Macular hole morphology and measurement using an automated three-dimensional image segmentation algorithm

Idiopathic Macular Hole Area to Foveal Avascular Zone Ratio and Its Effects on Visual Acuity Before and After Surgery

Purpose

This study investigates factors influencing visual acuity logarithmic minimum angle of resolution (logMAR) in idiopathic macular holes (IMH), with a focus on the foveal avascular zone (FAZ).

Methods

En face images from optical coherence tomography and optical coherence tomography angiography of 152 patients with stages 2, 3, and 4 IMH were analyzed. The minimum area (MA), base area, and FAZ of the macular hole were quantified, and the ratio of minimum area to FAZ (MFR) was calculated. In addition, central subfield thickness was extracted. Relationships between preoperative and postoperative visual acuity and these parameters, along with age, sex, axial length, and stage, were assessed using univariate and multivariate analyses.

Results

The study included 113 patients with high-quality images (113 eyes; median age, 69 years; interquartile range, 65-73 years). Multivariate analysis of factors significantly associated with pre- and postoperative visual acuity identified MFR as the only consistent independent factor (preoperative: β = 0.280, P < 0.05; postoperative: β = 0.357, P < 0.01).

Conclusions

The ratio of macular hole area to the FAZ may be a potentially important morphofunctional parameter influencing visual acuity outcomes in patients with IMH. These findings suggest that MFR could be useful in assessing surgical prognosis, although further research with larger, diverse cohorts is needed.

Translational Relevance

This study bridges the gap between basic retinal morphology and clinical outcomes by identifying MFR as a predictor of visual acuity in patients with IMH. Incorporating MFR into preoperative evaluations could improve surgical prognostication.

Factors influencing the reliability of measurements in eyes with full-thickness macular holes: are we measuring incorrectly?

PURPOSE

The calliper function is used for manual measurements of full thickness macular holes (FTMHs). We aimed to investigate whether a reproducible difference could be detected beyond interobserver variability between two commonly used manufacturers in their manual calliper facility in spectral domain optical coherence tomography (OCT) for metrics related to FTMH.

METHODS

This is a non-interventional, retrospective, observational study. Two independent observers examined 8 eyes (16 OCT) scans and 128 measurements (minimal linear diameter (MLD), basal diameter and hole height on both sides) of FTMHs, taken on Heidelberg Spectralis and Topcon Triton (OCT machines). The interobserver agreement and OCT machine agreement of measurements were analysed by Bland-Altman plots and intraclass correlation coefficient (ICC) analysis. Spectralis and Triton had 125 µm and 50 µm horizontal b-scan spacing, respectively.

RESULTS

Overall, we report high absolute agreement in interobserver (ICC 0.991 (95% CI 0.985 to 0.995, p<0.001)) and OCT machine (ICC 0.993 (95% CI 0.987 to 0.996, p<0.001)) variability. Lower horizontal resolution in Triton compared with Spectralis leads to interobserver variability, in smaller horizontal measurements. Lower horizontal scanning density in Spectralis lead to relatively large interobserver variation if different reference scans were chosen, and consistently smaller MLD measurements than Triton. Vertical measurements without 1:1 scaling lead to inaccurate exaggerated oblique vertical measurements. Calliper function appears otherwise identically calibrated.

CONCLUSIONS

We report excellent interobserver and OCT machine agreement in measurements. However, the paper shows several factors that could influence the reliability of measurements acquired in eyes with FTMHs, such as the dimension of the hole as well as different image resolution, density scanning protocols or vertical scaling of the OCT machines viewing platform.

Artificial Intelligence, Digital Imaging, and Robotics Technologies for Surgical Vitreoretinal Diseases

OBJECTIVE

To review recent technological advancement in imaging, surgical visualization, robotics technology, and the use of artificial intelligence in surgical vitreoretinal (VR) diseases.

BACKGROUND

Technological advancements in imaging enhance both preoperative and intraoperative management of surgical VR diseases. Widefield imaging in fundal photography and OCT can improve assessment of peripheral retinal disorders such as retinal detachments, degeneration, and tumors. OCT angiography provides a rapid and noninvasive imaging of the retinal and choroidal vasculature. Surgical visualization has also improved with intraoperative OCT providing a detailed real-time assessment of retinal layers to guide surgical decisions. Heads-up display and head-mounted display utilize 3-dimensional technology to provide surgeons with enhanced visual guidance and improved ergonomics during surgery. Intraocular robotics technology allows for greater surgical precision and is shown to be useful in retinal vein cannulation and subretinal drug delivery. In addition, deep learning techniques leverage on diverse data including widefield retinal photography and OCT for better predictive accuracy in classification, segmentation, and prognostication of many surgical VR diseases.

CONCLUSION

This review article summarized the latest updates in these areas and highlights the importance of continuous innovation and improvement in technology within the field. These advancements have the potential to reshape management of surgical VR diseases in the very near future and to ultimately improve patient care.

FINANCIAL DISCLOSURE(S)

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Association of macular hole intraretinal fluid and visual acuity

OBJECTIVE

This study investigates the association between preoperative intraretinal fluid (IRF) area and preoperative and postoperative best-corrected visual acuity (BCVA) in surgically repaired idiopathic macular holes (MH). This study further evaluates other prognostic indices related to MH repair that may assist clinicians' understanding of MH operative management.

DESIGN

Retrospective cohort study conducted at a single institution.

PARTICIPANTS

A total of 251 patients who underwent surgery for idiopathic MH between January 2012 and January 2021.

METHODS

Segmentation was performed on ocular coherence tomography scans of 251 eyes with MH and IRF. Associations between IRF area and preoperative and postoperative BCVA at 1, 3, and 6 months, preoperative and postoperative central subfield thickness, MH diameter, staging, closure status, and type of closure were evaluated using Spearman's correlation analysis.

RESULTS

Preoperative IRF area was moderately correlated with preoperative BCVA (r = -0.32; p < 0.001) and negligibly correlated with postoperative BCVA at 1, 3, and 6 months (r = -0.14, p = 0.026; r = -0.21, p < 0.001; and r = -0.19, p < 0.001, respectively). Preoperative IRF area was strongly correlated with MH minimum linear diameter (r = 0.56; p < 0.001) and MH base diameter (r = 0.65; p < 0.001). Other associations were not statistically significant.

CONCLUSION

Preoperative IRF area in patients with idiopathic MH demonstrated a moderate correlation with preoperative BCVA and a negligible or weak correlation with postoperative BCVA at up to 6 months, suggesting that vision may not have a clinically significant relationship with IRF in the setting of MH.

PREDICTION OF MACULAR HOLE SIZE PROGRESSION BASED ON BASELINE OPTICAL COHERENCE TOMOGRAPHY FINDINGS

PURPOSE

To quantify the rate of idiopathic macular hole progression from presentation and identify factors that may influence stratification and urgency for surgical listing based on the initial optical coherence tomography scans.

METHODS

The minimal linear diameter (MLD), base diameter (BD), and hole height on nasal and temporal sides of idiopathic macular hole were measured on spectral domain optical coherence tomographies, on initial presentation and just before surgery. Mean hole height, hole height asymmetry (absolute difference between nasal and temporal height), MLD/BD, and MLD change per day (MLD/day) were calculated for each patient. Multivariable linear regression analysis with MLD/day as the dependent variable was performed to identify significant risk factors for MLD progression. Minimal linear diameter was grouped to quintiles: 1: ≤290 µ m, 2: >290 µ m and ≤385 µ m, 3: >385 µ m and ≤490 µ m, 4: >490 µ m and ≤623 µ m, and 5: >623 µ m.

RESULTS

In 161 eyes (157 patients), we report significant associations with MLD/day: 1) MLD/BD ( P = 0.039) (i.e., wide BD relative to MLD lead to faster progression of MLD), 2) hole height asymmetry ( P = 0.006) (larger absolute difference between nasal and temporal hole height lead to faster progression), and 3) days between scans ( P < 0.001) (longer duration between scans had reduced MLD/day, indicating more rapid increase initially then plateaux), and relative to MLD Quintile 1, MLD Quintile 3 ( P = 0.002) and MLD Quintile 4 ( P = 0,008), and MLD Quintile 5 ( P < 0.001) all lead to a reduced MLD/day rate on multivariable regression.

CONCLUSION

In addition to finding that the previously reported initial smaller MLD is a risk factor for rapid MLD progression, we report two novel findings, large hole height asymmetry and a low MLD/BD (wide base relative to MLD), that represent significant risk factors. These factors should be taken into consideration on presentation to stratify timing of surgery.

Surgical classification for large macular hole: based on different surgical techniques results: the CLOSE study group

BACKGROUND

The CLOSE study group proposes an updated surgical classification for large macular holes based on a systematic review of new treatments. Recently, many new techniques have been introduced to treat large full-thickness macular holes (FTMH); although the indications are not clear. An updated surgical classification is needed to help surgical decision-making.

METHODS

We gathered published series by the CLOSE Study Group members and from literature search until June 2021. Techniques included: internal limiting membrane peeling (ILM peeling), ILM flaps, macular hydrodissection (macular hydro), human amniotic membrane graft (hAM), and autologous retinal transplantation (ART). Within each technique, chi-square test assessed association between the minimal linear diameter (MLD) (in µm) and closure rate; the postoperative best-corrected visual acuity (BCVA) gains were compared among groups.

RESULTS

Data extraction included 31 published articles: total of 1135 eyes. Eyes were divided into the following groups: ILM peel (n: 683), ILM Flap (n: 233), macular hydrodissection (n: 64), hAM (n: 59), and ART (n: 96). The initial BCVA and size were heterogenous between the groups. ILM peel showed the best results in large FTMH ≤ 535 µm (closure rate 96.8%); adjusted mean BCVA: 0.49 (LogMAR) with a statistical difference among groups. Large FTMH between 535 and 799 µm: ILM flap technique showed better results (closure rate 99.0%); adjusted mean BCVA: 0.67(LogMAR); also with a statistical difference. For large FTMH ≥ 800 µm more invasive techniques are required. Use of hAM, macular hydrodissection and ART showed higher closure rates for this category (100%, 83.3% and 90.5% respectively), and adjusted mean BCVA varied from 0.76 to 0.89. Although there was no statistical difference between those techniques for this group due to the smaller number of cases.

CONCLUSIONS

The CLOSE study group demonstrated the potential usefulness of a new surgical classification for large FTMHs and propose OCT biomarkers for use in clinical practice and future research. This new classification demonstrated that Large (400-550 µm) and X-Large (550-800 µm) holes can be treated highly successfully with ILM peel and ILM flap techniques, respectively. Further studies are necessary for the larger FTMHs (XX-Large and Giant), using the CLOSE classification, in order to determine which technique is better suited for each hole size and characteristics.

Automatic Determination of the Center of Macular Hole Using Optical Coherence Tomography En Face Images

To evaluate the automated determination of the center of an idiopathic macular hole (MH) by using swept-source optical coherence tomography (OCT) images with new macro-based algorithms in ImageJ and to compare the difference between the MH center measurements obtained automatically and manually. This cross-sectional study included 39 eyes of 39 elderly individuals (22 women, 17 men) with stage 3 and 4 MH. The MH center was automatically determined using the ImageJ macro. The foveal center was also manually identified by two masked examiners using horizontal and vertical serial B-scan OCT angiography images. The mean age was 68.8 ± 8.3 years. After adjusting for the effect of magnification, the mean distance between the MH center determined manually by Examiner 1 and that determined automatically was 15.5 ± 9.9 µm. The mean distance between the two manually determined measurements of the MH center was 20.3 ± 19.7 µm. These two mean distance values did not differ significantly (Welch t-test, p = 0.27) and was non-inferior (p < 0.0001). The automated ImageJ-based method for determining the MH center was comparable to manual methods. This study showed that automated measurements were non-inferior to manual measurements, and demonstrated a substitutable usefulness, at least for use in clinical practice.

Proof-of-Concept Analysis of a Deep Learning Model to Conduct Automated Segmentation of OCT Images for Macular Hole Volume

BACKGROUND AND OBJECTIVE

To determine whether an automated artificial intelligence (AI) model could assess macular hole (MH) volume on swept-source optical coherence tomography (OCT) images.

PATIENTS AND METHODS

This was a proof-of-concept consecutive case series. Patients with an idiopathic full-thickness MH undergoing pars plana vitrectomy surgery with 1 year of follow-up were considered for inclusion. MHs were manually graded by a vitreoretinal surgeon from preoperative OCT images to delineate MH volume. This information was used to train a fully three-dimensional convolutional neural network for automatic segmentation. The main outcome was the correlation of manual MH volume to automated volume segmentation.

RESULTS

The correlation between manual and automated MH volume was R² = 0.94 (n = 24). Automated MH volume demonstrated a higher correlation to change in visual acuity from preoperative to the postoperative 1-year time point compared with the minimum linear diameter (volume: R² = 0.53; minimum linear diameter: R² = 0.39).

CONCLUSION

MH automated volume segmentation on OCT imaging demonstrated high correlation to manual MH volume measurements. [Ophthalmic Surg Lasers Imaging Retina. 2022;53(4):208-214.].

A Tool for High-Resolution Volumetric Optical Coherence Tomography by Compounding Radial-and Linear Acquired B-Scans Using Registration

Optical coherence tomography (OCT) is a medical imaging modality that is commonly used to diagnose retinal diseases. In recent years, linear and radial scanning patterns have been proposed to acquire three-dimensional OCT data. These patterns show differences in A-scan acquisition density across the generated volumes, and thus differ in their suitability for the diagnosis of retinal diseases. While radial OCT volumes exhibit a higher A-scan sampling rate around the scan center, linear scans contain more information in the peripheral scan areas. In this paper, we propose a method to combine a linearly and radially acquired OCT volume to generate a single compound volume, which merges the advantages of both scanning patterns to increase the information that can be gained from the three-dimensional OCT data. We initially generate 3D point clouds of the linearly and radially acquired OCT volumes and use an Iterative Closest Point (ICP) variant to register both volumes. After registration, the compound volume is created by selectively exploiting linear and radial scanning data, depending on the A-scan density of the individual scans. Fusing regions from both volumes with respect to their local A-scan sampling density, we achieve improved overall anatomical OCT information in a high-resolution compound volume. We demonstrate our method on linear and radial OCT volumes for the visualization and analysis of macular holes and the surrounding anatomical structures.

Benchmarking automated detection of the retinal external limiting membrane in a 3D spectral domain optical coherence tomography image dataset of full thickness macular holes

In this article, we present a new benchmark for the segmentation of the retinal external limiting membrane (ELM) using an image dataset of spectral domain optical coherence tomography (OCT) scans in a patient population with idiopathic full-thickness macular holes. Specifically, the dataset used contains OCT images from one eye of 107 patients with an idiopathic full-thickness macular hole. In total, the dataset contains 5243 individual 2-dimensional (2-D) OCT image slices, with each patient contributing 49 individual spectral-domain OCT tagged image slices. We display precise image-wise binary annotations to segment the ELM line. The OCT images present high variations in image contrast, motion, brightness, and speckle noise which can affect the robustness of applied algorithms, so we performed an extensive OCT imaging and annotation data quality analysis. Imaging data quality control included noise, blurriness and contrast scores, motion estimation, darkness and average pixel scores, and anomaly detection. Annotation quality was measured using gradient mapping of ELM line annotation confidence, and idiopathic full-thickness macular hole detection. Finally, we compared qualitative and quantitative results with seven state-of-the-art machine learning-based segmentation methods to identify the ELM line with an automated system.

Internal limiting membrane peeling for large macular hole: Tailoring the rhexis to the shape of the hole

Purpose:

To report a simple modification of internal limiting membrane (ILM) peeling tailored to the shape of the macular hole to improve the closure rates.

Methods:

This is a single-center interventional case series. conducted between 2016 and 2020. The minimum follow-up was 4 months. All surgeries were performed by one surgeon. Twenty consecutive patients (21 eyes) with large idiopathic macular holes (horizontal diameter: ≥600 μm) were enrolled; vertical hole diameters were also measured using spectral-domain optical coherence tomography (OCT). Following vitrectomy, ILM peeling was performed over a horizontally oval area (additional 1 disc-diameter temporally); perfluoropropane gas (C3F8, 15%) tamponade was used. Hole closure and change in best-corrected visual acuity (BCVA) were monitored after absorption of the gas. Preoperative and postoperative visual acuities were compared using paired t-test. IBM SPSS (ver. 26) was used for analysis.

Results:

The macular holes were horizontally oval rather than circular without exception: mean horizontal and vertical diameters were 714 μm (range: 600–1020 μm) and 602 μm (490–844 μm), respectively. Following vitrectomy, macular hole closure was obtained in 20/21 eyes by the last follow-up (mean: 28 months, median: 34 months; range 4–48 months). Mean Snellen BCVA improved from 20/200 to 20/63 (P < 0.0001).

Conclusion:

All the macular holes in the study were observed to be horizontally oval. A corresponding horizontal enlargement of the ILM rhexis yielded excellent anatomical and satisfactory visual outcomes.

Defining a Cutoff for Progression of Macular Holes

Purpose

The purpose of this study was to determine a cutoff for progression of idiopathic full-thickness macular hole (MH) size.

Methods

Retrospective analysis of consecutive patients waiting 4 weeks for MH surgery. Two observers performed 3 repeat sets of MH size measurements on optical coherence tomography (OCT) high-density radial scans taken at first presentation and 4 weeks later before surgery. Primary outcome was the definition of a cutoff for true enlargement of MH size versus measurement error. Secondary outcomes were risk factors for change in minimum linear diameter (MLD) size and best-corrected visual acuity (BCVA).

Results

Fifty-one patients were included with a mean MH size of 334 µm (±179 µm; range 39 to 793 µm). The cutoff for an increase in MLD size calculated as the outer confidence limit for the 99.73% limits of agreement was 31 µm. This was independent of MH size. Using this cutoff, MLD size increased in 9/34 (26.5%) of patients without and in 14 of 17 (82.4%) of patients with vitreomacular traction (VMT; P < 0.001). Mean BCVA deteriorated in patients in whom the MH had progressed from 0.62 (±0.23) logMAR to 0.82 (±0.29; P < 0.001), whereas there was no significant change in BCVA in patients without MH progression (P = 0.25). In 31% (16/51) of patients, classification of their MHs (small ≤250 µm, medium 251–400 µm, and large >400 µm) changed over the 4-week period.

Conclusions

Using a cutoff discriminates change from measurement error. A significant proportion of MHs progressed by 4 weeks, particularly in the presence of VMT.

Translational Relevance

The established cutoff enables clinicians to differentiate true MH enlargement from measurement error.

OUTCOME OF REVISION PROCEDURES FOR FAILED PRIMARY MACULAR HOLE SURGERY

PURPOSE

To analyze the outcomes of revision surgery for idiopathic full-thickness macular holes that have failed to close after primary surgery, and also to assess factors predicting success and to review the relative effect of adjunctive surgical techniques.

METHODS

A multicenter retrospective study. Anatomical closure rates and visual acuity change between pre and postrevision surgery were assessed. Hole size, age, symptom duration, surgical interval, and reduced hole size were analyzed as predictive factors for success. Effectiveness of adjunctive surgical techniques was reviewed.

RESULTS

Seventy-seven eyes were included in the study. Anatomical closure was achieved in 71% (55/77) cases. There was a median gain of 11 Early Treatment of Diabetic Retinopathy Score letters in all holes and 14 letters in closed holes. Full-thickness macular holes that increased in size by more than 10% following primary surgery had a closure rate of 50% compared with 80% in holes that reduced by 10% or stayed the same (P = 0.015). Increasing hole size is associated with a modest reduction in odds of closure (odds ratio = 0.99; P = 0.04). Surgical interval <2 months is not associated with better outcomes compared with >2 months (P = 0.14).

CONCLUSION

Revision surgery for full-thickness macular holes that have failed to close after primary surgery is associated with high closure rates and significant visual gains.

Repeatability and Reproducibility of Macular Hole Size Measurements Using Optical Coherence Tomography

The purpose of this study was to assess the repeatability and reproducibility of measuring the minimum linear diameter (MLD) of macular holes (MHs) using horizontal linear and radial scan modes in optical coherence tomography (OCT). Patients with concurrent sets of radial and horizontal linear OCT volume scans were included. The MLD was measured twice in both scan modes by six raters of three different experience levels (groups). Outcome measures were the reliability and repeatability of MLD measurements. Fifty patients were included. Mean MLD was 317.21(±170.63) µm in horizontal linear and 364.52 (±161.71) µm in radial mode, a difference of 47.31 (±26.48) µm (p < 0.001). In the radial scan mode, MLD was identified within 15° of the horizontal meridian in 27% and within 15° of the vertical meridian in 26.7%, with the remainder (46.3%) in oblique meridians. The intra-group coefficients of repeatability (CR) for horizontal linear mode were 23 µm, 33 µm and 45 µm, and for radial mode 25 µm, 44 µm and 57 µm for groups 1, 2 and 3, respectively. The inter-group CR, taking group 1 as reference standard for groups 2 and 3, were 74 µm and 71 µm for the linear mode, and 62 µm and 78 µm for radial mode. The radial mode provides good repeatability and reliability for measurement of MLD. In a majority of cases the MLD does not lie in the horizontal meridian and would be underestimated using a horizontal OCT mode.