Spinal cord metrics derived from diffusion MRI: improvement in prognostication in cervical spondylotic myelopathy compared with conventional MRI
(2024) Journal of Neurosurgery. Spine, 41 (5), pp. 639-647.
Zhang, J.K.a b , Yakdan, S.b , Kaleem, M.I.b , Javeed, S.b , Greenberg, J.K.b , Botterbush, K.S.b , Benedict, B.b , Reis, M.c , Hongsermeier-Graves, N.a , Twitchell, S.a , Sherrod, B.a , Mazur, M.S.a , Mahan, M.A.a , Dailey, A.T.a , Bisson, E.F.a , Song, S.-K.c , Ray, W.Z.b
a Department of Neurosurgery, Clinical Neurosciences Center, University of Utah, Salt Lake City, UT, United States
b Departments of2Neurosurgery and
c 3Radiology, Washington University School of Medicine, St. Louis, MO, United States
Abstract
OBJECTIVE: A major shortcoming in optimizing care for patients with cervical spondylotic myelopathy (CSM) is the lack of robust quantitative imaging tools offered by conventional MRI. Advanced MRI modalities, such as diffusion MRI (dMRI), including diffusion tensor imaging (DTI) and diffusion basis spectrum imaging (DBSI), may help address this limitation by providing granular evaluations of spinal cord microstructure. METHODS: Forty-seven patients with CSM underwent comprehensive clinical assessments and dMRI, followed by DTI and DBSI modeling. Conventional MRI metrics included 10 total qualitative and quantitative assessments of spinal cord compression in both the sagittal and axial planes. The dMRI metrics included 12 unique measures including anisotropic tensors, reflecting axonal diffusion, and isotropic tensors, describing extraaxonal diffusion. The primary outcome was the modified Japanese Orthopaedic Association (mJOA) score measured at 2 years postoperatively. Extreme gradient boosting-supervised classification algorithms were used to classify patients into disease groups and to prognosticate surgical outcomes at 2-year follow-up. RESULTS: Forty-seven patients with CSM, including 24 (51%) with a mild mJOA score, 12 (26%) with a moderate mJOA score, and 11 (23%) with a severe mJOA score, as well as 21 control subjects were included. In the classification task, the traditional MRI metrics correctly assigned patients to healthy control versus mild CSM versus moderate/severe CSM cohorts, with an accuracy of 0.647 (95% CI 0.64-0.65). In comparison, the DTI model performed with an accuracy of 0.52 (95% CI 0.51-0.52) and the DBSI model’s accuracy was 0.81 (95% CI 0.808-0.814). In the prognostication task, the traditional MRI metrics correctly predicted patients with CSM who improved at 2-year follow-up on the basis of change in mJOA, with an accuracy of 0.58 (95% CI 0.57-0.58). In comparison, the DTI model performed with an accuracy of 0.62 (95% CI 0.61-0.62) and the DBSI model had an accuracy of 0.72 (95% CI 0.718-0.73). CONCLUSIONS: Conventional MRI is a powerful tool to assess structural abnormality in CSM but is inherently limited in its ability to characterize spinal cord tissue injury. The results of this study demonstrate that advanced imaging techniques, namely DBSI-derived metrics from dMRI, provide granular assessments of spinal cord microstructure that can offer better diagnostic and prognostic utility.
Author Keywords
cervical spondylotic myelopathy; diffusion MRI; machine learning; magnetic resonance imaging
Document Type: Article
Publication Stage: Final
Source: Scopus
Relationships between abdominal adipose tissue and neuroinflammation with diffusion basis spectrum imaging in midlife obesity
(2024) Obesity, .
Dolatshahi, M.a , Commean, P.K.a , Rahmani, F.a , Xu, Y.b , Liu, J.b , Hosseinzadeh Kassani, S.a , Naghashzadeh, M.a , Lloyd, L.a , Nguyen, C.a , McBee Kemper, A.a , Hantler, N.a , Ly, M.a , Yu, G.a , Flores, S.a c , Ippolito, J.E.a d , Song, S.-K.a , Sirlin, C.B.e , Dai, W.f , Mittendorfer, B.g , Morris, J.C.c h , Benzinger, T.L.S.a c , Raji, C.A.a c h
a Mallinckrodt Institute of Radiology, Washington University in St. Louis, St. Louis, MO, United States
b Division of Public Health Sciences, Department of Surgery, Washington University School of Medicine, St. Louis, MO, United States
c Knight Alzheimer Disease Research Center, Washington University School of Medicine, St. Louis, MO, United States
d Department of Biochemistry and Molecular Biophysics, Washington University School of Medicine, St. Louis, MO, United States
e Liver Imaging Group, Department of Radiology, University of California, Los Angeles, CA, United States
f Department of Computer Science, State University of New York at Binghamton, Binghamton, NY, United States
g Departments of Medicine and Nutrition & Exercise Physiology, University of Missouri School of Medicine, Columbia, MO, United States
h Department of Neurology, Washington University School of Medicine, St Louis, MO, United States
Abstract
Objective: This study investigated how obesity, BMI ≥ 30 kg/m2, abdominal adiposity, and systemic inflammation relate to neuroinflammation using diffusion basis spectrum imaging. Methods: We analyzed data from 98 cognitively normal midlife participants (mean age: 49.4 [SD 6.2] years; 34 males [34.7%]; 56 with obesity [57.1%]). Participants underwent brain and abdominal magnetic resonance imaging (MRI), blood tests, and amyloid positron emission tomography (PET) imaging. Abdominal visceral and subcutaneous adipose tissue (VAT and SAT, respectively) was segmented, and Centiloids were calculated. Diffusion basis spectrum imaging parameter maps were created using an in-house script, and tract-based spatial statistics assessed white matter differences in high versus low BMI values, VAT, SAT, insulin resistance, systemic inflammation, and Centiloids, with age and sex as covariates. Results: Obesity, high VAT, and high SAT were linked to lower axial diffusivity, reduced fiber fraction, and increased restricted fraction in white matter. Obesity was additionally associated with higher hindered fraction and lower fractional anisotropy. Also, individuals with high C-reactive protein showed lower axial diffusivity. Higher restricted fraction correlated with continuous BMI and SAT particularly in male individuals, whereas VAT effects were similar in male and female individuals. Conclusions: The findings suggest that, at midlife, obesity and abdominal fat are associated with reduced brain axonal density and increased inflammation, with visceral fat playing a significant role in both sexes. (Figure presented.). © 2024 The Author(s). Obesity published by Wiley Periodicals LLC on behalf of The Obesity Society.
Funding details
1Florida Alzheimer’s Disease Research CenterADRC
National Institute on AgingNIA1RF1AG072637‐01
National Institute on AgingNIA
Charles F. and Joanne Knight Alzheimer Disease Research Center, Washington University in St. LouisKGADP30AG066444
Charles F. and Joanne Knight Alzheimer Disease Research Center, Washington University in St. LouisKGAD
P01AG026276
P01AG003991
Document Type: Article
Publication Stage: Article in Press
Source: Scopus
Axonal damage and inflammation response are biological correlates of decline in small-world values: a cohort study in autosomal dominant Alzheimer’s disease
(2024) Brain Communications, 6 (5), art. no. fcae357, .
Vermunt, L.a b , Sutphen, C.L.c , Dicks, E.a d , de Leeuw, D.M.a , Allegri, R.F.e , Berman, S.B.f , Cash, D.M.g , Chhatwal, J.P.h , Cruchaga, C.c , Day, G.S.i , Ewers, M.j k , Farlow, M.R.l , Fox, N.C.m n , Ghetti, B.o , Graff-Radford, N.R.i , Hassenstab, J.c , Jucker, M.k o , Karch, C.M.c , Kuhle, J.p , Laske, C.k o , Levin, J.k q , Masters, C.L.r s , McDade, E.c , Mori, H.t , Morris, J.C.c , Perrin, R.J.c , Preische, O.k o , Schofield, P.R.u , Suárez-Calvet, M.v w x , Xiong, C.c , Scheltens, P.a y , Teunissen, C.E.b , Visser, P.J.z , Bateman, R.J.c , Benzinger, T.L.S.c , Fagan, A.M.c , Gordon, B.A.c , Tijms, B.M.a , on behalf of the Dominantly Inherited Alzheimer Networkaa
a Alzheimer center Amsterdam, Department of Neurology, Amsterdam Neuroscience, Programme Neurodegeneration, Amsterdam University Medical Centers, Vrije Universiteit, Amsterdam, 1081 HZ, Netherlands
b Neurochemistry Laboratory, Departmentt of Laboratory Medicine, Amsterdam Neuroscience, Programme Neurodegeneration, Amsterdam University Medical Centers, Vrije Universiteit, Amsterdam, 1081 HZ, Netherlands
c Washington University, School of Medicine, St. Louis, MO 63110, United States
d Department of Neurology, Mayo Clinic, Rochester, MN 55905, United States
e Instituto de Investigaciones Neurológicas FLENI, Buenos Aires, Argentina
f Department of Neurology, Alzheimer’s Disease Research Center, Pittsburgh Institute for Neurodegenerative Diseases, University of Pittsburgh, Pittsburgh, PA 15213, United States
g Dementia Research Centre, UCL Queen Square Institute of Neurology, London, WC1N 3AR, United Kingdom
h Department of Neurology, Massachusetts General Hospital, Boston, MA 02114, United States
i Mayo Clinic Florida, Jacksonville, FL 32224, United States
j Institute for Stroke and Dementia Research, University Hospital, Ludwig-Maximilian-University Munich, Munich, 81377, Germany
k German Center for Neurodegenerative Diseases (DZNE), Göttingen, 37075, Germany
l Department of Pathology and Laboratory Medicine, Indiana University, Indianapolis, IN 46202, United States
m Dementia Research Institute at UCL, University College London, Institute of Neurology, London, W1T 7NF, United Kingdom
n Department of Neurodegenerative Disease, Dementia Research Centre, London, WC1N 3AR, United Kingdom
o Section for Dementia Research, Hertie Institute for Clinical Brain Research, Department of Psychiatry and Psychotherapy, University of Tübingen, Tübingen, 72076, Germany
p Neurologic Clinic and Policlinic, University Hospital, University Basel, Basel, 4031, Switzerland
q Ludwig-Maximilians-Universität München, München, D-80539, Germany
r Florey Institute, Melbourne, VIC 3052, Australia
s The University of Melbourne, Melbourne, VIC 3052, Australia
t Department of Clinical Neuroscience, Osaka City University Medical School, Osaka, 558-8585, Japan
u Neuroscience Research Australia, School of Medical Sciences, Sydney, Sydney, NSW 2052, Australia
v Barcelonaβeta Brain Research Center (BBRC), Pasqual Maragall Foundation, Barcelona, 08005, Spain
w IMIM (Hospital del Mar Medical Research Institute), Barcelona, 08003, Spain
x Servei de Neurologia, Hospital del Mar, Barcelona, 08003, Spain
y Life Science Partners, Amsterdam, 1071 DV, Netherlands
z Department of Psychiatry and Neuropsychology, School for Mental Health and Neuroscience, Alzheimer Center Limburg, Maastricht University, Maastricht, 6229 ER, Netherlands
Abstract
The grey matter of the brain develops and declines in coordinated patterns during the lifespan. Such covariation patterns of grey matter structure can be quantified as grey matter networks, which can be measured with magnetic resonance imaging. In Alzheimer’s disease, the global organization of grey matter networks becomes more random, which is captured by a decline in the small-world coefficient. Such decline in the small-world value has been robustly associated with cognitive decline across clinical stages of Alzheimer’s disease. The biological mechanisms causing this decline in small-world values remain unknown. Cerebrospinal fluid (CSF) protein biomarkers are available for studying diverse pathological mechanisms in humans and can provide insight into decline. We investigated the relationships between 10 CSF proteins and small-world coefficient in mutation carriers (N = 219) and non-carriers (N = 136) of the Dominantly Inherited Alzheimer Network Observational study. Abnormalities in Amyloid beta, Tau, synaptic (Synaptosome associated protein-25, Neurogranin) and neuronal calcium-sensor protein (Visinin-like protein-1) preceded loss of small-world coefficient by several years, while increased levels in CSF markers for inflammation (Chitinase-3-like protein 1) and axonal injury (Neurofilament light) co-occurred with decreasing small-world values. This suggests that axonal loss and inflammation play a role in structural grey matter network changes. © The Author(s) 2024.
Author Keywords
autosomal dominant Alzheimer disease; axonal damage; inflammation; neuronal injury; structural covariance network
Funding details
GE Healthcare
Deutsches Zentrum für Neurodegenerative ErkrankungenDZNE
Fleni
National Institute on AgingNIA
Korea Health Industry Development InstituteKHIDI
Foundation for Barnes-Jewish HospitalFBJH1S10OD018091–01, 1S10RR022984–01A1, K01 AG053474
Foundation for Barnes-Jewish HospitalFBJH
Japan Agency for Medical Research and DevelopmentAMED17929884, 16815631
Japan Agency for Medical Research and DevelopmentAMED
National Institutes of HealthNIH1S10RR022984–01A1, 1S10OD018091–01
National Institutes of HealthNIH
Document Type: Article
Publication Stage: Final
Source: Scopus
Leveraging Normative Personality Data and Machine Learning to Examine the Brain Structure Correlates of Obsessive-Compulsive Personality Disorder Traits
(2024) Journal of Psychopathology and Clinical Science, 133 (8), pp. 656-666. Cited 1 time.
Moreau, A.L.a , Gorelik, A.J.a , Knodt, A.b , Barch, D.M.a c d , Hariri, A.R.b , Samuel, D.B.e , Oltmanns, T.F.a , Hatoum, A.S.a , Bogdan, R.a
a Department of Psychological and Brain Sciences, Washington University in St. Louis, United States
b Department of Psychology and Neuroscience, Duke University, United States
c Department of Psychiatry, School of Medicine, Washington University in St. Louis, United States
d Department of Radiology, School of Medicine, Washington University in St. Louis, United States
e Department of Psychological Sciences, Purdue University, United States
Abstract
Brain structure correlates of obsessive-compulsive personality disorder (OCPD) remain poorly understood as limited OCPD assessment has precluded well-powered studies. Here, we tested whether machine learning (ML; elastic net regression, gradient boosting machines, support vector regression with linear and radial kernels) could estimate OCPD scores from personality data and whetherML-predicted scores are associated with indices of brain structure (cortical thickness and surface area and subcortical volumes). Among older adults (ns= 898–1,606) who completed multiple OCPD assessments, ML elastic net regression with Revised NEO Personality Inventory personality items as features best predicted Five-Factor Obsessive-Compulsive Inventory—Short Form (FFOCI-SF) scores, root-mean-squared error (RMSE)/SD= 0.66; performance generalized to a sample of college students (n= 175; RMSE/SD= 0.51). Items from all five-factor model personality traits contributed to predicted FFOCI-SF (p-FFOCI-SF) scores; conscientiousness and openness items were the most influential. In college students (n = 1,253), univariate analyses of cortical thickness, surface area, and subcortical volumes revealed only a positive association between p-FFOCISF and right superior frontal gyrus cortical thickness after adjusting for multiple testing (b = 2.21, p =.0014; all other |b|s,1.04; all other ps..009). Multivariate ML models of brain features predicting FFOCI, conscientiousness, and neuroticism performed poorly (RMSE/SDs.1.00). These data reveal that all five-factor model traits contribute to maladaptive OCPD traits and identify greater right superior frontal gyrus cortical thickness as a promising correlate of OCPD for future study. Broadly, this study highlights the utility of ML to estimate unmeasured psychopathology phenotypes in neuroimaging data sets but that our application of ML to neuroimaging may not resolve unreliable associations and small effects characteristic of univariate psychiatric neuroimaging research. © 2024 American Psychological Association
Author Keywords
brain structure; machine learning; magnetic resonance imaging; neuroimaging; obsessive-compulsive personality disorder
Funding details
201102523
Pro00019095, Pro00014717
National Institutes of HealthNIHR01DA033369, R01AG061162, K01AA030083
Document Type: Article
Publication Stage: Final
Source: Scopus