A Single-center, Retrospective Study to Evaluate the Clinical Performance of Artificial Intelligence Medical Assisted Diagnostic Software (VeriSee AMD) for Screening of Age-Related Macular Degeneration
试验速览
- 阶段
- 不适用
- 状态
- 已完成
- 发起方
- 入组人数
- 986
- 试验地点
- 1
- 主要终点
- Specificity
研究概览
简要总结
An observational study is designed to evaluate the clinical performance of VeriSee AMD for potential age-related macular degeneration (AMD) screening from color fundus photography images. The sensitivity and specificity of VeriSee AMD's automated image analysis for screening AMD will be determined through the comparison with the gold standard, which is the judgment of AMD by the ophthalmologists.
详细描述
This is a retrospective, single-center, observational study to assess the clinical performance of VeriSee AMD in screening for potential age-related macular degeneration (AMD).
Potential subjects with images of color fundus photography will be selected for eligibility, and then the ophthalmologist further confirms the images with legible quality. Three independent ophthalmologists (or evaluators) having more than 6 years of experience in retina will evaluate AMD from the color fundus photography images with different sequences of images independently. These three evaluators are different from the ophthalmologist who is responsible for deciding subject's eligibility. The evaluators determine the severity of AMD based on the four-level scale which is the AMD Severity Scale stated in Age-Related Eye Disease Study (AREDS) Report No. 6. Before the study, the training section is necessary to ensure the consensus of the AMD severity grading among three evaluators. The severity determined by the majority of the grade the three evaluators gave is considered the gold/reference standard in this study. Therefore, level 1 and level 2 are classified as non-mtl3AMD, while level 3 and advanced AMD (level 4) are classified as mtl3AMD. The clinical performance displayed as sensitivity and specificity will be determined by comparing the AMD results between VeriSee AMD and the gold standard.
VeriSee AMD is intended to screen AMD from the images taken by color fundus photography, which can assist the physicians to assess whether further examination for retinopathy by the ophthalmologist is needed. The screening result of AMD will be non-mtl3AMD or mtl3AMD for the physicians' reference and does not intend to diagnose AMD, detect concomitant diseases, or treat AMD.
研究设计
- 研究类型
- Observational
- 观察模型
- Case Only
- 时间视角
- Retrospective
入排标准
- 年龄范围
- 50 Years 至 —（Adult, Older Adult）
- 性别
- All
- 接受健康志愿者
- 否
入选标准
- •Subject with age ≥ 50 years old
- •Subject with image taken by color fundus photography that meet the following requirement:
- •The resolution of image is 271x271 pixels or higher;
- •The angle view of image is 45 or 50 degree.
- •Subject's image includes macula as judged by the ophthalmologist.
排除标准
- •The color fundus photography image previously used by VeriSee AMD during the development process and pre-clinical test
- •The macula or other part in the image of color fundus photography is unclear to determine the disease condition as judged by the ophthalmologist.
结局指标
主要结局
Specificity
时间窗: 1 day
To evaluate the clinical performance of VeriSee AMD by determining the Specificity. Specificity = 100% x TN/(TN+FP)
Sensitivity
时间窗: 1 day
To evaluate the clinical performance of VeriSee AMD by determining the sensitivity. Sensitivity = 100% x TP/(TP+FN)
次要结局
- Positive Predictive Values (PPV)(1 day)
- The Percentage of Subjects' Images With Insufficient Quality as Judged by VeriSee AMD(1 day)
- Negative Predictive Values (NPV)(1 day)