⚡ Quick Summary
A recent pilot study utilized a convolutional neural network (CNN) to predict the response of treatment-naïve diabetic macular edema (DME) to a single injection of anti-VEGF therapy. The CNN achieved an impressive AUC of 0.81, indicating its potential as a predictive tool in ophthalmology.
🔍 Key Details
- 📊 Patient Population: 73 eyes from 53 patients with new DME diagnosis
- 🧩 Data Source: Optical coherence tomography (OCT) scans
- ⚙️ Technology: Convolutional Neural Network (CNN)
- 🏆 Main Outcome Measure: Prediction of treatment response defined as a CST reduction of 10 µm or more
🔑 Key Takeaways
- 🤖 CNN demonstrated the ability to predict treatment response in DME patients.
- 📈 AUC of 0.81 indicates a strong predictive capability.
- 🔍 Recall of 78.9% for identifying responders to anti-VEGF treatment.
- ⚠️ Specificity of 68.8% for identifying non-responders.
- 💡 Study highlights the potential of AI in enhancing treatment decision-making.
- 🏥 Conducted at the Storm Eye Institute, Medical University of South Carolina.
- 📅 Study timeframe: Retrospective analysis with a 1-month follow-up.
📚 Background
Diabetic macular edema (DME) is a common complication of diabetes that can lead to significant vision loss. Traditional methods of assessing treatment response often rely on subjective evaluations and can be inconsistent. The integration of advanced technologies, such as machine learning and convolutional neural networks, offers a promising avenue for improving predictive accuracy and personalizing treatment strategies.
🗒️ Study
This pilot study was conducted at the Storm Eye Institute, focusing on patients with a new diagnosis of DME who had not previously received treatment. The researchers reviewed charts and analyzed OCT scans taken at baseline and one month after the first anti-VEGF injection. The aim was to train a CNN to predict treatment responses based solely on the baseline OCT data.
📈 Results
The analysis revealed that out of the 73 eyes studied, 57 eyes were classified as responders to the anti-VEGF injection, while 16 eyes were non-responders. The CNN’s performance was quantified with an AUC of 0.81, showcasing its ability to accurately predict treatment outcomes based on initial OCT scans.
🌍 Impact and Implications
The findings from this study suggest that CNNs could play a transformative role in ophthalmology by providing clinicians with reliable tools for predicting treatment responses in DME patients. This could lead to more tailored treatment plans, ultimately improving patient outcomes and optimizing resource allocation in healthcare settings.
🔮 Conclusion
This pilot study underscores the potential of convolutional neural networks in predicting treatment responses for diabetic macular edema. As we continue to explore the integration of AI in clinical practice, such technologies may significantly enhance our ability to deliver personalized and effective care. Further research is encouraged to validate these findings and expand their application in ophthalmology and beyond.
💬 Your comments
What are your thoughts on the use of AI in predicting treatment responses for eye diseases? We would love to hear your insights! 💬 Share your comments below or connect with us on social media:
Use of a Convolutional Neural Network to Predict the Response of Diabetic Macular Edema to Intravitreal Anti-VEGF Treatment: A Pilot Study.
Abstract
PURPOSE: To utilize a convolutional neural network (CNN) to predict the response of treatment-naïve diabetic macular edema (DME) to a single injection of anti-vascular endothelial growth factor (anti-VEGF) with data from optical coherence tomography (OCT).
DESIGN: Retrospective study performed via chart review.
METHODS: Setting: This was a single-center study performed at the Storm Eye Institute, Medical University of South Carolina.
PATIENT POPULATION: Patients with a new diagnosis of DME who underwent intravitreal (IVT) anti-VEGF injections were eligible for inclusion, provided they had a baseline OCT scan at the time of diagnosis and a 1-month follow-up OCT scan after the first anti-VEGF injection. Exclusion criteria included prior treatment with anti-VEGF, lack of required OCT scans, coexistent macular degeneration, and macular edema due to other retinal diseases. Seventy-three (73) eyes from 53 patients were included.
INTERVENTION: The OCT scan from the baseline visit was compared to the follow-up OCT scan approximately 1 month after the first anti-VEGF injection to determine change in central subfield thickness (delta CST). The delta CST was fed into the CNN as a label to train the system to predict treatment response from only the baseline OCT scan.
MAIN OUTCOME MEASURE: CNN prediction of treatment response to anti-VEGF. Treatment response was defined as a CST reduction of10 µm or more.
RESULTS: Based on delta CST from two OCT scans, 57 eyes were responders and 16 eyes were non-responders to the initial anti-VEGF injection. Analyzing only the baseline OCT scan for each eye, the trained CNN demonstrated an area under the curve (AUC) of 0.81. At the reported operating point, the CNN correctly identified 45 of the 57 responder eyes (i.e., recall of 78.9%) and 11 of the 16 non-responder eyes (i.e., specificity of 68.8%).
CONCLUSIONS: The results of this study demonstrate the potential of a CNN to predict the response of treatment-naïve DME to a single injection of anti-VEGF therapy.
Author: [‘Magrath G’, ‘Luvisi J’, ‘Russakoff D’, ‘Oakley J’, ‘Say E’, ‘Blice J’, ‘Jayagopal A’, ‘Tucker S’, ‘Loayza A’, ‘Baker GH’, ‘Obeid JS’]
Journal: Am J Ophthalmol
Citation: Magrath G, et al. Use of a Convolutional Neural Network to Predict the Response of Diabetic Macular Edema to Intravitreal Anti-VEGF Treatment: A Pilot Study. Use of a Convolutional Neural Network to Predict the Response of Diabetic Macular Edema to Intravitreal Anti-VEGF Treatment: A Pilot Study. 2025; (unknown volume):(unknown pages). doi: 10.1016/j.ajo.2025.02.017