The use of artificial intelligence to assess diabetic eye disease among the Greenlandic population.

Background: Retina fundus images conducted in Greenland are telemedically assessed for diabetic retinopathy by ophthalmological nurses in Denmark. Applying an AI grading solution, in a Greenlandic setting, could potentially improve the efficiency and cost-effectiveness of DR screening.Method: We developed an AI model using retina fundus photos, performed on persons registered with diabetes in Greenland and Denmark, using Optos® ultra wide-field scanning laser ophthalmoscope, graded according to ICDR.Using the ResNet50 network we compared the model's ability to distinguish between different images of ICDR severity levels in a confusion matrix.Results: Comparing images with ICDR level 0 to images of ICDR level 4 resulted in an accuracy of 0.9655, AUC of 0.9905, sensitivity and specificity of 96.6%.Comparing ICDR levels 0,1,2 with ICDR levels 3,4, we achieved a performance with an accuracy of 0.8077, an AUC of 0.8728, a sensitivity of 84.6% and a specificity of 78.8%. For the other comparisons, we achieved a modest performance.Conclusion: We developed an AI model using Greenlandic data, to automatically detect DR on Optos retina fundus images. The sensitivity and specificity were too low for our model to be applied directly in a clinical setting, thus optimising the model should be prioritised.

Blood oxygen saturation is lower in persons with pre-diabetes and screen-detected diabetes compared with non-diabetic individuals: A population-based study of the Lolland-Falster Health Study cohort.

Aims: Low blood oxygen saturation is associated with increased mortality and persons with diabetes have sub-clinical hypoxemia. We aimed to confirm the presence of sub-clinical hypoxemia in pre-diabetes, screen-detected diabetes and known diabetes. Methods: Pre-diabetes was defined as hemoglobin A1C (HbA1C) ≥ 42 mmol/mol and <48 mmol/mol; known diabetes as history or treatment of diabetes; screen-detected diabetes as no history or treatment of diabetes and HbA1C ≥ 48 mmol/mol. Blood oxygen saturation was measured with pulse oximetry. Urine albumin-to creatinine ratio (UACR) was measured on a single spot urine. Results: The study included 829 adults (≥18 years) with diabetes (713 (86%) with known diabetes; 116 (14%) with screen-detected diabetes) and 12,747 without diabetes (11,981 (94%) healthy controls; 766 (6%) with pre-diabetes). Mean (95% CI) blood oxygen saturation was 96.3% (96.3% to 96.4%) in diabetes which was lower than in non-diabetes [97.3% (97.2-97.3%)] after adjustment for age, gender, and smoking (p < 0.001), but significance was lost after adjustment for BMI (p = 0.25). Sub-groups with pre-diabetes and screen-detected diabetes had lower blood oxygen saturations than healthy controls (p-values < 0.01). Lower blood oxygen saturation was associated with higher UACR. Conclusions: Persons with pre-diabetes and screen-detected diabetes have sub-clinical hypoxemia, which is associated with albuminuria.