Retinopatia cukrzycowa
Rokowania, prognozy i postęp choroby

Retinopatia cukrzycowa (RC) jest istotnym powikłaniem mikronaczyniowym cukrzycy, prowadzącym do uszkodzenia naczyń i neuronów siatkówki, stanowiąc jedną z głównych przyczyn ślepoty w populacji osób w wieku produkcyjnym (20-64 lata). Rokowanie zależy od stadium choroby i obecności cukrzycowego obrzęku plamki, który jest najczęstszą przyczyną utraty wzroku. Wczesne wykrycie i interwencja mogą zapobiec upośledzeniu widzenia w ponad 90% przypadków. U pacjentów z proliferacyjną retinopatią cukrzycową (PDR) 10-letnie przeżycie wynosi 76% (95% CI: 63-85%), a ryzyko zgonu jest czterokrotnie wyższe u osób ≥50 lat. Leczenie laserowe i witrektomia pozostają kluczowymi metodami terapeutycznymi, redukując ryzyko ciężkiej utraty wzroku (≤ 5/200) o ponad 50%. Obrzęk plamki wymaga często powtarzanych iniekcji leków, a laseroterapia zmniejsza częstość umiarkowanej utraty wzroku z 30% do 15% w ciągu 3 lat.

Retinopatia cukrzycowa – rokowanie (prognoza wyników)

Retinopatia cukrzycowa, określana również jako cukrzycowa choroba oczu, stanowi poważne powikłanie mikronaczyniowe cukrzycy, będące jedną z głównych przyczyn ślepoty w populacji osób w wieku produkcyjnym (20-64 lata). Jest to stan, w którym dochodzi do uszkodzenia małych naczyń krwionośnych i neuronów siatkówki wskutek cukrzycy.1 Rokowanie w retinopatii cukrzycowej zależy od wielu czynników, w tym od stadium choroby w momencie diagnozy oraz skuteczności leczenia. Wczesne wykrycie i interwencja znacząco poprawiają wyniki i zmniejszają ryzyko utraty wzroku.2

Czynniki wpływające na rokowanie

Rokowanie wzrokowe silnie zależy od stopnia zaawansowania retinopatii cukrzycowej, a przede wszystkim od tego, czy towarzyszy jej cukrzycowy obrzęk plamki. Gdy choroba zostanie zdiagnozowana wcześnie, można ją ustabilizować przy pomocy nowoczesnych metod leczenia, co zwykle nie wpływa na pogorszenie widzenia pacjenta.3 Należy jednak zaznaczyć, że u niektórych osób z przewlekłym, opornym na leczenie obrzękiem plamki może nigdy nie dojść do odzyskania centralnego widzenia.4

Badania epidemiologiczne wskazują, że wczesne wykrycie i szybka interwencja mogą zapobiec upośledzeniu widzenia w ponad 90% przypadków retinopatii cukrzycowej, co podkreśla kliniczne i zdrowotne znaczenie wdrażania programów przesiewowych.5 Im dłużej osoba choruje na cukrzycę, tym większe jest ryzyko rozwoju retinopatii cukrzycowej – dotyczy to nawet 80% osób chorujących na cukrzycę przez 20 lat lub dłużej.6

Wpływ retinopatii cukrzycowej na śmiertelność

Obecność retinopatii cukrzycowej u osób starszych z cukrzycą jest istotnie związana ze zwiększoną ogólną śmiertelnością oraz śmiertelnością z powodu chorób układu sercowo-naczyniowego (CVD). Wyniki analiz National Health and Nutrition Examination Survey (NHANES) wskazują, że zarówno sama obecność retinopatii cukrzycowej, jak i jej nasilenie (łagodna, umiarkowana do ciężkiej postać nieproliferacyjnej retinopatii cukrzycowej) są istotnie związane ze zwiększonym ryzykiem śmiertelności ogólnej i śmiertelności z powodu chorób układu sercowo-naczyniowego.7

Obecność retinopatii cukrzycowej i jej ciężkość są również znacząco związane ze zwiększonym ryzykiem śmiertelności specyficznej dla cukrzycy, a ryzyko zgonu z powodu cukrzycy wzrasta wykładniczo wraz z nasileniem retinopatii. W porównaniu z pacjentami bez retinopatii cukrzycowej, ryzyko śmiertelności spowodowanej cukrzycą wzrasta wykładniczo wraz ze wzrostem ciężkości retinopatii.8

Duńskie ogólnokrajowe badanie kohortowe wykazało, że u dorosłych z cukrzycą typu 1 ryzyko występowania i zachorowania na choroby układu sercowo-naczyniowego jest większe w przypadku obecności retinopatii cukrzycowej.9

Długoterminowe wyniki kliniczne

Analizy dziesięcioletnich wyników u pacjentów skierowanych z powodu proliferacyjnej retinopatii cukrzycowej (PDR) z programów badań przesiewowych w kierunku retinopatii cukrzycowej do ośrodków opieki trzeciorzędowej w Wielkiej Brytanii wykazały, że dziesięcioletnie przeżycie u wszystkich pacjentów z PDR wynosiło 76% (95% CI: 63 do 85%), a średni czas do zgonu dla wszystkich zmarłych pacjentów wynosił 5,4±3,6 lat.10 Ryzyko zgonu w dowolnym momencie było czterokrotnie wyższe u osób w wieku 50 lat lub starszych w porównaniu z grupą poniżej 50 roku życia (95% CI: 1,33 do 13,3, p=0,008).11

W 10-letniej obserwacji ciężka utrata wzroku (SVL) nie była częstym zjawiskiem (7% oczu), ale prawie jedna piąta oczu spadła poniżej brytyjskich standardów widzenia wymaganych do prowadzenia pojazdów. Witrektomia jest ważną interwencją w leczeniu tych pacjentów, wymaganą w około 25% oczu, dlatego zaleca się ścisłą obserwację, szczególnie w ciągu pierwszych 3 lat po rozpoznaniu.12

Skuteczność interwencji terapeutycznych

Obrzęk plamki żółtej jest najczęstszą formą utraty wzroku u osób z cukrzycą, szczególnie jeśli pozostaje nieleczony. Jest to przewlekłe powikłanie i nawet przy prawidłowym leczeniu nie ma definitywnego wyleczenia. Plany leczenia powinny być zindywidualizowane dla każdego pacjenta, ale często wymagane są powtarzane iniekcje leków, aby kontrolować nawroty obrzęku.13

Badanie Early Treatment for Diabetic Retinopathy Study wykazało, że zabieg laserowy w leczeniu obrzęku plamki zmniejsza częstość występowania umiarkowanej utraty wzroku (podwojenie kąta widzenia lub w przybliżeniu utrata 2 linii ostrości wzroku) z 30% do 15% w okresie 3 lat.14

Z kolei badanie Diabetic Retinopathy Study wykazało, że odpowiednia panretinalna fotokoagulacja laserowa redukuje ryzyko ciężkiej utraty wzroku (≤ 5/200) o ponad 50%.15

Nowoczesne metody prognozowania retinopatii cukrzycowej

Obecnie rozwijane są zaawansowane metody prognozowania rozwoju retinopatii cukrzycowej w oparciu o uczenie maszynowe. Model oparty na algorytmie XGBoost skutecznie ocenia ryzyko retinopatii cukrzycowej u pacjentów z cukrzycą typu 2 przy użyciu rutynowych danych laboratoryjnych, pomagając klinicystom w identyfikacji osób z wysokim ryzykiem i kierowaniu zindywidualizowanych strategii postępowania, szczególnie na obszarach o ograniczonym dostępie do opieki medycznej.16

Proponowany model oparty na XGBoost, wykorzystujący wartości SHAP, zidentyfikował skurczowe ciśnienie tętnicze (SBP) jako najistotniejszy czynnik ryzyka retinopatii cukrzycowej.17

Sztuczna inteligencja (AI), w szczególności głębokie uczenie i konwolucyjne sieci neuronowe (CNN), wykazała niezwykłą dokładność w analizie obrazów siatkówki, identyfikując wczesne stadium retinopatii cukrzycowej z wysoką czułością i swoistością.18 Integracja AI z telemedycyną dodatkowo zwiększyła dostępność, umożliwiając zdalne badania przesiewowe za pomocą urządzeń przenośnych i obrazowania opartego na smartfonach. Pod względem ekonomicznym systemy oparte na AI zmniejszają koszty opieki zdrowotnej poprzez optymalizację alokacji zasobów i minimalizację niepotrzebnych skierowań.19

Algorytmy AI są coraz częściej stosowane do analizy obrazów siatkówki, wykorzystując zarówno konwolucyjne sieci neuronowe (CNN), jak i techniki głębokiego uczenia do identyfikacji subtelnych zmian naczyniowych i mikromalakteryzmatów, które wskazują na retinopatię cukrzycową w jej najwcześniejszych stadiach. Potencjał tych metod w pomaganiu okulistom w zmniejszaniu zmienności interpretacji obrazów i poprawie przepustowości badań przesiewowych jest podkreślany przez gwałtowne postępy obserwowane w ciągu ostatniej dekady, które znacznie zmniejszyły odsetek błędnych diagnoz, jednocześnie zwiększając możliwości prognostyczne.20

Narzędzia prognostyczne dla lekarzy

Opracowano narzędzia oparte na uczeniu maszynowym, które mogą prognozować retinopatię cukrzycową na podstawie wyników badań krwi. Celem jest przewidywanie, czy pacjent będzie cierpiał na retinopatię cukrzycową na podstawie cech z testu krwi. Prognoza przyjmuje wartość 1, jeśli pacjent ma retinopatię, i 0, jeśli jej nie ma.21

Proces prognostyczny polega na obliczeniu wyników sieci neuronowej na podstawie zmiennych wejściowych pacjenta. Przykładowo, prognoza 0,85 oznacza, że pacjent ma 85% prawdopodobieństwo cierpienia na retinopatię cukrzycową. Wyrażenie matematyczne modelu może być eksportowane do oprogramowania diagnostycznego używanego przez lekarzy.22

Podsumowanie rokowania w retinopatii cukrzycowej

Rokowanie w retinopatii cukrzycowej różni się w zależności od wielu czynników, w tym stadium choroby, obecności obrzęku plamki oraz odpowiedzi na leczenie. Wczesne wykrycie i leczenie znacząco poprawiają rokowanie, a niedawne innowacje w leczeniu pokazały obiecujące wyniki w zachowaniu wzroku i poprawie wyników u pacjentów.23

Najnowsze badania podkreślają również znaczenie retinopatii cukrzycowej jako markera zwiększonego ryzyka śmiertelności ogólnej oraz z przyczyn sercowo-naczyniowych u osób z cukrzycą. Stopień nasilenia retinopatii cukrzycowej może odzwierciedlać ciężkość stanu choroby sercowo-naczyniowej i ogólne ryzyko śmiertelności u pacjentów z cukrzycą.24

Włączenie sztucznej inteligencji do protokołów badań przesiewowych w kierunku retinopatii cukrzycowej może zrewolucjonizować opiekę nad pacjentem, zapewniając wcześniejszą diagnozę, umożliwiając terminowe interwencje terapeutyczne i ostatecznie zmniejszając częstość występowania ślepoty związanej z cukrzycą.25

Kolejne rozdziały

Zapraszamy do dalszego czytania naszego leksykonu.

Wybierz kolejny rozdział z menu poniżej, aby otworzyć nową podstronę kompedium wiedzy i uzyskać szczegółowe informację o leku, substancji lub chorobie.

  1. 12.04.2026
  2. www.leksykon.com.pl

Materiały źródłowe

  • #1 Convolutional Neural Network (CNN) in Machine Learning | GeeksforGeeks
    https://www.geeksforgeeks.org/convolutional-neural-network-cnn-in-machine-learning/
    Diabetic retinopathy also known as diabetic eye disease, is a medical state in which destruction occurs to the retina due to diabetes mellitus, It is a major cause of blindness in advance countries. […] Diabetic retinopathy influence up to 80 percent of those who have had diabetes for 20 years or more. […] The overlong a person has diabetes, the higher his or her chances of growing diabetic retinopathy. […] It is also the main cause of blindness in people of age group 20-64. […] Diabetic retinopathy is the outcome of destruction to the small blood vessels and neurons of the retina.
  • #2
    https://www.linkedin.com/pulse/from-pixels-predictions-how-artificial-intelligence-can-kulkarni
    Prognosis […] The prognosis of diabetic retinopathy depends on various factors, including the stage of the disease at diagnosis and the effectiveness of treatment. Early detection and intervention significantly improve outcomes and reduce the risk of vision loss. However, if left untreated or inadequately managed, diabetic retinopathy can progress to advanced stages where irreversible vision loss occurs. […] „Diabetic retinopathy is a preventable complication of diabetes and a leading cause of blindness in adults worldwide.” […] Conclusion […] Recent innovations in its management have shown promising results in preserving vision and improving patient outcomes. Laser therapy, anti-VEGF injections, and AI-based screening methods are among the notable advancements in this field. Furthermore, AI has the potential to assist in various aspects of managing diabetic complications, including screening, risk stratification, treatment planning, and remote monitoring. While challenges remain in implementing these innovations globally, AI still offers a lot of hope for reducing complications in LMICs with a high prevalence of diabetes.
  • #3 Prognosis for Diabetic Retinopathy | Hospital Clínic Barcelona
    https://www.clinicbarcelona.org/en/assistance/diseases/diabetic-retinopathy/evolution-of-the-disease
    The visual prognosis strongly depends on the stage of the diabetic retinopathy and, above all, whether it is accompanied by diabetic macular oedema. […] When the condition is diagnosed early, it can be stabilised with the aid of modern treatments and usually does not affect the patients vision. […] Some people with chronic, treatment-resistant macular oedema may never recover their central vision. […] Macular oedema is the most common form of vision loss in people with diabetes, particularly if it goes untreated. […] It is a chronic complication and even when treated correctly there is no definitive cure. […] Treatment plans should be individualised to each patient, but repeated drug injections are often required to control recurrences of the oedema.
  • #4 Prognosis for Diabetic Retinopathy | Hospital Clínic Barcelona
    https://www.clinicbarcelona.org/en/assistance/diseases/diabetic-retinopathy/evolution-of-the-disease
    The visual prognosis strongly depends on the stage of the diabetic retinopathy and, above all, whether it is accompanied by diabetic macular oedema. […] When the condition is diagnosed early, it can be stabilised with the aid of modern treatments and usually does not affect the patients vision. […] Some people with chronic, treatment-resistant macular oedema may never recover their central vision. […] Macular oedema is the most common form of vision loss in people with diabetes, particularly if it goes untreated. […] It is a chronic complication and even when treated correctly there is no definitive cure. […] Treatment plans should be individualised to each patient, but repeated drug injections are often required to control recurrences of the oedema.
  • #5 Artificial Intelligence for Early Detection and Prognosis Prediction of Diabetic Retinopathy | medRxiv
    https://www.medrxiv.org/content/10.1101/2025.03.29.25324873v1.full-text
    Epidemiological studies indicate that early detection and prompt intervention can prevent visual impairment in over 90% of DR cases, a fact that underscores the clinical and public health significance of deploying AI in screening programs. […] The incorporation of AI into DR screening protocols is poised to revolutionize patient care by ensuring earlier diagnosis, enabling timely therapeutic interventions, and ultimately reducing the incidence of diabetes-related blindness. […] The integration of artificial intelligence (AI) in the early detection and prognosis prediction of diabetic retinopathy (DR) represents a transformative advancement in ophthalmology, addressing critical challenges in screening accuracy, accessibility, and cost-effectiveness. […] This review highlights the dominance of deep learning techniques, particularly convolutional neural networks (CNNs), in analyzing retinal images with high sensitivity and specificity, outperforming traditional manual grading methods. […] By addressing current limitations and fostering innovation, AI can fulfill its promise in reducing global diabetes-related blindness.
  • #6 Convolutional Neural Network (CNN) in Machine Learning | GeeksforGeeks
    https://www.geeksforgeeks.org/convolutional-neural-network-cnn-in-machine-learning/
    Diabetic retinopathy also known as diabetic eye disease, is a medical state in which destruction occurs to the retina due to diabetes mellitus, It is a major cause of blindness in advance countries. […] Diabetic retinopathy influence up to 80 percent of those who have had diabetes for 20 years or more. […] The overlong a person has diabetes, the higher his or her chances of growing diabetic retinopathy. […] It is also the main cause of blindness in people of age group 20-64. […] Diabetic retinopathy is the outcome of destruction to the small blood vessels and neurons of the retina.
  • #7 Association of diabetic retinopathy on all-cause and cause-specific mortality in older adults with diabetes: National Health and Nutrition Examination Survey, 2005–2008 | Scientific Reports
    https://www.nature.com/articles/s41598-024-58502-z
    To evaluate the effect of diabetic retinopathy (DR) status or severity on all-cause and cause-specific mortality among diabetic older adults in the United States using the most recent National Health and Nutrition Examination Survey (NHANES) follow-up mortality data. […] DR was associated with increased all-cause, cardiovascular disease (CVD) and diabetes mellitus (DM)-specific mortality, which remained consistent after propensity score matching (PSM). […] The presence of DR in elderly individuals with diabetes is significantly associated with the elevated all-cause and CVD mortality. […] The grading or severity of DR may reflect the severity of cardiovascular disease status and overall mortality risk in patients with diabetes. […] The results showed that the presence of any DR, as well as mild, moderate to severe NPDR, are significantly associated with an increased risk of all-cause mortality and CVD-specific mortality.
  • #8 Association of diabetic retinopathy on all-cause and cause-specific mortality in older adults with diabetes: National Health and Nutrition Examination Survey, 2005–2008 | Scientific Reports
    https://www.nature.com/articles/s41598-024-58502-z
    The presence of any DR and its severity are also significantly associated with an increased risk of DM-specific mortality, and the risk of mortality due to DM increases exponentially with the severity of DR. […] Compared with patients without DR, the risk of DM-caused mortality increases exponentially with increasing DR severity. […] Our study has several advantages. Firstly, we used a large-scale nationally representative sample of non-hospitalized elderly individuals with diabetes aged 40 and above in the United States and a standardized DR grading assessment method. […] The use of multiple confounding factor-adjusted competing risk regression models and PSM methods clarifies the relationship between DR and increased mortality risk due to CVD and DM, emphasizing that DR grading may serve as an effective predictive indicator for continuous monitoring of vascular status in diabetes patients and could have significant value in regular screening and future prognostic assessment.
  • #9 Diabetic Retinopathy: Practice Essentials, Pathophysiology, Etiology
    https://emedicine.medscape.com/article/1225122-overview
    Prognostic factors that are favorable for visual loss include the following: […] Prognostic factors that are unfavorable for visual loss include the following: […] The Early Treatment for Diabetic Retinopathy Study has found that laser surgery for macular edema reduces the incidence of moderate visual loss (doubling of visual angle or roughly a 2-line visual loss) from 30% to 15% over a 3-year period. […] The Diabetic Retinopathy Study has found that adequate scatter laser panretinal photocoagulation reduces the risk of severe visual loss ( 5/200) by more than 50%. […] In a Danish nationwide matched case-cohort study, Mabala et al found evidence that in adults with type 1 diabetes mellitus, the risk for prevalent and incident cardiovascular disease (CVD) is greater in the presence of diabetic retinopathy.
  • #10 Referrals for proliferative diabetic retinopathy from two UK diabetic retinopathy screening services: a 10-year analysis of visual outcomes, requirement for vitrectomy, and mortality | Eye
    https://www.nature.com/articles/s41433-024-03078-1
    To determine long-term outcomes of patients referred with proliferative diabetic retinopathy (PDR) from diabetic eye screening programmes (DESP) to tertiary care centres in the United Kingdom (UK). […] Critical outcomes included severe vision loss (SVL) and the need for vitrectomy. […] The 10-year survival in all PDR patients was 76% (95% CI: 63 to 85%) with the mean time to death for all deceased patients being 5.43.6 years. […] During the 10 year follow up SVL was uncommon, but MVL occurred in almost one-fifth of the eyes. Approximately 1 in 4 eyes required vitrectomy, highlighting its significance in patient management. […] The 10-year survival in all R3a patients was 76.0% (95% CI: 63.0% to 85.0%). […] The risk of death at any time was four times higher in those aged 50 years or over compared to the under 50 years group (95% CI: 1.33 to 13.3, p=0.008).
  • #11 Referrals for proliferative diabetic retinopathy from two UK diabetic retinopathy screening services: a 10-year analysis of visual outcomes, requirement for vitrectomy, and mortality | Eye
    https://www.nature.com/articles/s41433-024-03078-1
    To determine long-term outcomes of patients referred with proliferative diabetic retinopathy (PDR) from diabetic eye screening programmes (DESP) to tertiary care centres in the United Kingdom (UK). […] Critical outcomes included severe vision loss (SVL) and the need for vitrectomy. […] The 10-year survival in all PDR patients was 76% (95% CI: 63 to 85%) with the mean time to death for all deceased patients being 5.43.6 years. […] During the 10 year follow up SVL was uncommon, but MVL occurred in almost one-fifth of the eyes. Approximately 1 in 4 eyes required vitrectomy, highlighting its significance in patient management. […] The 10-year survival in all R3a patients was 76.0% (95% CI: 63.0% to 85.0%). […] The risk of death at any time was four times higher in those aged 50 years or over compared to the under 50 years group (95% CI: 1.33 to 13.3, p=0.008).
  • #12 Referrals for proliferative diabetic retinopathy from two UK diabetic retinopathy screening services: a 10-year analysis of visual outcomes, requirement for vitrectomy, and mortality | Eye
    https://www.nature.com/articles/s41433-024-03078-1
    Our study provides UK real-world long-term outcomes of patients referred with PDR from systematic diabetic retinopathy screening programmes. Severe vision loss was not a common occurrence at 7% of eyes, but nearly one-fifth of the eyes fell below UK driving standards for vision. Vitrectomy is an important intervention for managing these patients, required in approximately 25% of eyes, and close observation is recommended especially during the initial 3 years following diagnosis for any complications and conversion of fellow-eyes to PDR. The outcomes of this study can assist any future risk-modelling studies to predict outcome in PDR patients.
  • #13 Prognosis for Diabetic Retinopathy | Hospital Clínic Barcelona
    https://www.clinicbarcelona.org/en/assistance/diseases/diabetic-retinopathy/evolution-of-the-disease
    The visual prognosis strongly depends on the stage of the diabetic retinopathy and, above all, whether it is accompanied by diabetic macular oedema. […] When the condition is diagnosed early, it can be stabilised with the aid of modern treatments and usually does not affect the patients vision. […] Some people with chronic, treatment-resistant macular oedema may never recover their central vision. […] Macular oedema is the most common form of vision loss in people with diabetes, particularly if it goes untreated. […] It is a chronic complication and even when treated correctly there is no definitive cure. […] Treatment plans should be individualised to each patient, but repeated drug injections are often required to control recurrences of the oedema.
  • #14 Diabetic Retinopathy: Practice Essentials, Pathophysiology, Etiology
    https://emedicine.medscape.com/article/1225122-overview
    Prognostic factors that are favorable for visual loss include the following: […] Prognostic factors that are unfavorable for visual loss include the following: […] The Early Treatment for Diabetic Retinopathy Study has found that laser surgery for macular edema reduces the incidence of moderate visual loss (doubling of visual angle or roughly a 2-line visual loss) from 30% to 15% over a 3-year period. […] The Diabetic Retinopathy Study has found that adequate scatter laser panretinal photocoagulation reduces the risk of severe visual loss ( 5/200) by more than 50%. […] In a Danish nationwide matched case-cohort study, Mabala et al found evidence that in adults with type 1 diabetes mellitus, the risk for prevalent and incident cardiovascular disease (CVD) is greater in the presence of diabetic retinopathy.
  • #15 Diabetic Retinopathy: Practice Essentials, Pathophysiology, Etiology
    https://emedicine.medscape.com/article/1225122-overview
    Prognostic factors that are favorable for visual loss include the following: […] Prognostic factors that are unfavorable for visual loss include the following: […] The Early Treatment for Diabetic Retinopathy Study has found that laser surgery for macular edema reduces the incidence of moderate visual loss (doubling of visual angle or roughly a 2-line visual loss) from 30% to 15% over a 3-year period. […] The Diabetic Retinopathy Study has found that adequate scatter laser panretinal photocoagulation reduces the risk of severe visual loss ( 5/200) by more than 50%. […] In a Danish nationwide matched case-cohort study, Mabala et al found evidence that in adults with type 1 diabetes mellitus, the risk for prevalent and incident cardiovascular disease (CVD) is greater in the presence of diabetic retinopathy.
  • #16 Predicting diabetic retinopathy based on routine laboratory tests by machine learning algorithms | European Journal of Medical Research | Full Text
    https://eurjmedres.biomedcentral.com/articles/10.1186/s40001-025-02442-5
    This study aimed to identify risk factors for diabetic retinopathy (DR) and develop machine learning (ML)-based predictive models using routine laboratory data in patients with type 2 diabetes mellitus (T2DM). […] The XGBoost-based prediction model effectively assesses DR risk in T2DM patients using routine laboratory data, aiding clinicians in identifying high-risk individuals and guiding personalized management strategies, especially in medically underserved areas. […] The XGBoost model demonstrated superior performance compared to the others. […] The proposed XGBoost-based model, incorporating SHAP values, identified SBP as the most significant risk factor for DR. […] In this study, utilizing routine laboratory tests, we developed and evaluated a ML-based model for predicting DR risk in patients with T2DM.
  • #17 Predicting diabetic retinopathy based on routine laboratory tests by machine learning algorithms | European Journal of Medical Research | Full Text
    https://eurjmedres.biomedcentral.com/articles/10.1186/s40001-025-02442-5
    This study aimed to identify risk factors for diabetic retinopathy (DR) and develop machine learning (ML)-based predictive models using routine laboratory data in patients with type 2 diabetes mellitus (T2DM). […] The XGBoost-based prediction model effectively assesses DR risk in T2DM patients using routine laboratory data, aiding clinicians in identifying high-risk individuals and guiding personalized management strategies, especially in medically underserved areas. […] The XGBoost model demonstrated superior performance compared to the others. […] The proposed XGBoost-based model, incorporating SHAP values, identified SBP as the most significant risk factor for DR. […] In this study, utilizing routine laboratory tests, we developed and evaluated a ML-based model for predicting DR risk in patients with T2DM.
  • #18 Artificial Intelligence for Early Detection and Prognosis Prediction of Diabetic Retinopathy | medRxiv
    https://www.medrxiv.org/content/10.1101/2025.03.29.25324873v1
    This review explores the transformative role of artificial intelligence (AI) in the early detection and prognosis prediction of diabetic retinopathy (DR), a leading cause of vision loss in diabetic patients. […] AI, particularly deep learning and convolutional neural networks (CNNs), has demonstrated remarkable accuracy in analyzing retinal images, identifying early-stage DR with high sensitivity and specificity. […] The integration of AI with telemedicine has further enhanced accessibility, enabling remote screening through portable devices and smartphone-based imaging. […] Economically, AI-based systems reduce healthcare costs by optimizing resource allocation and minimizing unnecessary referrals. […] Future research should focus on multi-center validation, diverse AI methodologies, and clinician-friendly tools to ensure equitable adoption. […] By addressing these gaps, AI can revolutionize DR management, reducing the global burden of diabetes-related blindness through early intervention and scalable solutions.
  • #19 Artificial Intelligence for Early Detection and Prognosis Prediction of Diabetic Retinopathy | medRxiv
    https://www.medrxiv.org/content/10.1101/2025.03.29.25324873v1
    This review explores the transformative role of artificial intelligence (AI) in the early detection and prognosis prediction of diabetic retinopathy (DR), a leading cause of vision loss in diabetic patients. […] AI, particularly deep learning and convolutional neural networks (CNNs), has demonstrated remarkable accuracy in analyzing retinal images, identifying early-stage DR with high sensitivity and specificity. […] The integration of AI with telemedicine has further enhanced accessibility, enabling remote screening through portable devices and smartphone-based imaging. […] Economically, AI-based systems reduce healthcare costs by optimizing resource allocation and minimizing unnecessary referrals. […] Future research should focus on multi-center validation, diverse AI methodologies, and clinician-friendly tools to ensure equitable adoption. […] By addressing these gaps, AI can revolutionize DR management, reducing the global burden of diabetes-related blindness through early intervention and scalable solutions.
  • #20 Artificial Intelligence for Early Detection and Prognosis Prediction of Diabetic Retinopathy | medRxiv
    https://www.medrxiv.org/content/10.1101/2025.03.29.25324873v1.full-text
    This review explores the transformative role of artificial intelligence (AI) in the early detection and prognosis prediction of diabetic retinopathy (DR), a leading cause of vision loss in diabetic patients. […] AI algorithms have been increasingly applied to analyze retinal images, leveraging both convolutional neural networks (CNNs) and deep learning techniques to identify subtle vascular changes and microaneurysms that are indicative of DR at its earliest stages. […] The potential of these methods to assist ophthalmologists in reducing variability in image interpretation and improving screening throughput is underscored by the rapid advancements observed over the last decade, which have significantly decreased the rate of misdiagnosis while simultaneously enhancing prognostic capabilities.
  • #21 Prognose diabetic retinopathy using machine learning
    https://www.neuraldesigner.com/learning/examples/diabetic-retinopathy-prognosis/
    In this example, we build a machine learning model to prognose diabetic retinopathy. […] The goal here is to predict whether a patient will suffer from diabetic retinopathy conditioned on blood test features. […] The prognosis is 1 if the patient has retinopathy and 0 if he doesn’t. […] The objective of model selection is to find the network architecture with the best generalization properties, which minimizes the error on the selected instances of the data set. […] The objective of the testing analysis is to validate the generalization performance of the trained neural network. […] We can prognosticate new patients by calculating the neural network outputs. For that, we need to know the input variables for them. An example is the following: prognosis: 0.85, so the patient has 85% probability of suffering diabetic retinopathy. […] The objective of the response optimization algorithm is to exploit the mathematical model to look for optimal operating conditions. […] The above expression can be exported anywhere, for instance, to a diagnosis software to be used by doctors.
  • #22 Prognose diabetic retinopathy using machine learning
    https://www.neuraldesigner.com/learning/examples/diabetic-retinopathy-prognosis/
    In this example, we build a machine learning model to prognose diabetic retinopathy. […] The goal here is to predict whether a patient will suffer from diabetic retinopathy conditioned on blood test features. […] The prognosis is 1 if the patient has retinopathy and 0 if he doesn’t. […] The objective of model selection is to find the network architecture with the best generalization properties, which minimizes the error on the selected instances of the data set. […] The objective of the testing analysis is to validate the generalization performance of the trained neural network. […] We can prognosticate new patients by calculating the neural network outputs. For that, we need to know the input variables for them. An example is the following: prognosis: 0.85, so the patient has 85% probability of suffering diabetic retinopathy. […] The objective of the response optimization algorithm is to exploit the mathematical model to look for optimal operating conditions. […] The above expression can be exported anywhere, for instance, to a diagnosis software to be used by doctors.
  • #23
    https://www.linkedin.com/pulse/from-pixels-predictions-how-artificial-intelligence-can-kulkarni
    Prognosis […] The prognosis of diabetic retinopathy depends on various factors, including the stage of the disease at diagnosis and the effectiveness of treatment. Early detection and intervention significantly improve outcomes and reduce the risk of vision loss. However, if left untreated or inadequately managed, diabetic retinopathy can progress to advanced stages where irreversible vision loss occurs. […] „Diabetic retinopathy is a preventable complication of diabetes and a leading cause of blindness in adults worldwide.” […] Conclusion […] Recent innovations in its management have shown promising results in preserving vision and improving patient outcomes. Laser therapy, anti-VEGF injections, and AI-based screening methods are among the notable advancements in this field. Furthermore, AI has the potential to assist in various aspects of managing diabetic complications, including screening, risk stratification, treatment planning, and remote monitoring. While challenges remain in implementing these innovations globally, AI still offers a lot of hope for reducing complications in LMICs with a high prevalence of diabetes.
  • #24 Association of diabetic retinopathy on all-cause and cause-specific mortality in older adults with diabetes: National Health and Nutrition Examination Survey, 2005–2008 | Scientific Reports
    https://www.nature.com/articles/s41598-024-58502-z
    To evaluate the effect of diabetic retinopathy (DR) status or severity on all-cause and cause-specific mortality among diabetic older adults in the United States using the most recent National Health and Nutrition Examination Survey (NHANES) follow-up mortality data. […] DR was associated with increased all-cause, cardiovascular disease (CVD) and diabetes mellitus (DM)-specific mortality, which remained consistent after propensity score matching (PSM). […] The presence of DR in elderly individuals with diabetes is significantly associated with the elevated all-cause and CVD mortality. […] The grading or severity of DR may reflect the severity of cardiovascular disease status and overall mortality risk in patients with diabetes. […] The results showed that the presence of any DR, as well as mild, moderate to severe NPDR, are significantly associated with an increased risk of all-cause mortality and CVD-specific mortality.
  • #25 Artificial Intelligence for Early Detection and Prognosis Prediction of Diabetic Retinopathy | medRxiv
    https://www.medrxiv.org/content/10.1101/2025.03.29.25324873v1.full-text
    Epidemiological studies indicate that early detection and prompt intervention can prevent visual impairment in over 90% of DR cases, a fact that underscores the clinical and public health significance of deploying AI in screening programs. […] The incorporation of AI into DR screening protocols is poised to revolutionize patient care by ensuring earlier diagnosis, enabling timely therapeutic interventions, and ultimately reducing the incidence of diabetes-related blindness. […] The integration of artificial intelligence (AI) in the early detection and prognosis prediction of diabetic retinopathy (DR) represents a transformative advancement in ophthalmology, addressing critical challenges in screening accuracy, accessibility, and cost-effectiveness. […] This review highlights the dominance of deep learning techniques, particularly convolutional neural networks (CNNs), in analyzing retinal images with high sensitivity and specificity, outperforming traditional manual grading methods. […] By addressing current limitations and fostering innovation, AI can fulfill its promise in reducing global diabetes-related blindness.