Spectrally fat-suppressed coronal 2D TSE sequences may be more sensitive than 2D STIR for the detection of hyperintense optic nerve lesions

Aliphatic and Olefinic Fat Suppression in the Orbit Using Polarity-altered Spectral and Spatial Selective Acquisition (PASTA) with Opposed Phase

PURPOSE

Fatty acid composition of the orbit makes it challenging to achieve complete fat suppression during orbit MR imaging. Implementation of a fat suppression technique capable of suppressing signals from saturated (aliphatic) and unsaturated (olefinic or protons at double-bonded carbon sites) fat would improve the visualization of an optical nerve. Furthermore, the ability to semi-quantify the fractions of aliphatic and olefinic fat may potentially provide valuable information in assessing orbit pathology.

METHODS

A phantom study was conducted on various oil samples on a clinical 3 Tesla scanner. The imaging protocol included three 2D fast spin echo (FSE) sequences: in-phase, polarity-altered spectral and spatial selective acquisition (PASTA), and a combination of PASTA with opposed phase in olefinic and aliphatic chemical shift. The results were validated against high-resolution 11.7T NMR and compared with images acquired with spectral attenuated inversion recovery (SPAIR) and chemical shift selective (CHESS) fat suppression techniques. In-vivo data were acquired on eight healthy subjects and were compared with the prior histological studies.

RESULTS

PASTA with opposed phase achieved complete suppression of fat signals in the orbits and provided images of well-delineated optical nerves and muscles in all subjects. The olefinic fat fraction in the olive, walnut, and fish oil phantoms at 3T was found to be 5.0%, 11.2%, and 12.8%, respectively, whereas 11.7T NMR provides the following olefinic fat fractions: 6.0% for olive, 11.5% for walnut, and 12.6% for fish oils. For the in-vivo study, on average, olefinic fat accounted for 9.9% ± 3.8% of total fat while the aliphatic fat fraction was 90.1% ± 3.8%, in the normal orbits.

CONCLUSION

We have introduced a new fat suppression technique using PASTA with opposed phase and applied it to human orbits. The purposed method achieves an excellent orbital fat suppression and the quantification of aliphatic and olefinic fat signals.

Detection of lesions in the optic nerve with magnetic resonance imaging using a 3D convolutional neural network

BACKGROUND

Optic neuritis (ON) is one of the first manifestations of multiple sclerosis, a disabling disease with rising prevalence. Detecting optic nerve lesions could be a relevant diagnostic marker in patients with multiple sclerosis.

OBJECTIVES

We aim to create an automated, interpretable method for optic nerve lesion detection from MRI scans.

MATERIALS AND METHODS

We present a 3D convolutional neural network (CNN) model that learns to detect optic nerve lesions based on T2-weighted fat-saturated MRI scans. We validated our system on two different datasets (N = 107 and 62) and interpreted the behaviour of the model using saliency maps.

RESULTS

The model showed good performance (68.11% balanced accuracy) that generalizes to unseen data (64.11%). The developed network focuses its attention to the areas that correspond to lesions in the optic nerve.

CONCLUSIONS

The method shows robustness and, when using only a single imaging sequence, its performance is not far from diagnosis by trained radiologists with the same constraint. Given its speed and performance, the developed methodology could serve as a first step to develop methods that could be translated into a clinical setting.

Optical coherence tomography for detection of asymptomatic optic nerve lesions in clinically isolated syndrome

Objective

To evaluate the ability of intereye retinal thickness difference (IETD) measured by optical coherence tomography (OCT) to detect asymptomatic optic nerve involvement in clinically isolated syndrome (CIS).

Methods

We conducted a cross-sectional study of patients who recently presented a CIS (≤4.5 months). All patients underwent OCT and brain/optic nerve MRI. Optic nerve involvement was defined clinically (episode of optic neuritis [ON] or not) and radiologically (optic nerve hypersignal on 3D double inversion recovery [3D-DIR]). We evaluated the sensitivity and specificity of previously published IETD thresholds and report the observed optimal thresholds for identifying symptomatic optic nerve involvement but also for identifying asymptomatic optic nerve involvement (optic nerve hypersignal without ON history). Primary outcomes were ganglion cell–inner plexiform layer (GC-IPL) and peripapillary retinal nerve fiber layer IETD.

Results

The study group consisted of 130 patients. In the CIS with ON group, 3D-DIR showed a hypersignal in all 41 symptomatic optic nerves and in 11 asymptomatic optic nerves. In the CIS without ON group, 3D-DIR showed a unilateral optic nerve hypersignal in 22 patients and a bilateral optic nerve hypersignal in 7 patients. For the detection of symptomatic and asymptomatic optic nerve lesion, GC-IPL IETD had better performance. We found an optimal GC-IPL IETD threshold ≥2.83 µm (sensitivity 88.2, specificity 83.3%) for the detection of symptomatic lesions and an optimal GC-IPL IETD ≥1.42 µm (sensitivity 89.3%, specificity 72.6%) for the detection of asymptomatic lesions.

Conclusions

Detection of asymptomatic optic nerve lesions in CIS requires lower IETD thresholds than previously reported. GC-IPL IETD represents an alternative biomarker to MRI for the detection of asymptomatic optic nerve lesions.

Classification of evidence

This study provides Class I evidence that OCT accurately identifies asymptomatic optic nerve involvement in patients with CIS.