Intricate eye layers with data grid symbolizing image analysis in DME diagnosis

Decoding DME: How Image Analysis Impacts Your Eye Health

Jordan Keane in Health & Wellness August 2025 • 5 min read.

"Discover how different algorithms affect the accuracy of diabetic macular edema (DME) diagnosis and treatment, empowering you to make informed decisions about your eye care."

Optical coherence tomography (OCT) has transformed the way doctors diagnose and monitor diabetic macular edema (DME), a complication of diabetes that can lead to vision loss. Initially, time-domain OCT was the standard, but spectral-domain OCT (SD-OCT) technology has since revolutionized the process. SD-OCT offers faster scans and higher resolution images of the retina, making it easier to detect subtle changes associated with DME.

One of the most commonly used SD-OCT devices is the Heidelberg Spectralis. It boasts a scan rate of 40,000 axial scans per second and an axial resolution of 5 µm, providing detailed views of the retinal layers. However, the repeatability of OCT measurements is essential for accurate monitoring and treatment decisions.

This raises a critical question: how much do the image analysis algorithms used by these devices impact the accuracy and consistency of DME assessments? Different algorithms may interpret the same scan data differently, potentially affecting the diagnosis and treatment plan. This article explores the effect of image analysis algorithms from SD-OCT. Understanding this impact can empower patients to have informed discussions with their eye care professionals.

The Algorithm Advantage: Comparing Iowa Reference and Heidelberg Spectralis

Intricate eye layers with data grid symbolizing image analysis in DME diagnosis

Researchers at the University of Iowa investigated the intersession repeatability of retinal thickness measurements in patients with DME. They compared the performance of the Heidelberg Spectralis OCT algorithm with a publicly available, three-dimensional graph search-based multilayer OCT segmentation algorithm known as the Iowa Reference Algorithm.

The study involved 30 eyes from 21 patients diagnosed with clinically significant DME. Each patient underwent consecutive, registered macula-centered SD-OCT scans using the Heidelberg Spectralis device. The OCT scans were then analyzed using both the Iowa Reference Algorithm and the manufacturer-supplied Spectralis software.

The Iowa Reference Algorithm segmented the OCT scans into separate surfaces and calculated the average thickness between the internal limiting membrane and the outer retinal pigment epithelium complex surfaces.
The researchers then analyzed the variability between paired scans and compared the results with the retinal thickness measurements obtained from the Spectralis software.

The results showed that the choice of image analysis algorithm significantly affected the reproducibility of central macular thickness measurements. The coefficient of repeatability (variation) for central macular thickness using the Iowa Reference Algorithm was 5.26 µm (0.62% [95% confidence interval (CI), 0.43-0.71]), while for the Spectralis algorithm, it was 6.84 µm (0.81% [95% CI, 0.55-0.92]). When analyzing the central 3 mm, the coefficient of repeatability (variation) was 2.46 µm (0.31% [95% CI, 0.23-0.38]) for the Iowa Reference Algorithm and 4.23 µm (0.53% [95% CI, 0.39-0.65]) for the Spectralis software.

Empowering Your Eye Health: What This Means for You

The study findings suggest that while both the Iowa Reference Algorithm and the Spectralis software offer excellent reproducibility between serial scans in patients with DME, the Iowa Reference Algorithm may have lower between-measurement variation for the central 3 mm subfield. This has significant implications for the management of patients with DME.

The researchers emphasize that the Iowa Reference Algorithm utilizes all three-dimensional information when identifying and segmenting retinal layers, potentially contributing to its increased robustness. In contrast, current segmentation algorithms may lose some of this information.

It's important to discuss the image analysis methods used by your eye care professional. The choice of algorithm can impact the consistency and reliability of your DME assessments, ultimately affecting treatment decisions and visual outcomes. If you have concerns about the accuracy of your OCT scans, ask your doctor about the possibility of using alternative image analysis algorithms or seeking a second opinion.

About this Article -

This article was crafted using a human-AI hybrid and collaborative approach. AI assisted our team with initial drafting, research insights, identifying key questions, and image generation. Our human editors guided topic selection, defined the angle, structured the content, ensured factual accuracy and relevance, refined the tone, and conducted thorough editing to deliver helpful, high-quality information.See our About page for more information.

This article is based on research published under:

DOI-LINK: 10.1167/iovs.12-10420, Alternate LINK

Title: Reproducibility Of Diabetic Macular Edema Estimates From Sd-Oct Is Affected By The Choice Of Image Analysis Algorithm

Subject: General Medicine

Journal: Investigative Opthalmology & Visual Science

Publisher: Association for Research in Vision and Ophthalmology (ARVO)

Authors: Elliott H. Sohn, John J. Chen, Kyungmoo Lee, Meindert Niemeijer, Milan Sonka, Michael D. Abràmoff

Published: 2013-06-19

Everything You Need To Know

What is Optical Coherence Tomography (OCT) and how does it relate to Diabetic Macular Edema (DME)?

Optical coherence tomography, or OCT, is a technology that is used to diagnose and monitor diabetic macular edema, or DME. Initially time-domain OCT was the standard, but Spectral-domain OCT, or SD-OCT, technology has since changed the process. SD-OCT offers faster scans and higher resolution images of the retina, making it easier to detect subtle changes associated with DME. One of the most used SD-OCT devices is the Heidelberg Spectralis.

What is the Iowa Reference Algorithm and how does it compare to the Heidelberg Spectralis OCT algorithm in assessing DME?

The Iowa Reference Algorithm is a publicly available, three-dimensional graph search-based multilayer OCT segmentation algorithm. Researchers at the University of Iowa compared the performance of the Heidelberg Spectralis OCT algorithm with the Iowa Reference Algorithm to determine the intersession repeatability of retinal thickness measurements in patients with DME.

What were the key findings of the study comparing the Iowa Reference Algorithm and the Spectralis algorithm in measuring retinal thickness?

The study found that the choice of image analysis algorithm significantly affected the reproducibility of central macular thickness measurements. The coefficient of repeatability (variation) for central macular thickness using the Iowa Reference Algorithm was 5.26 µm, while for the Spectralis algorithm, it was 6.84 µm. For the central 3 mm, the coefficient of repeatability (variation) was 2.46 µm for the Iowa Reference Algorithm and 4.23 µm for the Spectralis software. This suggests that while both algorithms offer reproducibility between serial scans in patients with DME, the Iowa Reference Algorithm may have lower between-measurement variation for the central 3 mm subfield.

What are the implications of variations in measurement of the retinal thickness between different image analysis algorithms for managing patients with Diabetic Macular Edema (DME)?

The variation in measurements between different algorithms has significant implications for managing patients with DME. If measurements of retinal thickness vary depending on the algorithm used, it can affect the assessment of disease progression and response to treatment. This can affect decisions about when to initiate or modify treatment, potentially leading to suboptimal outcomes. Therefore, it is important for eye care professionals to be aware of the potential impact of different image analysis algorithms on DME assessments and to consider this factor when interpreting OCT results and making treatment decisions.

What is the Heidelberg Spectralis, and how does the Iowa Reference Algorithm relate to its accuracy in diagnosing Diabetic Macular Edema (DME)?

The Heidelberg Spectralis is an SD-OCT device with a scan rate of 40,000 axial scans per second and an axial resolution of 5 µm, providing detailed views of the retinal layers. While the Heidelberg Spectralis provides detailed retinal images, the repeatability of its measurements is essential for accurate monitoring and treatment decisions. The Iowa Reference Algorithm can be used to assess the accuracy of the Heidelberg Spectralis. Understanding the impact of these algorithms can empower patients to have informed discussions with their eye care professionals.