Orbital and ocular imaging using 3- and 1.5-T MR imaging systems

Deep Learning on Conventional Magnetic Resonance Imaging Improves the Diagnosis of Multiple Sclerosis Mimics

OBJECTIVES

The aims of this study were to present a deep learning approach for the automated classification of multiple sclerosis and its mimics and compare model performance with that of 2 expert neuroradiologists.

MATERIALS AND METHODS

A total of 268 T2-weighted and T1-weighted brain magnetic resonance imagin scans were retrospectively collected from patients with migraine (n = 56), multiple sclerosis (n = 70), neuromyelitis optica spectrum disorders (n = 91), and central nervous system vasculitis (n = 51). The neural network architecture, trained on 178 scans, was based on a cascade of 4 three-dimensional convolutional layers, followed by a fully dense layer after feature extraction. The ability of the final algorithm to correctly classify the diseases in an independent test set of 90 scans was compared with that of the neuroradiologists.

RESULTS

The interrater agreement was 84.9% (Cohen κ = 0.78, P < 0.001). In the test set, deep learning and expert raters reached the highest diagnostic accuracy in multiple sclerosis (98.8% vs 72.8%, P < 0.001, for rater 1; and 81.8%, P < 0.001, for rater 2) and the lowest in neuromyelitis optica spectrum disorders (88.6% vs 4.4%, P < 0.001, for both raters), whereas they achieved intermediate values for migraine (92.2% vs 53%, P = 0.03, for rater 1; and 64.8%, P = 0.01, for rater 2) and vasculitis (92.1% vs 54.6%, P = 0.3, for rater 1; and 45.5%, P = 0.2, for rater 2). The overall performance of the automated method exceeded that of expert raters, with the worst misdiagnosis when discriminating between neuromyelitis optica spectrum disorders and vasculitis or migraine.

CONCLUSIONS

A neural network performed better than expert raters in terms of accuracy in classifying white matter disorders from magnetic resonance imaging and may help in their diagnostic work-up.

Ophthalmic Magnetic Resonance Imaging: Where Are We (Heading To)?

Magnetic resonance imaging of the eye and orbit (MReye) is a cross-domain research field, combining (bio)physics, (bio)engineering, physiology, data sciences and ophthalmology. A growing number of reports document technical innovations of MReye and promote their application in preclinical research and clinical science. Realizing the progress and promises, this review outlines current trends in MReye. Examples of MReye strategies and their clinical relevance are demonstrated. Frontier applications in ocular oncology, refractive surgery, ocular muscle disorders and orbital inflammation are presented and their implications for explorations into ophthalmic diseases are provided. Substantial progress in anatomically detailed, high-spatial resolution MReye of the eye, orbit and optic nerve is demonstrated. Recent developments in MReye of ocular tumors are explored, and its value for personalized eye models derived from machine learning in the treatment planning of uveal melanoma and evaluation of retinoblastoma is highlighted. The potential of MReye for monitoring drug distribution and for improving treatment management and the assessment of individual responses is discussed. To open a window into the eye and into (patho)physiological processes that in the past have been largely inaccessible, advances in MReye at ultrahigh magnetic field strengths are discussed. A concluding section ventures a glance beyond the horizon and explores future directions of MReye across multiple scales, including in vivo electrolyte mapping of sodium and other nuclei. This review underscores the need for the (bio)medical imaging and ophthalmic communities to expand efforts to find solutions to the remaining unsolved problems and technical obstacles of MReye, with the objective to transfer methodological advancements driven by MR physics into genuine clinical value.

The Evaluation of Optic Nerves Using 7 Tesla "Silent" Zero Echo Time Imaging in Patients with Leber's Hereditary Optic Neuropathy with or without Idebenone Treatment

Magnetic Resonance Imaging (MRI) of the Optic Nerve is difficult due to the fine extended nature of the structure, strong local magnetic field distortions induced by anatomy, and large motion artefacts associated with eye movement. To address these problems we used a Zero Echo Time (ZTE) MRI sequence with an Adiabatic SPectral Inversion Recovery (ASPIR) fat suppression pulse which also imbues the images with Magnetisation Transfer contrast. We investigated an application of the sequence for imaging the optic nerve in subjects with Leber's hereditary optic neuropathy (LHON). Of particular note is the sequence's near-silent operation, which can enhance image quality of the optic nerve by reducing the occurrence of involuntary saccades induced during Magnetic Resonance (MR) scanning.

Normal measurements of the optic nerve, optic nerve sheath and optic chiasm in the adult population

Background

Imaging assessment of the anterior visual pathway structures, particularly the optic nerves (ON), requires knowledge of normal dimensions. Several studies suggesting techniques and normal ranges have been performed, but most suffer from various methodological flaws. This study is the first to be performed in a South African population.

Objectives

The aim of this study was to establish normal measurements of the ON, optic nerve sheath (ONS) and optic chiasm (OC) on magnetic resonance imaging (MRI).

Method

Eighty normal adults between ages of 12 and 65 years were included in this prospective, quantitative, observational, descriptive study to establish normal measurement of the ON, ONS and OC using a T2W 3D MRI sequence. Measurements (width and height) were undertaken by two observers independently.

Results

A total of 80 participants with a mean age of 35 years were studied: 49 females (61.25%) and 31 males (38.75%). There were no statistical differences in the measurements between gender and age correlation. Interobserver agreement was best for larger structures, that is, OC width and intracranial ON width, respectively. The overall mean of OC width was 13.63 mm (range: 11.13 mm–16.92 mm, standard deviation [s.d.] 1.21); intraorbital ON height at 5 mm behind the globe 2.29 mm (range: 1.63 mm–3.33 mm, s.d. 0.43), and intracranial ON width 4.27 mm (range: 2.46 mm–5.19 mm, s.d. 0.53).

Conclusion

Normal measurements of the anterior visual pathway structures on MRI are best reflected in the larger structures. Interobserver variability was poor for the orbital ON, ONS, intracranial ON height and OC height. We recommend that measurements be obtained for the OC width and intracranial ON width. The overall mean for the OC width is 13.63 mm and intracranial ON width 4.27 mm.

Characterization of diffuse orbital mass using Apparent diffusion coefficient in 3-tesla MRI

Purpose

To evaluate if the apparent diffusion coefficient (ADC) value in diffusion-weighted magnetic resonance imaging (DW-MRI) improves the diagnostic accuracy of diffuse orbital masses.

Materials and methods

ADC DW-MRI was used to evaluate cases of diffuse orbital masses at our institution from 2000 to 2015. Lesions were grouped according to histopathologic diagnosis as, benign, pre-malignant and malignant. Lymphoproliferative lesions were further subgrouped as lymphoma or other lymphoproliferative lesions. The validity of the ADC value for the diffuse orbital mass was compared between groups. The area under curve (AUC) was also calculated.

Results

Thirty-nine cases of diffuse orbital masses were evaluated. The median ADC was 0.58 (25% quartile 0.48; minimum: 0.45; maximum: 1.72 × 10⁽⁻³⁾) for the malignant tumors and 1.19 (25% quartile 0.7; minimum: 0.5; maximum: 1.95 × 10⁽⁻³⁾ mm⁽²⁾ s⁽⁻¹⁾) for benign lesions. This difference in ADC between lesions was statistically significant (Mann Whitney U test P < 0.001). The median ADC was 0.51 (25% quartile 0.48) for lymphomas and 0.9 (25% quartile 0.7) for other lymphoproliferative lesions. This difference in ADC was statistically significant (Mann Whitney U test P = 0.02). An ADC value of 0.8 × 10⁽⁻³⁾ mm⁽²⁾ s⁽⁻¹⁾ was noted as the ideal threshold value for differentiating malignant from benign diffuse orbital masses. The validity of ADC in predicting a malignant or benign diffuse orbital mass had a sensitivity of 87%, specificity of 67% and accuracy of 88%.

Conclusion

ADC is a promising imaging metric to characterize malignant and benign diffuse orbital masses and to distinguish lymphomas from other non−lymphoproliferative lesions.

Uveal melanoma: quantitative evaluation of diffusion-weighted MR imaging in the response assessment after proton-beam therapy, long-term follow-up

PURPOSE

The purpose of this prospective study was to investigate the proton-beam-induced changes in apparent diffusion coefficient (ADC) values of ocular melanoma treated with proton-beam therapy (PBT) in patients undergoing long-term magnetic resonance imaging (MRI) follow-up and to assess whether variations in ADC constitute a reliable biomarker for predicting and detecting the response of ocular melanoma to PBT.

METHODS

Seventeen patients with ocular melanoma treated with PBT were enrolled. All patients underwent conventional MRI and diffusion-weighted imaging (DWI) at baseline and 1, 3, 6, and 18 months after the beginning of therapy. Tumor volumes and ADC values of ocular lesions were measured at each examination. Tumor volumes and mean ADC measurements of the five examination series were compared; correlation of ADC values and tumor regression was estimated.

RESULTS

Mean ADC values of ocular melanomas significantly increased already 1 month after therapy whereas tumor volume significantly decreased only 6 months after therapy. Pretreatment ADC value of ocular melanomas and early change in ADC value 1 month after therapy significantly correlated with tumor regression.

CONCLUSIONS

In ocular melanoma treated with PBT, ADC variations precede volume changes. Both pretreatment ADC and early change in ADC value may predict treatment response, thus expanding the role of DWI from diagnostic to prognostic.

Magnetic resonance imaging of optic nerve

Optic nerves are the second pair of cranial nerves and are unique as they represent an extension of the central nervous system. Apart from clinical and ophthalmoscopic evaluation, imaging, especially magnetic resonance imaging (MRI), plays an important role in the complete evaluation of optic nerve and the entire visual pathway. In this pictorial essay, the authors describe segmental anatomy of the optic nerve and review the imaging findings of various conditions affecting the optic nerves. MRI allows excellent depiction of the intricate anatomy of optic nerves due to its excellent soft tissue contrast without exposure to ionizing radiation, better delineation of the entire visual pathway, and accurate evaluation of associated intracranial pathologies.

MRI helps depict clinically undetectable risk factors in advanced stage retinoblastomas

This study compared high-resolution MRI with histology in advanced stage retinoblastomas in which ophthalmoscopy and ultrasonography did not give an exhaustive depiction of the tumour and/or its extension. MRI of orbits and head in 28 retinoblastoma patients (28 eyes) treated with primary enucleation were evaluated. Iris neoangiogenesis, infiltrations of optic nerve, choroid, anterior segment and sclera suspected at MR and histology were compared. Abnormal anterior segment enhancement (AASE) was also correlated with histologically proven infiltrations. Brain images were also evaluated. Significant values were obtained for: prelaminar optic nerve (ON) sensitivity (0.88), positive predictive value (PPV) (0.75) and negative predictive value (NPV) (0.71); post-laminar ON sensitivity (0.50), specificity (0.83), PPV (0.50) and NPV (0.83); overall choroid sensitivity (0.82), and massive choroid NPV (0.69); scleral specificity (1), and NPV (1). AASE correlated with iris neoangiogenesis in 14 out of 19 eyes, and showed significant values for: overall ON PPV (0.65), prelaminar ON sensitivity (0.65), and PPV (0.61), post-laminar ON NPV (0.64); overall choroid sensitivity (0.77), PPV (0.59) and NPV (0.73); scleral NPV (0.83); anterior segment sensitivity (1), and NPV (1). Odds ratios (OR) and accuracy were significant in scleral and prelaminar optic nerve infiltration. Brain examination was unremarkable in all cases. High-resolution MRI may add important findings to clinical evaluation of advanced stage retinoblastomas.

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.

High-resolution MRI of uveal melanoma using a microcoil phased array at 7 T

High-field MRI is a promising technique for the characterisation of ocular tumours, both in vivo and after enucleation. For in vivo imaging at 7 T, a dedicated three-element microcoil array was constructed as a high-sensitivity receive-only device. Using a dedicated blink/fixation protocol, high-resolution in vivo images could be acquired within 3 min in volunteers and patients with no requirement for post-acquisition image registration. Quantitative measures of axial length, aqueous depth and lens thickness in a healthy volunteer were found to agree well with standard ocular biometric techniques. In a patient with uveal melanoma, in vivo MRI gave excellent tumour/aqueous body contrast. Ex vivo imaging of the enucleated eye showed significant heterogeneity within the tumour.

Guidelines for imaging retinoblastoma: imaging principles and MRI standardization

Retinoblastoma is the most common intraocular tumor in children. The diagnosis is usually established by the ophthalmologist on the basis of fundoscopy and US. Together with US, high-resolution MRI has emerged as an important imaging modality for pretreatment assessment, i.e. for diagnostic confirmation, detection of local tumor extent, detection of associated developmental malformation of the brain and detection of associated intracranial primitive neuroectodermal tumor (trilateral retinoblastoma). Minimum requirements for pretreatment diagnostic evaluation of retinoblastoma or mimicking lesions are presented, based on consensus among members of the European Retinoblastoma Imaging Collaboration (ERIC). The most appropriate techniques for imaging in a child with leukocoria are reviewed. CT is no longer recommended. Implementation of a standardized MRI protocol for retinoblastoma in clinical practice may benefit children worldwide, especially those with hereditary retinoblastoma, since a decreased use of CT reduces the exposure to ionizing radiation.

Diffusion-weighted imaging of malignant ocular masses: initial results and directions for further study

BACKGROUND AND PURPOSE

Ocular masses represent a spectrum of malignant tumors and benign lesions that are sometimes difficult to detect and differentiate by conventional imaging techniques. The aim of this study was to characterize a group of malignant ocular masses with DWI, with the goals of establishing reference data and identifying potential clinical applications for improved noninvasive characterization.

MATERIALS AND METHODS

With institutional review board approval, 26 malignant ocular masses in 22 patients were retrospectively analyzed. Five masses were excluded from further analysis due to nonvisualization. Fifteen retinoblastomas, 5 melanomas, and 1 highly undifferentiated carcinoma were studied. Region-of-interest analysis was performed, and the ADC of each mass was measured and also compared with a normal-appearing thalamus. Lesion thickness was measured, the amount of susceptibility artifact was qualitatively assessed and graded, and the correlation between these factors and retinoblastoma ADC was determined.

RESULTS

Retinoblastomas had an ADC of 0.93 ± 0.3 × 10(-3) mm(2)/s (mean). Melanoma had an ADC of 1.18 ± 0.16 × 10(-3) mm(2)/s. The ADC of retinoblastoma was strongly inversely correlated with lesion thickness, likely representing the effect of partial volume averaging. ADC was not correlated with the amount of subjectively determined susceptibility artifact.

CONCLUSIONS

Malignant ocular tumors were consistently characterized with DWI, though with limitations due to artifact and partial volume averaging. Additional description of DWI of ocular masses and further technical improvements may lead to a clinical role for DWI.

MRI of retinoblastoma

We review the role of MRI in retinoblastoma and simulating lesions. Retinoblastoma is the most common paediatric intra-ocular tumour. It may be endophytic, exophytic or a diffuse infiltrating tumour. MRI can detect intra-ocular, extra-ocular and intracranial extension of the tumour. MRI is essential for monitoring patients after treatment and detection of associated second malignancies. It helps to differentiating the tumour from simulating lesions with leukocoria.

Differentiation between benign and malignant orbital tumors at 3-T diffusion MR-imaging

INTRODUCTION

To differentiate between malignant and benign orbital tumors at 3-T diffusion MR imaging.

METHODS

A retrospective study was conducted on 47 patients (34 males and 13 females aged 4-74 years) with orbital masses. They underwent echo-planar diffusion-weighted MR imaging of the orbit with b-factor of 0, 500, and 1,000 s/mm(2) at 3-T MR unit. Apparent diffusion coefficient (ADC) maps were reconstructed, and the ADC value of the orbital mass was calculated.

RESULTS

The mean ADC value of the malignant orbital tumors (0.84 ± 0.34 × 10(-3) mm(2)/s) was significantly lower (P = 0.001) than that of the benign orbital tumors (1.57 ± 0.33 × 10(-3) mm(2)/s). The selection of an ADC value of 1.15 × 10(-3) mm(2)/s as a threshold value for differentiating malignant orbital tumors from benign lesions has a sensitivity of 95%, a specificity of 91%, and an accuracy of 93%. There was a significant difference in the ADC value between well- and poorly differentiated malignancies (P = 0.005).

CONCLUSION

Apparent diffusion coefficient value at 3 T is an additional noninvasive imaging parameter that can be used for the differentiation of malignant orbital tumors from benign lesions, the characterization of some orbital tumors, as well as the grading of orbital malignancy.

7 Tesla MR imaging of the human eye in vivo

PURPOSE

To develop a protocol which optimizes contrast, resolution and scan time for three-dimensional (3D) imaging of the human eye in vivo using a 7 Tesla (T) scanner and custom radio frequency (RF) coil.

MATERIALS AND METHODS

Initial testing was conducted to reduce motion and susceptibility artifacts. Three-dimensional FFE and IR-TFE images were obtained with variable flip angles and TI times. T(1) measurements were made and numerical simulations were performed to determine the ideal contrast of certain ocular structures. Studies were performed to optimize resolution and signal-to-noise ratio (SNR) with scan times from 20 s to 5 min.

RESULTS

Motion and susceptibility artifacts were reduced through careful subject preparation. T(1) values of the ocular structures are in line with previous work at 1.5T. A voxel size of 0.15 x 0.25 x 1.0 mm(3) was obtained with a scan time of approximately 35 s for both 3D FFE and IR-TFE sequences. Multiple images were registered in 3D to produce final SNRs over 40.

CONCLUSION

Optimization of pulse sequences and avoidance of susceptibility and motion artifacts led to high quality images with spatial resolution and SNR exceeding prior work. Ocular imaging at 7T with a dedicated coil improves the ability to make measurements of the fine structures of the eye.

MR imaging of orbital inflammatory syndrome, orbital cellulitis, and orbital lymphoid lesions: the role of diffusion-weighted imaging

BACKGROUND AND PURPOSE

Orbital inflammatory syndrome (OIS) has clinical features that overlap with orbital lymphoid lesions and orbital cellulitis. Prompt diagnosis is needed in all 3 conditions because the management of each one differs greatly. CT and MR imaging, though useful, do not always distinguish among these conditions. The aim of this study was to identify the role of diffusion-weighted imaging (DWI) in differentiating these 3 diagnoses.

MATERIALS AND METHODS

A retrospective analysis of orbital MR imaging was conducted. T1- and T2-weighted and postcontrast images were analyzed. Region-of-interest analysis was performed by using measurements in areas of abnormality seen on conventional MR imaging sequences and measurements of the ipsilateral thalamus for each patient. The DWI signal intensity of the lesion was expressed as a percentage of average thalamic intensity in each patient. Similarly, lesion apparent diffusion coefficients (ADCs) and lesion-thalamus ADC ratios were calculated. Statistical significance was determined by the Kruskal-Wallis test, and post hoc pairwise comparisons, by the Mann-Whitney U test for DWI-intensity ratio, ADC, and ADC ratio.

RESULTS

A significant difference was noted in DWI intensities, ADC, and ADC ratio between OIS, orbital lymphoid lesions, and orbital cellulitis (P < .05). Lymphoid lesions were significantly brighter than OIS, and OIS lesions were significantly brighter than cellulitis. Lymphoid lesions showed lower ADC than OIS and cellulitis. A trend was seen toward lower ADC in OIS than in cellulitis (P = .17).

CONCLUSIONS

DWI may help differentiate OIS from lymphoid lesions and cellulitis and may allow more rapid management.