2023-10-19 17:47:04
In Italy, there are over 1 million people with diabetes who suffer from diabetic retinopathy, the main complication of diabetes and the leading cause of blindness in working age. All national and international diabetes guidelines, for both type 1 and type 2 diabetes, recommend regular and early evaluation of the state of the retina and immediate intervention, if necessary. However, access to this type of screening in developed countries is low, partly due to the complexity and cost of the procedures and partly because in many countries, including Italy, this screening is carried out by ophthalmologists, burdening waiting lists. An algorithm can make the difference.
In a study recently published in Acta Diabetologica, Dairet* (Diabetes Artificial Intelligence for RETinopathy), an artificial intelligence system for the automated evaluation of diabetic retinopathy, demonstrated 100% sensitivity in identifying moderate or severe cases, that is, forms of the disease that can put vision at risk and therefore require the intervention of an ophthalmologist. And consequently to be a tool to speed up the diagnostic process, reducing the burden on specialists and the waiting time for patients. The evidence emerging from the Italian study “Feasibility and accuracy of the screening for diabetic retinopathy using a fundus camera and an artificial intelligence pre-evaluation”, – reports a note – offers scientific diabetes societies food for thought regarding the possibility of applying this new method screening in daily clinical practice.
The study – the first in Italy on artificial intelligence in diabetic retinopathy conducted on 637 patients followed by the diabetes and ophthalmology centers of the ASL Torino 5 – is based on the need to find simplified and low-cost procedures to implement retinopathy screening in large populations , as is that of people with diabetes, and request the intervention of ophthalmologists only in strictly necessary cases.
“The artificial intelligence algorithm has proven to be very efficient in identifying moderate and severe retinopathy, with the certainty that no patient requiring an ophthalmologist is incorrectly diagnosed as negative – explains Carlo Bruno Giorda, principal investigator of the study – Given the ever-increasing number of people with diabetes and the importance of this screening, which is often not carried out due to long waiting lists, it has become necessary to use automated image evaluation systems to speed up the diagnostic process , reducing the burden on specialists and the waiting time for patients.”
Obviously “we are not talking about replacing the professionalism of the ophthalmologist, but about giving him important support in the most complex phases – underlines the expert – Furthermore, we hope that the evidence emerging from the study will offer scientific diabetes societies food for thought regarding the possibility of apply this new screening method in daily clinical practice”.
Dairet* – concludes the note – has also shown a sensitivity level of 100 percent for moderate or severe forms in subjects over 70 years of age, even if senile eye diseases slightly decrease the number of patients subjected to this test . The system is developed by Retmarker, a Portuguese company controlled by the Italian Meteda group, which boasts extensive international experience with over 500 thousand patients examined. The device, which can be used by non-specialised nursing staff, is managed within the MètaClinic electronic medical record, currently installed in 90 percent of Italian diabetes care centers and used regularly by healthcare staff, thus allowing the doctor to have all the patient’s clinical data available with a simple click.
#Diabetic #retinopathy #million #Italians #early #diagnosis #algorithm