Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
his study evaluates the clinical utility of an artificial intelligence (AI)-assisted lesion-based urgent referral triage system for ultra-widefield (UWF) retinal images.
Unlike disease-classification systems, the AI system identifies predefined vision-threatening retinal findings and generates lesion-level urgent referral recommendations. Participating ophthalmologists will evaluate UWF retinal images under randomized AI-assisted and unassisted conditions.
The primary objective is to determine whether lesion-based AI assistance improves urgent referral triage performance compared with unaided image interpretation.
Ultra-widefield retinal imaging is increasingly used for retinal disease screening and referral triage. Many vision-threatening retinal abnormalities require timely identification and referral to retinal specialists.
The AI system evaluated in this study is designed as a lesion-based triage tool rather than a disease-diagnosis system. The model identifies predefined urgent referral retinal findings and generates referral recommendations based on lesion-level evidence.
Urgent referral findings include:
A total of 600 UWF retinal images acquired using Zeiss and Optos imaging systems will be included.
Participating ophthalmologists will independently evaluate images in randomized AI-assisted and unassisted settings.
The primary objective is to determine whether AI assistance improves lesion-based urgent referral triage accuracy.
Not provided
Not provided
Not provided
Not provided
| Label | Type | Description | Intervention Names |
|---|---|---|---|
| AI-Assisted Interpretation | Experimental | Readers interpret UWF retinal images with lesion-level AI findings and urgent referral recommendations. |
|
| Unassisted Interpretation | Active Comparator | Readers interpret UWF retinal images without AI assistance. |
|
| Name | Type | Description | Arm Group Labels | Other Names |
|---|---|---|---|---|
| AI-Assisted UWF Lesion-Based Triage System | Diagnostic Test | Readers interpret UWF retinal images with lesion-level AI findings and urgent referral recommendations. |
|
| Measure | Description | Time Frame |
|---|---|---|
| Correct Lesion-Based Urgent Referral Triage Rate | Proportion of reader referral decisions consistent with expert-adjudicated lesion-based urgent referral classifications. | Through study completion, up to 2 months |
| Measure | Description | Time Frame |
|---|---|---|
| Sensitivity for Urgent Referral Findings | Sensitivity for correctly classifying non-urgent referral images according to expert-adjudicated lesion-based triage labels. | Through study completion, up to 2 months |
| Specificity for Urgent Referral Findings |
Not provided
Inclusion Criteria:
Exclusion Criteria:
Not provided
Not provided
Not provided
Not provided
Not provided
| Name | Role | Phone | Extension | |
|---|---|---|---|---|
| Xiuju Chen, md | Contact | +8618060955810 | joyychen@aliyun.com |
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Each participating ophthalmologist will independently review a library of 600 UWF retinal images.
For each reader, cases will be randomly assigned to either:
Not provided
Not provided
Not provided
Not provided
| Unassisted Interpretation | Diagnostic Test | Readers interpret UWF retinal images without AI assistance. |
|
Specificity for correctly classifying non-urgent referral images according to expert-adjudicated lesion-based triage labels.
| Through study completion, up to 2 months |
| False-Negative Rate for Urgent Referral Findings | Proportion of urgent referral images incorrectly classified as non-urgent referral by readers. | Through study completion, up to 2 months |
| False-Positive Rate for Urgent Referral Findings | Proportion of non-urgent referral images incorrectly classified as urgent referral by readers. | Through study completion, up to 2 months |
| Reader Confidence Score | Reader-reported confidence level for referral decisions measured using a 5-point Likert scale, ranging from 1 (very uncertain) to 5 (very confident). | Immediately after image interpretation. |
| Change in Correct Urgent Referral Decisions After AI Assistance | Number and proportion of cases in which AI assistance changed an incorrect referral decision to a correct referral decision. | Through study completion, up to 2 months |
| ID | Term |
|---|---|
| D012163 | Retinal Detachment |
| D015861 | Retinal Neovascularization |
| ID | Term |
|---|---|
| D012164 | Retinal Diseases |
| D005128 | Eye Diseases |
| D009389 | Neovascularization, Pathologic |
| D008679 | Metaplasia |
| D010335 | Pathologic Processes |
| D013568 | Pathological Conditions, Signs and Symptoms |
Not provided
Not provided