka Metropolitan University Ophthalmology and Visual sciences

tagami.mizuki@med.osaka-cu.ac.jp+81666453867

Outcome results

None listed

Source: JPRN (via WHO ICTRP) · Data processed: Jul 3, 2026