Measuring extraocular muscle volume using dynamic contours

Semantic Segmentation of Extraocular Muscles on Computed Tomography Images Using Convolutional Neural Networks

Computed tomography (CT) imaging of the orbit with measurement of extraocular muscle size can be useful for diagnosing and monitoring conditions that affect extraocular muscles. However, the manual measurement of extraocular muscle size can be time-consuming and tedious. The purpose of this study is to evaluate the effectiveness of deep learning algorithms in segmenting extraocular muscles and measuring muscle sizes from CT images. Consecutive CT scans of orbits from 210 patients between 1 January 2010 and 31 December 2019 were used. Extraocular muscles were manually annotated in the studies, which were then used to train the deep learning algorithms. The proposed U-net algorithm can segment extraocular muscles on coronal slices of 32 test samples with an average dice score of 0.92. The thickness and area measurements from predicted segmentations had a mean absolute error (MAE) of 0.35 mm and 3.87 mm², respectively, with a corresponding mean absolute percentage error (MAPE) of 7 and 9%, respectively. On qualitative analysis of 32 test samples, 30 predicted segmentations from the U-net algorithm were accepted while 2 were rejected. Based on the results from quantitative and qualitative evaluation, this study demonstrates that CNN-based deep learning algorithms are effective at segmenting extraocular muscles and measuring muscles sizes.

Magnetic resonance imaging of the orbit, Part 2: Characterization of orbital pathologies

In this article we focus on a systematic approach to assess common orbital lesions on magnetic resonance imaging (MRI). The identification of the probable compartment or structure of origin helps narrow the differential diagnosis of a lesion. Analyzing the morphology, appearance, and signal intensity on various sequences, the pattern, and degree of contrast enhancement are key to characterize lesions on MRI. Imaging features suggesting cellularity and vascularity can also be determined to help plan for biopsy or surgery of these lesions. MRI can also distinguish active from chronic disease in certain pathologies and aids in selecting appropriate medical management. MRI may thus serve as a diagnostic tool and help in guiding therapeutic strategies and posttreatment follow-up.

Age-related difference in extraocular muscles and its relation to clinical manifestations in an ethnically homogenous group of patients with Graves' orbitopathy

PURPOSE

To evaluate the age-related difference in EOMs and its relation to clinical manifestations by computed tomography (CT) measurement of EOMs.

METHODS

The medical records and CT image review of 40 patients (80 orbits) with moderate-to-severe Graves' orbitopathy were performed. The patients were divided into two age groups, group 1 (≤ 40 years) and group 2 (> 40 years). CT scans of 30 gender- and age-matched normal controls were also obtained. The maximal cross-sectional area (MCA) and its position (pMCA) of each EOM were measured.

RESULTS

Group 1 presented with more severe proptosis (p < 0.001), while group 2 had a higher risk of diplopia (p < 0.001). Motility restriction in supraduction was more likely to occur in Group 2 (p = 0.027) with even higher severity (p = 0.047). The pMCA was higher in the inferior (p = 0.001), medial (p = 0.021), and lateral rectus (p = 0.013) in group 1. Proptosis was positively correlated to pMCA while diplopia was correlated to MCA in both groups. Significant correlation was noted between restrictions levels and MCA (superior, r = 0.467, p < 0.001; inferior, r = 0.358, p = 0.007; medial, r = 0.314, p = 0.018; lateral, r = 0.308, p = 0.021) or pMCA (inferior, r =  - 0.534, p < 0.001) only in group 2.

CONCLUSIONS

The muscle enlargement patterns are significantly different between younger and older patients. Older patients tended to have enlarged muscle bellies more posterior in the orbit, which is responsible for more diplopia and motility restriction. Proptosis is more likely to be affected by the most enlarged position than muscle size. So younger patients tended to develop more proptosis and be less bothered by motility restriction even with enlarged muscles.

Semi-automatic magnetic resonance imaging based orbital fat volumetry: reliability and correlation with computed tomography

Post-processing analysis can provide valuable information for diagnosis and planning of orbital disorders. This cross-sectional study aims to evaluate the reliability of semi-automatic, orbital fat volumetry using magnetic resonance imaging (MRI). Two observers assessed the orbital fat volume using a standard MRI protocol (3T, T1w sequence) in 12 orbits diagnosed with Graves' orbitopathy (GO) and 10 healthy control orbits. MRI and computed tomography (CT) based analysis were compared. Intra-observer variability was good (intraclass correlation coefficient (ICC) 0.88; 95% confidence interval (CI) [0.70, 0.95]) and interobserver agreement was moderate (ICC 0.55; 95% CI [-0.09, 0.81]), which corresponds to a mean percentage difference of 1.3% and 17.9% of the total orbital fat volume. Mean differences between MRI and CT measurements were, respectively, 1.1 cm³ (P= 0.064, 95% CI [-0.20, 2.43]) and 1.4 cm³ (P=0.016, 95% CI [0.21, 2.56]) for the control and the GO group. MRI volumetry was strongly correlated with CT (Pearson's r= 0.7, P<0.001). We conclude that orbital fat volumetry is feasible with a semi-automatic segmentation procedure and standard MRI protocol. Correlation with CT volumetry is good, but considerable bias may derive from observer variability and these errors should be taken into account for the purpose of volumetric analysis. Better definition of error sources may increase measurement accuracy.

Extraocular muscle sampled volume in Graves' orbitopathy using 3-T fast spin-echo MRI with iterative decomposition of water and fat sequences

Background

Current magnetic resonance imaging (MRI) techniques for measuring extraocular muscle (EOM) volume enlargement are not ideally suited for routine follow-up of Graves’ ophthalmopathy (GO) because the difficulty of segmenting the muscles at the tendon insertion complicates and lengthens the study protocol.

Purpose

To measure the EOM sampled volume (SV) and assess its correlation with proptosis.

Material and Methods

A total of 37 patients with newly diagnosed GO underwent 3-T MRI scanning with iterative decomposition of water and fat (IDEAL) sequences with and without contrast enhancement. In each patient, the three largest contiguous coronal cross-sectional areas (CSA) on the EOM slices were segmented using a polygon selection tool and then summed to compute the EOM-SV. Proptosis was evaluated with the Hertel index (HI). The relationships between the HI value and EOM-SV and between HI and EOM-CSA were compared and assessed with Pearson’s correlation coefficient and the univariate regression coefficient. Inter-observer and intra-observer variability were calculated.

Results

HI showed a stronger correlation with EOM-SV (P < 0.001; r = 0.712, r²= 0.507) than with EOM-CSA (P < 0.001; r = 0.645 and r²= 0.329). The intraclass correlation coefficient indicated that the inter-observer agreement was high (0.998). The standard deviation between repeated measurements was 1.9–5.3%.

Conclusion

IDEAL sequences allow for the measurement EOM-SV both on non-contrast and contrast-enhanced scans. EOM-SV predicts proptosis more accurately than does EOM-CSA. The measurement of EOM-SV is practical and reproducible. EOM-SV changes of 3.5–8.3% can be assumed to reflect true volume changes.

Correlation between Extraocular Muscle Size Measured by Computed Tomography and the Vertical Angle of Deviation in Thyroid Eye Disease

The aim of this study was to investigate extraocular muscle (EOM) volume and cross-sectional area using computed tomography (CT), and to determine the relationship between EOM size and the vertical angle of deviation in thyroid eye disease (TED). Twenty-nine TED patients (58 orbits) with vertical strabismus were enrolled in the study. All patients underwent complete ophthalmic examination including prism, alternate cover, and Krimsky tests. Orbital CT scans were also performed on each patient. Digital image analysis was used to quantify superior rectus (SR) and inferior rectus (IR) muscle cross-sectional areas and volumes. Measurements were compared with those of controls. The correlation between muscle size and degree of vertical angle deviation was evaluated. The mean vertical angle of deviation was 26.2 ± 4.1 prism diopters. The TED group had a greater maximum cross-sectional area and EOM volume in the SR and IR than the control group (all p<0.001). Area and volume of the IR were correlated with the angle of deviation, but the SR alone did not show a significant correlation. The maximum cross-sectional area and volume of [Right IR + Left SR − Right SR − Left IR] was strongly correlated with the vertical angle of deviation (P<0.001). Quantitative CT of the orbit with evaluation of the area and volume of EOMs may be helpful in anticipating and monitoring vertical strabismus in TED patients.

Graves' ophthalmopathy: the role of diffusion-weighted imaging in detecting involvement of extraocular muscles in early period of disease

OBJECTIVE

To evaluate involvement of the extraocular muscle (EOM) using diffusion-weighted imaging (DWI), to determine whether there is correlation with conventional orbital MRI and apparent diffusion coefficient (ADC) values in patients with Graves' ophthalmopathy (GO).

METHODS

35 patients known clinically with GO and 21 healthy controls were studied. Patients were assessed with clinical activity scores. All subjects underwent conventional MRI and DWI study. Involvement of the EOM was evaluated. The patients were classified as involved or uninvolved on orbital MRI and their ADC values in DWI compared.

RESULTS

There was significant difference in the mean ADC value of all the EOMs in patients vs controls. The ADC values of all the EOMs were higher in patients. There were significant differences in ADC values between uninvolved muscles on conventional MRI and controls for the MR, SR and LR. There was no significant difference in ADC value between the two groups when considering the IR. ADC values of medial, lateral and superior rectus muscles were increased.

CONCLUSION

Increased ADC values of the EOM in patients with GO suggest that EOM damage begins at a very early stage before being detected on routine orbital MRI. The routine MRI with DWI sequence will be a useful adjunct in the selection of a group of patients most likely to benefit from early treatment.

ADVANCES IN KNOWLEDGE

This study can help to evaluate the involvement of GO in early period with MRI added DWI.

Imaging anatomy and pathology of extraocular muscles in adults

The extraocular muscles (EOM) are involved in a variety of disease processes with characteristic findings on imaging. EOM anatomy is described, followed by a review of adult EOM pathology. The imaging characteristics are explained with examples. The pattern of EOM disease on imaging, in corroboration with clinical findings, can often lead the radiologist towards a specific diagnosis.

Differential involvement of orbital fat and extraocular muscles in graves' ophthalmopathy

Graves' ophthalmopathy (GO) is characterized by swelling of orbital fat and extraocular muscles, but little attention has been given to differential involvement of fat and muscles. Advancements in imaging allow rather accurate measurements of orbital bony cavity volume (OV), fat volume (FV) and muscle volume (MV), and are the topics of this review. Ratios of FV/OV and MV/OV neutralize gender differences. In adult Caucasian controls, mean values ± SD of FV/OV are 0.56 ± 0.11 and of MV/OV are 0.15 ± 0.02. FV increases substantially and MV decreases slightly with advancing age, requiring age-specific reference ranges. In 95 consecutive untreated Caucasian GO patients, both FV and MV were within normal limits in 25%, increased FV but normal MV was present in 5%, normal FV but increased MV was detected in 61%, and both increased FV and MV was evident in 9%. Increased FV was associated with more proptosis and longer GO duration. Increased MV was associated with older age, more severe GO (more proptosis and diplopia, worse eye muscle ductions), higher TBII and current smoking. At the cellular and molecular level differential involvement of fat and muscles might be related to differences between fibroblast phenotypes and cytokine profiles in each compartment, to different orbital T cell subsets during the course of the disease and to peroxisome proliferator activator receptor-γ polymorphisms and modulation of 11β-hydroxysteroid dehydrogenase-1. Enlarged muscles are apparently a rather early phenomenon in GO, whereas increases in fat mass occur relatively late. Why a minor subset of GO patients presents with an increase of only fat remains poorly understood.

A quantitative method for assessing the degree of axial proptosis in relation to orbital tissue involvement in Graves' orbitopathy

PURPOSE

To define a method of quantifying axial proptosis in patients with Graves' orbitopathy (GO) and to validate a score that correlates with the orbital involvement and helps determine the degree of proptosis correction for elective orbital decompression.

DESIGN

Retrospective, case series.

PARTICIPANTS

The study included 50 patients (group A) and 29 control subjects who underwent orbital computed tomography (CT). The method was then validated in another group of 21 patients with GO (group B).

METHODS

The orbital area (OA) was measured manually on the central axial section of the CT scan at a level where the lens is visualized. The OA intersects the projection of the globe and delimitates the chord of an arch (globe chord [OC]). The area of the circular sector under the chord (CA) represents the portion of the globe within the orbit.

MAIN OUTCOME MEASURES

A CA-to-OA ratio was calculated to reduce the error due to variability of the measurements and to perform correlations with some of the clinical parameters of GO.

RESULTS

Measurement error was low (<2%). We did not observe significant differences in the mean OA of patients with GO (783.6 ± 12.1 mm(2)) and controls (758.5 ± 20.4 mm(2); P = not significant). The OC value in patients with GO was 130.2 ± 11.5 mm(2), significantly lower than in controls (281.8 ± 9.7 mm(2); P<0.0001). The CA-to-OA ratio also was lower in patients with GO than in controls (0.16 ± 0.01 vs. 0.38 ± 0.01; P<0.0001). A significant correlation was found in patients with GO between the CA-to-OA ratio and proptosis (P<0.001), lid fissure (P = 0.004), and intraocular pressure (P<0.001). In group B, the CA-to-OA ratio was 0.18 ± 0.02, significantly different from that of controls (P<0.0001) and inversely correlated with proptosis (P<0.0001) and lid fissure (P<0.045).

CONCLUSIONS

By measuring the CA-to-OA ratio, we were able to quantify the degree of axial proptosis in patients with GO. The significant correlation of CA/OA with some orbital parameters confirms that this parameter also may be used as a measure of orbital involvement in GO.

FINANCIAL DISCLOSURE(S)

The author(s) have no proprietary or commercial interest in any materials discussed in this article.

Eyeball muscles' diameters versus volume estimated by numerical image segmentation

PURPOSE

To determine the clinical usefulness of the numerical segmentation image technique (NSI) in estimating the volume of extraocular muscles and to compare this value to widely used measurements of single diameters of the muscles.

METHODS

Forty-five patients underwent magnetic resonance examinations in 1.5-T scanner. SE T1 sequences in transversal and coronal planes were provided and data were sent to a personal computer, where the degree of exophthalmos, horizontal diameter of medial rectus muscles, and vertical diameter of inferior rectus muscles were determined on the basis of two-dimensional images. The quantity estimation of all eye muscles volumes using NSI application in three-dimensional space was carried out with use of level set segmentation algorithm.

RESULTS

A strong correlation between the total eye muscle volume and degree of exophthalmos was determined. The usefulness of measuring single diameters for estimating the muscles' enlargement was confirmed. The difference between a single muscle's volume and its width also was confirmed. Estimates of muscle volume correlate with the degree of exophthalmos more accurately than measurements of single diameters.

CONCLUSIONS

The NSI technique is a clinically useful application, providing objective data calculated individually for each orbit. It allows an objective estimation of the pathologic processes leading to exophthalmos and may be especially helpful in monitoring discrete changes in the muscles volume during treatment.

Magnetic resonance imaging and ultrasound measurements of extraocular muscles in thyroid-associated ophthalmopathy at different stages of the disease

PURPOSE

To assess extraocular muscle (EOM) involvement in thyroid-associated orbitopathy (TAO) of different stages with magnetic resonance imaging (MRI) and ultrasound techniques.

METHODS

A total of 32 patients with TAO were divided into three groups according to whether they had mild active, pronounced active or longstanding inactive disease. Six patients with Graves' disease but no clinical signs of TAO and 10 healthy control subjects were also studied. Muscle volume and cross-sectional area were measured with MRI. A-scan ultrasound was used to measure muscle thickness.

RESULTS

The average MRI volume and maximal cross-sectional area of the EOM were significantly larger in patients with pronounced active and longstanding inactive TAO than in control subjects. Increased average muscle thickness measured by ultrasound was found mainly in patients with longstanding disease. Muscle enlargement was seen with MRI and ultrasound in individual patients in all patient groups, including those with Graves' disease but no TAO. Bilateral muscle enlargement was revealed by MRI in about two-thirds of patients with mild active TAO and in all patients with pronounced active and longstanding inactive TAO. Bilateral involvement estimated with ultrasound was less common in all patient groups. The MRI and ultrasound findings were not well correlated in any patient group.

CONCLUSIONS

Extraocular muscle enlargement was seen in all patient groups with TAO of differing levels of severity. Measurements with MRI of muscle volume or maximal cross-sectional area are considered good indicators of muscle enlargement in TAO.

Extraocular muscle quantification using mathematical morphology: A semi-automatic method for analyzing muscle enlargement in orbital diseases

Mathematical morphology was used to quantify the dimensions of the human extraocular muscles in computed tomography orbital scans. Coronal images of 28 patients with Graves orbitopathy and 5 controls (218 orbital scans) were analyzed with an algorithm for extraocular muscle segmentation. The results showed that measurements of extraocular muscle area obtained with semi-automatic segmentation are highly correlated with manual tracing and provides a simple method to quantify orbital structures including the extraocular muscles.

Radiologic measurement of extraocular muscle volumes in patients with Graves' orbitopathy: a review and guideline

OBJECTIVE

To evaluate and compare techniques for extraocular muscle (EOM) volume measurement and to provide guidelines for future measurements.

DESIGN

Systematic review.

RESULTS

Existing techniques used to measure extraocular muscle volumes on radiologic scans can be divided into manual outlining, computer assisted and automated segmentation. Both computed tomography (CT) and magnetic resonance (MR) image datasets can be used. On CT scans, one best measures muscle volume using region grow segmentation, accepting an overestimation of true volume by inevitable inclusion of non-muscular tissue. On high resolution MRI scans, single muscles can be outlined manually, but measurements include only part of the muscle due to poor tissue contrast at the orbital apex. Measurement errors can be reduced 3.5% by exact horizontal repositioning. A measured volume change of at least 6-17% is required to demonstrate a significant difference.

CONCLUSION

Currently the best choice for EOM volume measurements on CT images is computer assisted grey value segmentation and on MRI images is manual outlining of individual muscles. Because of the time required and the complexity of the measurements, present EOM volume measurement is as yet only suitable for research purposes.

Volume changes in intra- and extraorbital compartments in patients with Graves' ophthalmopathy: effect of smoking

The objective of this study was to demonstrate the effect of smoking history on soft tissue expansion in specific orbital compartments in patients with Graves' ophthalmopathy. The volumes of the rectus muscles, intra and extraorbital connective, and soft tissues were measured in 110 orbits of 35 patients and 20 control subjects. Data sets from current smokers, ex-smokers, and non-smokers were compared. The total number of cigarettes smoked was calculated, and it was used as an estimate for the severity of smoking (cumulative smoking). The volume measurements were performed on T1-weighted contiguous transversal magnetic resonance images of the orbits. Connective tissue volumes were influenced by smoking history, while muscle volumes were not affected. Ex-smokers had larger amount of extraorbital connective tissue than current smokers (p = 0.012), and this volume showed a good correlation with the number of cigarettes smoked (r = 0.539, p < 0.05). In current smokers, the amount of intraorbital connective tissue correlated well with the cumulative smoking (r = 0.635, p < 0.001). We conclude that connective tissue volumes in certain orbital compartments correlate well with cumulative smoking. Extraocular muscle volumes are not influenced by smoking in patients with Graves' ophthalmopathy.

Using morphologic parameters of extraocular muscles for diagnosis and follow-up of Graves' ophthalmopathy: diameters, areas, or volumes?

OBJECTIVE

The objective of this study was to find the most appropriate, accurate, and convenient muscle parameter that can be used as a substitute for volume in monitoring the effectiveness of therapy of patients with Graves' ophthalmopathy.

SUBJECTS AND METHODS

The four rectus muscles in 110 orbits (35 patients and 20 control subjects) were evaluated with MR imaging. The diameter at the largest extent of the muscle belly, as well as the long and the short diameters and the cross-sectional areas in a preselected coronal scan, were measured for each muscle and were compared with the corresponding muscle volume measured on contiguous T1-weighted transverse slices.

RESULTS

The measured coronal area correlated well with the volume of the superior (r = 0.694, p < 0.0001) and inferior (r = 0.783, p < 0.0001) recti, and the largest transverse diameter showed strong correlation with the volume of the lateral (r = 0.868, p < 0.0001) and medial (r = 0.869, p < 0.0001) recti. For the latter muscle, the coronal area also exhibited a good correlation with the volume (r = 0.838, p < 0.0001). Coronal cross-sectional areas can be well estimated by measuring both the short and long coronal muscle diameters (r values were between 0.914 and 0.966; p < 0.0001).

CONCLUSION

In Graves' ophthalmopathy, the volume of three of the rectus muscles can be well estimated by simple measurements on a single coronal slice, and the largest transverse diameter of the lateral rectus is suitable for the same purpose.