Irene Margaret Vavasour^1,2, Pierre Becquart¹, Tigris Joseph^1,2, Jasmine Gill¹, Guojun Zhao³, Farhia Abdullahi¹, Emily Kamma¹, Kaya Frese¹, Robert Carruthers¹, Shannon H Kolind^1,2,3, Alice Schabas¹, Ana-Luiza Sayao¹, Virginia Devonshire¹, Roger Tam^1,3, GR Wayne Moore^1,2, Sophie Stukas¹, Cheryl Wellington¹, Jacqueline Quandt¹, Anthony Traboulsee^1,3, David Li^1,3, and Cornelia Laule^1,2
1University of British Columbia, Vancouver, BC, Canada, ²International Collaboration on Repair Discoveries, Vancouver, BC, Canada, ³MSMRI Research Group, Vancouver, BC, Canada

Diffusely abnormal white matter (DAWM) is commonly found on conventional brain MRI in all stages of multiple sclerosis (MS). Blood markers associated with prior or ongoing myelin and axonal damage were not found to be elevated in people with DAWM relative to people without DAWM. Within areas of DAWM, no differences between MS phenotypes were found possibly suggesting that the tissue abnormalities within DAWM are similar across the clinical course, from early to late stages of MS.

Ryan K Robison^1,2,3, Atlee A Witt⁴, Anna JE Combes^2,3, Kristin P O'Grady^2,3, Logan E Prock³, Delaney Houston⁵, Dann C Martin⁵, Margareta Clarke⁵, and Seth A Smith^2,3
1Philips, Nashville, TN, United States, ²Radiology and Radiological Sciences, Vanderbilt University Medical Center, Nashville, TN, United States, ³Vanderbilt University Institute of Imaging Sciences, Nashville, TN, United States, ⁴Vanderbilt University School of Medicine, Nashville, TN, United States, ⁵Vanderbilt University Medical Center, Nashville, TN, United States

Central vein sign is believed to be a specific marker for the diagnosis of multiple sclerosis (MS) but has not yet been reported in the spine. This work used SWI in the cervical spinal cord of MS patients and healthy controls at 7T to look for evidence of central veins. SWI hypointensities were observed frequently around the cord periphery, between the white and grey matter within the cord, and along nerve roots. Hypointensities were also observed within spinal cord lesions, providing candidate evidence of central veins in lesions of the spinal cord.

Till Huelnhagen^1,2,3, Omar al Louzi^4,5, Lynn Daboul⁴, Jonas Richiardi², Daniel S. Reich⁴, Tobias Kober^1,2,3, and Pascal Sati^5
1Advanced Clinical Imaging Technology, Siemens Healthcare AG, Lausanne, Switzerland, ²Department of Radiology, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, ³Signal Processing Laboratory (LTS 5), Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland, ⁴Translational Neuroradiology Section, National Institute of Neurological Disorders and Stroke, National Institutes of Health (NIH), Bethesda, MD, United States, ⁵Department of Neurology, Cedars-Sinai Medical Center, Los Angeles, CA, United States

Central vein sign (CVS) assessment has shown potential to improve differential diagnosis in multiple sclerosis, but automating this task remains non-trivial. As human inter-rater agreement was reported to improve by separating the tasks of lesion exclusion and CVS assessment, we hypothesized that this could also benefit automated CVS assessment. To test this hypothesis, we implemented a novel multi-level classifier for automated CVS assessment and trained and evaluated it in more than 9400 expert-reviewed lesions. The new approach outperforms previous methods, achieving per-class accuracies of 76%–83% in an unseen testing set and >90% accuracy to identify MS cases.

Loredana Storelli¹, Matteo Azzimonti^1,2, Mor Gueye^1,2, Paolo Preziosa^1,2, Carmen Vizzino¹, Gioacchino Tedeschi³, Nicola De Stefano⁴, Patrizia Pantano^5,6, Massimo Filippi^1,2,7,8,9, and Maria A. Rocca^1,2,9
1Neuroimaging Research Unit, Division of Neuroscience, IRCCS San Raffaele Scientific Institute, Milan, Italy, ²Neurology Unit, IRCCS San Raffaele Scientific Institute, Milan, Italy, ³Department of Advanced Medical and Surgical Sciences, and 3T MRI-Center, University of Campania “Luigi Vanvitelli”, Maples, Italy, ⁴Department of Medicine, Surgery and Neuroscience, University of Siena, Siena, Italy, ⁵Department of Human Neurosciences, Sapienza University of Rome, Rome, Italy, ⁶IRCCS NEUROMED, Pozzilli, Italy, ⁷Neurorehabilitation Unit, IRCCS San Raffaele Scientific Institute, Milan, Italy, ⁸Neurophysiology Service, IRCCS San Raffaele Scientific Institute, Milan, Italy, ⁹Vita-Salute San Raffaele University, Milan, Italy

Artificial intelligence (AI) approaches have been applied in several fields of multiple sclerosis (MS) in recent years. However, their application to predict disease progression remains largely unexplored. In this study, we obtained a robust and accurate AI tool for predicting clinical and cognitive evolution at two years for MS patients, based on just T1-weighted and T2-weighted brain MRI scans at baseline visit, which exceeded the performance of two expert physicians blinded to patients’ clinical history. This algorithm has the potential to be an important tool to support physicians for a prompt recognition of MS patients at risk of disease worsening.

Junghwa Kang¹, Siyun Jung¹, Jeongmin Yim², Jinhee Jang³, and Yoonho Nam^1
1Division of biomedical engineering, Hankuk University of Foreign Studies, gyunggi-do,yongin-si, Korea, Republic of, ²College of Medicine, Catholic University of Korea, Seoul, Korea, Republic of, ³Department of Radiology, Seoul St. Mary's Hospital, Seoul, Korea, Republic of

Detecting changes in multiple sclerosis (MS) lesions through follow-up MR images is an important but time-consuming and subjective process. In this work, we propose a fully automatic deep learning-based method to detect new MS lesions. The model was trained and tuned using the MSSEG2 challenge dataset. First, the brain and spinal cord masks were generated, and registration between two time points was performed. Then, new lesions and whole lesions were segmented by patch-wise inputs, respectively. The final mask for new lesions was produced by comparing these two segmentations, and in this way we could effectively reduce false-positives.

Tigris S. Joseph^1,2, Hanwen Liu^2,3,4, Guojun Zhao⁴, Shannon H. Kolind^1,2,4,5, Robert Carruthers⁴, Alice Schabas⁴, Ana-Luiza Sayao⁴, Virginia Devonshire⁴, Roger Tam^5,6, G. R. Wayne Moore^2,4,7, David K. B. Li^4,5, Anthony Traboulsee⁴, Irene M. Vavasour^2,5, and Cornelia Laule^1,2,5,7
1Physics and Astronomy, University of British Columbia, Vancouver, BC, Canada, ²International Collaboration on Repair Discoveries, University of British Columbia, Vancouver, BC, Canada, ³Montreal Neurological Institute - Hospital, McGill University, Montreal, QC, Canada, ⁴Medicine, University of British Columbia, Vancouver, BC, Canada, ⁵Radiology, University of British Columbia, Vancouver, BC, Canada, ⁶School of Biomedical Engineering, University of British Columbia, Vancouver, BC, Canada, ⁷Pathology & Laboratory Medicine, University of British Columbia, Vancouver, BC, Canada

Multiple sclerosis (MS) lesion volume is commonly reported but is not pathologically specific or strongly associated with MS disability. Lesions are not necessarily demyelinated, which may be why lesion volume and disability are not strongly correlated. MS lesion volume was compared to Myelin Water Imaging (MWI) derived metrics related to myelin content (myelin water fraction) and inflammation (geometric mean T₂) in MS normal appearing white matter and lesions. Weak correlations were observed between lesion volume and MWF, highlighting that the amount of lesional tissue does not reflect the degree of myelin damage within the brain.

Emily Louise Baadsvik¹, Markus Weiger¹, Romain Froidevaux¹, Wolfgang Faigle², Benjamin Victor Ineichen³, and Klaas Paul Pruessmann^1
1Institute for Biomedical Engineering, ETH Zurich and University of Zurich, Zurich, Switzerland, ²Neuroimmunology and MS Research Section, Neurology Clinic, University of Zurich, University Hospital Zurich, Zurich, Switzerland, ³Department of Neuroradiology, Clinical Neuroscience Center, University Hospital Zurich, University of Zurich, Zurich, Switzerland

Signal from the myelin bilayer can be captured using dedicated short-T₂ imaging techniques, and analysis methods to isolate the contribution of the bilayer from other signal sources have been proposed. However, such analysis methods are often validated on samples in which the background water signal has been suppressed. Thus, it is unclear whether these techniques can successfully map myelin content in a clinical setting. Here, we apply and evaluate a previously proposed technique for myelin mapping to multiple sclerosis brain tissue without water suppression. We conclude that the quality of the myelin maps is comparable to that of water-suppressed samples.

Tigris S. Joseph^1,2, Hanwen Liu^2,3,4, Shannon H. Kolind^1,2,4,5, Guojun Zhao⁴, Peng Sun⁶, Robert Carruthers⁴, Alice Schabas⁴, Ana-Luiza Sayao⁴, Virginia Devonshire⁴, Roger Tam^5,7, G. R. Wayne Moore^2,4,8, David K. B. Li^4,5, Sheng-Kwei Song⁹, Anthony Traboulsee⁴, Irene M. Vavasour^2,5, and Cornelia Laule^1,2,5,8
1Physics and Astronomy, University of British Columbia, Vancouver, BC, Canada, ²International Collaboration on Repair Discoveries, University of British Columbia, Vancouver, BC, Canada, ³Montreal Neurological Institute - Hospital, McGill University, Montreal, QC, Canada, ⁴Medicine, University of British Columbia, Vancouver, BC, Canada, ⁵Radiology, University of British Columbia, Vancouver, BC, Canada, ⁶Imaging Physics, MD Anderson Cancer Center, Houston, TX, United States, ⁷School of Biomedical Engineering, University of British Columbia, Vancouver, BC, Canada, ⁸Pathology & Laboratory Medicine, University of British Columbia, Vancouver, BC, Canada, ⁹Radiology, Washington University, St. Louis, MO, United States

Myelin Water Imaging (MWI), Diffusion Basis Spectrum Imaging (DBSI) and Neurite Orientation Dispersion and Density Imaging (NODDI) were used to assessed relationships between myelin and axon-related measures in 122 multiple sclerosis (MS) participants and 16 healthy controls. Neurite density index (NDI) correlated strongly with radial diffusivity and weakly with myelin water fraction, suggesting that radial diffusivity also captures diffusion in dendrites. The lack of correlation between NDI and fiber fraction was surprising given that both metrics are meant to relate to axon density. MWI, DBSI and NODDI provide unique and complementary information about MS damage.

Tsen-Hsuan (Abby) Lin¹, William M Spees¹, Michael Wallendorf², Peng Sun^1,3, Junqian Xu⁴, Anne H Cross^5,6, and Sheng-Kwei Song^1,6
1Radiology, Washington University School of Medicine, St. Louis, MO, United States, ²Biostatistics, Washington University School of Medicine, St. Louis, MO, United States, ³Imaging Physics, MD Anderson Cancer Center, Houston, TX, United States, ⁴Radiology, Baylor College of Medicine, Houston, TX, United States, ⁵Neurology, Washington Univeristy School of Medicine, St. Louis, MO, United States, ⁶Hope Center for Neurological Disorders, Washington University School of Medicine, St. Louis, MO, United States

We have introduced diffusion basis spectrum imaging (DBSI) to detect, differentiate, and quantify coexisting pathologies in people with multiple sclerosis (MS). Recently, we performed functional DBSI and DTI with flashing-checkerboard stimulation. DBSI-derived radial diffusivity (DBSI-RD) decreased significantly during visual stimulation while DTI-RD did not change. In this study, we employed fDBSI to assess optic nerve function and pathology simultaneously in MS. Axonal loss and vasogenic edema/increased extracellular space attenuated optic nerve response to visual stimulation.  

Rajiv G Menon¹, Azadeh Sharafi¹, Marco Muccio¹, Tyler E. Smith², Ilya Kister², Yulin Ge¹, and Ravinder R Regatte^1
1Center for Biomedical Imaging, Grossman school of Medicine, NYU Langone Health, New York, NY, United States, ²Department of Neurology, Grossman School of Medicine, NYU Langone Health, New York, NY, United States

A 3D multi-parameter MRF imaging technique capable of quantifying 4 different parameters (T₁, T₂, T_1ρ, B₁) was developed for multiple sclerosis(MS) application. The 3D-MRF technique was tested with multiple shots on a standardized NIST/ISMRM phantom, and brain imaging in 5 healthy volunteers. The 3D-MRF was also tested on two MS patients with chronic and enhancing MS lesions. The phantom studies demonstrate that quantitative maps are in agreement with reference, healthy volunteer studies show the agreement between multiple shots and SNR improvement with increasing shots. The MS patient results demonstrate the potential use of mutli-parameter quantitative mapping for WM characterization.