Edad retiniana como biomarcador predictivo del grado de retinopatía diabética / Retinal age as a predictive biomarker of the diabetic retinopathy grade

Objetivo Aplicar técnicas de inteligencia artificial, mediante algoritmos de aprendizaje profundo, para el desarrollo y optimización de un sistema de predicción de la edad de una persona con base en una retinografía color, y estudiar una posible relación entre la evolución de la retinopatía diabética (RD) y un envejecimiento prematuro de la retina. Métodos Se entrenó una red convolucional para calcular la edad de una persona con base en una retinografía. Dicho entrenamiento fue realizado sobre un conjunto de retinografías de pacientes con diabetes previamente dividido en 3 subconjuntos (entrenamiento, validación y test). La diferencia entre la edad cronológica del paciente y la edad biológica de la retina se definió como gap de edad retiniano. Resultados Se utilizó un conjunto de 98.400 imágenes para la fase de entrenamiento, 1.000 imágenes para la fase de validación y 13.544 para la fase de test. El gap retiniano de los pacientes sin RD fue de 0,609 años y el de los pacientes con RD de 1,905 años (p<0,001), siendo la distribución por grado de RD de: RD leve 1,541 años; RD moderada 3,017 años; RD severa 3,117 años, y RD proliferativa 8,583 años. Conclusiones El gap de edad retiniano muestra una diferencia en positivo de media entre las personas diabéticas con RD frente a las que no tienen RD, y además aumenta progresivamente, de acuerdo con el grado de RD. Estos resultados podrían indicar la existencia de una relación entre la evolución de la enfermedad y un envejecimiento prematuro de la retina (AU)

Objective To apply artificial intelligence techniques, through deep learning algorithms, for the development and optimization of a system for predicting the age of a person based on a color retinography, and to study a possible relationship between the evolution of retinopathy diabetes (RD) and premature aging of the retina. Methods A convolutional network was trained to calculate the age of a person based on a retinography. Said training was carried out on a set of retinographies of patients with diabetes previously divided into 3 subsets (training, validation and test). The difference between the chronological age of the patient and the biological age of the retina was defined as the retinal age gap. Results A set of 98,400 images was used for the training phase, 1000 images for the validation phase and 13,544 for the test phase. The retinal gap of the patients without RD was 0.609 years and that of the patients with RD was 1905 years (p<0.001), with the distribution by degree of RD being: mild RD 1541 years; moderate RD 3017 years; RD severe 3117 years, and proliferative RD 8583 years. Conclusions The retinal age gap shows a positive mean difference between diabetics with RD versus those without RD, and it increases progressively, according to the degree of RD. These results could indicate the existence of a relationship between the evolution of the disease and premature aging of the retina (AU)

Retinal age as a predictive biomarker of the diabetic retinopathy grade.

OBJECTIVE: To apply artificial intelligence (AI) techniques, through deep learning algorithms, for the development and optimization of a system for predicting the age of a person based on a color retinography and to study a possible relationship between the evolution of retinopathy diabetes and premature ageing of the retina. METHODS: A convolutional network was trained to calculate the age of a person based on a retinography. Said training was carried out on a set of retinographies of patients with diabetes previously divided into three subsets (training, validation and test). The difference between the chronological age of the patient and the biological age of the retina was defined as the retinal age gap. RESULTS: A set of 98,400 images was used for the training phase, 1000 images for the validation phase and 13,544 for the test phase. The retinal gap of the patients without DR was 0.609 years and that of the patients with DR was 1905 years (p < 0.001), with the distribution by degree of DR being: mild DR: 1541 years, moderate DR: 3017 years, DR severe: 3117 years and proliferative DR: 8583 years. CONCLUSIONS: The retinal age gap shows a positive mean difference between diabetics with DR versus those without DR, and it increases progressively, according to the degree of DR. These results could indicate the existence of a relationship between the evolution of the disease and premature ageing of the retina.