Diabetic retinopathy (DR) is a common microvascular complication of diabetes mellitus and the leading cause of preventable blindness in working-age individuals. Early screening for DR is a public health issue. The gold standard is fundoscopy interpreted by an ophthalmologist. In France, in 2020, it was estimated that one-third of diabetic patients had not seen an ophthalmologist in over two years. Recently, a portable electroretinogram (ERG), which provides a physiological and non-invasive assessment of retinal cell function, has been developed. It combines pupillary responses through an infrared-sensitive camera, the implicit time and the response amplitude to brief white light stimuli. The result is expressed as an overall score (DR score), considered abnormal if it is \< 7 or \> 23.4. The sensitivity of the ERG for screening diabetic retinopathy is estimated at 80%, and its specificity at 82%, compared to conventional imaging. Additionally, it has been shown that poor glycemic control is associated with an increase in implicit time in preclinical DR. This test could therefore detect earlier changes than imaging results. To date, it is not considered the gold standard in France. The investigators aim to test this hypothesis. The investigators propose a prospective, bi-center study to evaluate the performance of the ERG, performed and interpreted by an Advanced Practice Nurse (APN), compared to the gold standard of retinal imaging (retinography) interpreted remotely by an ophthalmologist.