Reproducibility of Lesion Count in Various Subregions on MRI Scans in Multiple Sclerosis

Consensus of algorithms for lesion segmentation in brain MRI studies of multiple sclerosis

Segmentation of multiple sclerosis (MS) lesions on brain MRI scans is crucial for diagnosis, disease and treatment monitoring but is a time-consuming task. Despite several automated algorithms have been proposed, there is still no consensus on the most effective method. Here, we applied a consensus-based framework to improve lesion segmentation on T1-weighted and FLAIR scans. The framework is designed to combine publicly available state-of-the-art deep learning models, by running multiple segmentation tasks before merging the outputs of each algorithm. To assess the effectiveness of the approach, we applied it to MRI datasets from two different centers, including a private and a public dataset, with 131 and 30 MS patients respectively, with manually segmented lesion masks available. No further training was performed for any of the included algorithms. Overlap and detection scores were improved, with Dice increasing by 4-8% and precision by 3-4% respectively for the private and public dataset. High agreement was obtained between estimated and true lesion load (ρ = 0.92 and ρ = 0.97) and count (ρ = 0.83 and ρ = 0.94). Overall, this framework ensures accurate and reliable results, exploiting complementary features and overcoming some of the limitations of individual algorithms.

Automatic lesion detection at Multiple Sclerosis patients - Comparison of 2D- and 3D-FLAIR-datasets

BACKGROUND

Multiple Sclerosis (MS) is a common autoimmune inflammatory disease of the central nervous system (CNS). Magnetic Resonance Imaging (MRI) allows a sensitive assessment of the CNS and is established for diagnostic, prognostic and (therapy-) monitoring purposes. Especially lesion counting in T2- or Fluid Attenuated Inversion Recovery (FLAIR)-weighted images plays a decisive role in clinical routine. Software-packages allowing an automatic evaluation of image data are increasingly established aiming a faster and improved workflow. These programs allow e.g. the counting, spatial attribution and volumetry of MS-lesions in FLAIR-weighted images. Research has shown that 3D-FLAIR-sequences are superior to 2D-FLAIR-sequences in visual evaluation of lesion burden in MS. An influence on the automatic analysis is expectable but not yet systematically studied. This work will therefore investigate the influence of 2D- and 3D datasets on the results of an automatic assessment.

MATERIAL AND METHODS

In this prospective study, 80 Multiple Sclerosis patients underwent a clinically indicated routine MRI examination. The clinical routine protocol already including a 3D-FLAIR sequence was adapted by an additional 2D-FLAIR sequence also conform to the 2021 MAGNIMS-CMSCNAIMS consensus recommendations. To obtain a quantitative analysis for assessment of amount, dissemination and volume of the lesions, the acquired MR images were post-processed using the CE-certified Software mdbrain (mediaire, Berlin, Germany). The resulting data were statistically analysed using the paired t-test for normally distributed data and the Wilcoxon-signed-rank-test for not normally distributed data respectively. Demographic data and data such as the subtype, duration, severity and therapy of the disease were collected, pseudonymized and evaluated.

RESULTS

There is a significant difference concerning the total number and lesion volume with more lesions being detected (2D: 29.7, +/- 20.22 sd; 3D: 40.1 +/- 31.67 sd; p < 0.0001) but lower total volume (2D: 6.24 +/- 6.11 sd; 3D: 5.39 +/- 6.37 sd; p < 0.0001) when using the 3D- sequence. Especially significantly more small lesions in the unspecific white matter and infratentorial region were detected by using the 3D-FLAIR sequence (p < 0.0001) compared to the 2D-FLAIR image. Main reason for the lower total volume in the 3D-FLAIR sequence was the calculated volume for periventricular lesions which was significantly beneath the calculated volume from the 2D-FLAIR sequence (p < 0.0001).

CONCLUSION

Automatic lesion counting and volumetry is feasible with both 2D- and 3D-weightend FLAIR images. Still, it leads to partly significant differences even between two sequences that both are conform to the 2021 MAGNIMS-CMSCNAIMS consensus recommendations. This study contributes valuable insights into the impact of using different input data from the same patient for automated MS lesion evaluation.

The impact of image contrast, resolution and reader expertise on black hole identification in Multiple Sclerosis

OBJECTIVES

In the neuroradiological work-up of Multiple Sclerosis (MS), the detection of "black holes" (BH) represent an information of undeniable importance. Nevertheless, different sequences can be used in clinical practice to evaluate BH in MS. Aim of this study was to investigate the possible impact of different sequences, resolutions, and levels of expertise on the intra- and inter-rater reliability identification of BH in MS.

METHODS

Brain MRI scans of 85 MS patients (M/F = 22/63; mean age = 36.0 ± 10.2 years) were evaluated in this prospective single-center study. The acquisition protocol included a 3 mm SE-T1w sequence, a 1 mm 3D-GrE-T1w sequence from which a resliced 3 mm sequence was also obtained. Images were evaluated independently by two readers of different expertise at baseline and after a wash-out period of 30 days. The intraclass correlation coefficient (ICC) was calculated as an index of intra and inter-reader reliability.

RESULTS

For both readers, the intra-reader ICC analysis showed that the 3 mm SE-T1w and 3 mm resliced GrE-T1w images achieved an excellent performance (both with an ICC ≥ 0.95), while 1 mm 3D-GrE-T1w scans achieved a moderate one (ICC < 0.90). The inter-reader analysis showed that each of the three sequences achieved a moderate performance (all ICCs < 0.90).

CONCLUSIONS

The 1 mm 3D-GrE-T1w sequence seems to be prone to a greater intra-reader variability compared to the 3 mm SE-T1w, with this effect being driven by the higher spatial resolution of the first sequence. To ensure reliability levels comparable with the standard SE-T1w in BH count, an assessment on a 3 mm resliced GrE-T1w sequence should be recommended.

Robust Ensemble of Two Different Multimodal Approaches to Segment 3D Ischemic Stroke Segmentation Using Brain Tumor Representation Among Multiple Center Datasets

Ischemic stroke segmentation at an acute stage is vital in assessing the severity of patients' impairment and guiding therapeutic decision-making for reperfusion. Although many deep learning studies have shown attractive performance in medical segmentation, it is difficult to use these models trained on public data with private hospitals' datasets. Here, we demonstrate an ensemble model that employs two different multimodal approaches for generalization, a more effective way to perform on external datasets. First, after we jointly train a segmentation model on diffusion-weighted imaging (DWI) and apparent diffusion coefficient (ADC) MR modalities, the model is inferred on the DWI images. Second, a channel-wise segmentation model is trained by concatenating the DWI and ADC images as input, and then is inferred using both MR modalities. Before training with ischemic stroke data, we utilized BraTS 2021, a public brain tumor dataset, for transfer learning. An extensive ablation study evaluates which strategy learns better representations for ischemic stroke segmentation. In our study, nnU-Net well-known for robustness is selected as our baseline model. Our proposed method is evaluated on three different datasets: the Asan Medical Center (AMC) I and II, and the 2022 Ischemic Stroke Lesion Segmentation (ISLES). Our experiments are widely validated over a large, multi-center, and multi-scanner dataset with a huge amount of 846 scans. Not only stroke lesion models can benefit from transfer learning using brain tumor data, but combining the MR modalities using different training schemes also highly improves segmentation performance. The method achieved a top-1 ranking in the ongoing ISLES'22 challenge and performed particularly well on lesion-wise metrics of interest to neuroradiologists, achieving a Dice coefficient of 78.69% and a lesion-wise F1 score of 82.46%. Also, the method was relatively robust on the AMC I (Dice, 60.35%; lesion-wise F1, 68.30%) and II (Dice; 74.12%; lesion-wise F1, 67.53%) datasets in different settings. The high segmentation accuracy of our proposed method could improve radiologists' ability to detect ischemic stroke lesions in MRI images. Our model weights and inference code are available on https://github.com/MDOpx/ISLES22-model-inference .

Image harmonization improves consistency of intra-rater delineations of MS lesions in heterogeneous MRI

Clinical magnetic resonance images (MRIs) lack a standard intensity scale due to differences in scanner hardware and the pulse sequences used to acquire the images. When MRIs are used for quantification, as in the evaluation of white matter lesions (WMLs) in multiple sclerosis, this lack of intensity standardization becomes a critical problem affecting both the staging and tracking of the disease and its treatment. This paper presents a study of harmonization on WML segmentation consistency, which is evaluated using an object detection classification scheme that incorporates manual delineations from both the original and harmonized MRIs. A cohort of ten people scanned on two different imaging platforms was studied. An expert rater, blinded to the image source, manually delineated WMLs on images from both scanners before and after harmonization. It was found that there is closer agreement in both global and per-lesion WML volume and spatial distribution after harmonization, demonstrating the importance of image harmonization prior to the creation of manual delineations. These results could lead to better truth models in both the development and evaluation of automated lesion segmentation algorithms.

Evaluation of transorbital sonography measures of optic nerve diameter in the context of global and regional brain volume in multiple sclerosis

Transorbital sonography (TOS) could be a swift and convenient method to detect the atrophy of the optic nerve, possibly providing a marker that might reflect other quantitative structural markers of multiple sclerosis (MS). Here we evaluate the utility of TOS as a complementary tool for assessing optic nerve atrophy, and investigate how TOS-derived measures correspond to volumetric brain markers in MS. We recruited 25 healthy controls (HC) and 45 patients with relapsing-remitting MS and performed B-mode ultrasonographic examination of the optic nerve. Patients additionally underwent MRI scans to obtain T1-weighted, FLAIR and STIR images. Optic nerve diameters (OND) were compared between HC, MS patients with and without history of optic neuritis (non-ON) using a mixed-effects ANOVA model. The relationship between within-subject-average OND and global and regional brain volumetric measures was investigated using FSL SIENAX, voxel-based morphometry and FSL FIRST. OND was significantly different between HC-MS (HC = 3.2 ± 0.4 mm, MS = 3 ± 0.4 mm; p < 0.019) and we found significant correlation between average OND and normalised whole brain (β = 0.42, p < 0.005), grey matter (β = 0.33, p < 0.035), white matter (β = 0.38, p < 0.012) and ventricular cerebrospinal fluid volume (β = - 0.36, p < 0.021) in the MS group. History of ON had no impact on the association between OND and volumetric data. In conclusion, OND is a promising surrogate marker in MS, that can be simply and reliably measured using TOS, and its derived measures correspond to brain volumetric measures. It should be further explored in larger and longitudinal studies.