AI models offer hope for personalized AMD treatment plans

(Image credit: Adobe Stock/RocknRoller Studios)

Age-related macular degeneration (AMD), a leading cause of blindness in developed countries, continues to present significant challenges as the number of cases rises globally. Neovascular AMD (nAMD) is commonly managed with intravitreal anti-vascular endothelial growth factor (VEGF) injections, guided by optical coherence tomography (OCT) imaging to evaluate macular anatomy.

However, while OCT remains indispensable for identifying fluid-related abnormalities such as intraretinal fluid (IRF), subretinal fluid (SRF), and pigment epithelial detachment (PED), its ability to predict visual and anatomical outcomes following treatment is limited.

A team of Korean researchers, led by co-first authors Jeong Mo Han, Kong Eye Hospital, Seoul, Korea; Department of Ophthalmology, Seoul National University College of Medicine, Seoul, Korea; Junseo Ko, Department of Applied Artificial Intelligence, Sungkyunkwan University, Seoul, Korea; and Jinyoung Han, Department of Applied Artificial Intelligence, Sungkyunkwan University, Seoul, Korea, Department of Human-Artificial Intelligence Interaction, Sungkyunkwan University, Seoul, Korea, presented the development and evaluation of an AI model that predicts the outcome of intravitreal anti-VEGF injections in nAMD based on OCT images.¹

Emerging research suggests artificial intelligence (AI) may bridge this gap, offering predictive insights to refine treatment plans. Recent studies have used machine learning algorithms and generative adversarial networks (GANs) to analyze OCT images and predict responses to anti-VEGF therapy. By segmenting fluid volumes such as SRF, IRF, and PED, AI models have demonstrated the potential to forecast post-treatment macular status and functional improvement more accurately than traditional imaging alone. These advancements could empower ophthalmologists to provide patients with tailored prognoses and optimize treatment strategies during the early stages of therapy.

This study also investigated whether the incorporation of OCT images acquired after the first and second anti-VEGF injections, as well as the images acquired prior to treatment, improved the predictive value of the model. The results revealed that the AI model outperformed ophthalmologists in treatment outcome prediction, which further improved with additional OCT images during the loading phase.

The study included 2068 images from 517 eyes of patients with nAMD (mean age, 71.4 ± 9.0 years; range, 65–78 years). Ranibizumab was used in 71% of cases, and aflibercept in 29%. Researchers investigated whether incorporating OCT images acquired before treatment, as well as after the first and second anti-VEGF injections, improved the AI model's predictive accuracy.¹

Many studies have demonstrated the high accuracy of AI in aiding diagnosis. However, few studies have predicted treatment outcomes or recommended individualized and tailored treatments.

In a study that predicted the treatment outcomes of patients with macular disease treated with anti-VEGF injections using AI, investigators attempted to predict the treatment burden using OCT images and demographic information and developed an AI model that classified whether the average treatment interval between injections was low, high, or moderate.²

In that study, investigators constructed a model utilizing a GAN to predict the effects of a single treatment based on OCT images obtained before treatment in patients with typical nAMD. The accuracy of predicting the wet or dry macular state by doctors assessing OCT images generated by AI following treatment has been examined. To predict the outcome of AMD treatment, a study was conducted to estimate the prognosis of treatment based on OCT images obtained by a conditional GAN using OCT images prior to treatment and after three loading treatments.^3,4

A total of 90% of the synthetic OCT images produced by this model revealed pathological lesions similar to the actual post-treatment images. Based on these OCT images, clinicians assessed the treatment effect. The dry-up prediction sensitivity and specificity of IRF and SRF were 33.3% and 95.1%, and 21.2% and 95.1%, respectively. The addition of fluorescein angiography (FA) and indocyanine green angiography (ICGA) images improved the IRF and SRF to 33.3% and 98.4%, and 24.2% and 99.0% in the SRF.

“In our study, pre-injection OCT images were used to predict the inactive state after three anti-VEGF injections,” the researchers noted.

“In contrast to earlier studies,^5,6 we utilized a convolutional neural network (CNN) instead of a GAN,” the researchers wrote. “The clinicians did not evaluate the pathological lesions in the GAN-generated images, and the rate of the inactive state was immediately depicted quantitatively.”

Specifically, the pre-injection images and the OCT images during the loading injection treatment were learned using the AI model in various ways (average, attention, LSTM, and concatenation), and the prediction value was enhanced.

Fusion methods (concatenation, average, attention, LSTM) were chosen based on clinical considerations to enhance OCT image interpretability in nAMD treatment. Each method has a specific role: LSTM captures temporal information, attention focuses on critical regions, and average assesses overall disease severity.^7,8

These choices ensure the model interprets temporal changes, highlights crucial areas, and assesses disease severity accurately. Most significantly, the concatenation fusion layer was selected to preserve detailed information without reducing the vector size, preventing potential information loss. This approach allows the model to capture and utilize a richer set of features for improved performance in predicting treatment outcomes for nAMD, demonstrating superior results.

“Our study is the first study to attempt to compare the prediction results of AI with those of ophthalmologists and retinal specialists regarding whether the nAMD status would become inactive after treatment, as well as predict the results after three loading injections using not only images before treatment but also images taken during the treatment process,” the researchers wrote.

Upon predicting the treatment outcomes after 3 injections, retinal specialists demonstrated only approximately 80% accuracy in their predictions, as cases with slight SRF or IRF remain, rendering determination of treatment effectiveness challenging.

“The use of AI also did not demonstrate a significantly better performance than the experts in this complex task,” they wrote.

However, the researchers acknowledged the study has a number of limitations.

“First, this was a retrospective study with a small number of patients and was conducted in only one hospital using only one OCT examination device. Second, the use of ranibizumab and aflibercept was not differentiated, and the AMD subtypes were not categorized separately. Third, the study was conducted only in the Korean population, and the treatment response and performance of patients of other races, especially in Western countries, should be investigated in future studies. Fourth, only anatomical responses were evaluated using OCT without assessing functional improvements such as visual acuity. Fifth, recently developed drugs such as brolucizumab and faricimab were not included in this study.”

Although the differences in these drugs could potentially lead to variations in anatomical outcomes, this study did not investigate differences between the drugs. Finally, the study did not investigate whether the performance of the model could be improved by using other modalities, such as FA or ICGA, in addition to OCT.

Future directions such as functional improvement assessment, exploring other imaging modalities, and conducting multicenter trials with long-term follow-up would be invaluable in investigating the AI model’s predictive accuracy.

The researchers noted the study led to the development of an AI model that predicts the dry-up status after 3 loading treatments using OCT images before treatment in patients with nAMD and compared the model’s performance with that of ophthalmologists.

“The model demonstrated a higher mean performance than ophthalmologists, and the model’s treatment prediction performance further improved with additional OCT images during the loading phase,” they concluded.

Future studies should address the limitations of the present study to improve the generalizability and clinical applicability of this model.

References

1. Han, J.M., Han, J., Ko, J. et al. Anti-VEGF treatment outcome prediction based on optical coherence tomography images in neovascular age-related macular degeneration using a deep neural network. Sci Rep 14, 28253 (2024). https://doi.org/10.1038/s41598-024-79034-6

2. Hwang, D. D. et al. Distinguishing retinal angiomatous proliferation from polypoidal choroidal vasculopathy with a deep neural network based on optical coherence tomography. Sci. Rep. 11, 9275. https://doi.org/10.1038/s41598-021-88543-7 (2021).

3. Yoon, J. et al. Classifying central serous chorioretinopathy subtypes with a deep neural network using optical coherence tomography images: A cross-sectional study. Sci. Rep. 12, 422. https://doi.org/10.1038/s41598-021-04424-z (2022).

4. Ferrara, D., Newton, E. M. & Lee, A. Y. Artificial intelligence-based predictions in neovascular age-related macular degeneration. Curr. Opin. Ophthalmol. 32, 389–396. https://doi.org/10.1097/ICU.0000000000000782 (2021).

5. Jung, J. et al. Prediction of neovascular age-related macular degeneration recurrence using optical coherence tomography images with a deep neural network. Sci. Rep. 14, 5854 (2024).

6. Gallardo, M. et al. Machine learning can predict anti-VEGF treatment demand in a treat-and-extend regimen for patients with neovascular AMD, DME, and RVO associated macular edema. Ophthalmol. Retina 5, 604–624. https://doi.org/10.1016/j.oret.2021.05.002 (2021).

7. Liu, Y. et al. Prediction of OCT images of short-term response to anti-VEGF treatment for neovascular age-related macular degeneration using generative adversarial network. Br. J. Ophthalmol. 104, 1735–1740. https://doi.org/10.1136/bjophthalmol-2019-315338 (2020).

8. Lee, H., Kim, S., Kim, M. A., Chung, H. & Kim, H. C. Post-treatment prediction of optical coherence tomography using a conditional generative adversarial network in age-related macular degeneration. Retina 41, 572–580. https://doi.org/10.1097/IAE.0000000000002898 (2021).