Multimodal IVA fusion for detection of linked neuroimaging biomarkers

With the increasing availability of large-scale multimodal neuroimaging datasets, it is necessary to develop data fusion methods which can extract cross-modal features. A general framework, multidataset independent subspace analysis (MISA), has been developed to encompass multiple blind source separation approaches and identify linked cross-modal components in multiple datasets. In this work we utilized the multimodal independent vector analysis model in MISA to directly identify meaningful linked features across three neuroimaging modalities --- structural magnetic resonance imaging (MRI), resting state functional MRI and diffusion MRI --- in two large independent datasets, one comprising of healthy subjects and the other including patients with schizophrenia. Results show several linked subject profiles (the sources/components) that capture age-associated reductions, schizophrenia-related biomarkers, sex effects, and cognitive performance.

Neuroimaging cerebrovascular biomarkers in Parkinson’s disease

Background and Purpose The cardiovascular risk in Parkinson’s disease (PD) remains uncertain and controversial. Some studies suggest PD patients present an increased risk of cerebrovascular disease. We aimed to study the prevalence of neuroimaging cerebrovascular biomarkers in PD patients compared to controls, using an accurate and complete magnetic resonance (MR) imaging evaluation. Material and Methods Neuroimaging sub-study within a larger cross-sectional case–control study. An enriched subgroup of PD patients (≤10 years since diagnosis) with at least a moderate cardiovascular mortality risk based on a Systematic COronary Risk Evaluation (SCORE) was compared to community-based controls regarding neuroimaging biomarkers. Patients underwent a high-resolution T1-weighted MR imaging sequence at 3.0 T to visualize neuromelanin. A 3D SWI FFE, sagittal 3D T1-weighted, axial FLAIR and diffusion-weighted image sequences were obtained. Results The study included 47 patients, 24 with PD and 23 controls. PD patients presented a reduced area and signal intensity of the substantia nigra and locus coeruleus on neuromelanin-sensitive MR. The median SCORE was 5% in both groups. No significant differences regarding white matter hyperintensities (OR 4.84, 95% CI 0.50, 47.06), lacunes (OR 0.43, 95% CI 0.07, 2.63), microbleeds (OR 0.64, 95% CI 0.13, 3.26), or infarcts (0.95, 95% CI 0.12, 7.41) was found. The frequency of these neuroimaging biomarkers was very low in both groups. Conclusion The present study does not support an increased prevalence of neuroimaging cerebrovascular biomarkers in PD patients.

Alzheimer’s Disease-Related Metabolic Pattern in Diverse Forms of Neurodegenerative Diseases

Dementia is broadly characterized by cognitive and psychological dysfunction that significantly impairs daily functioning. Dementia has many causes including Alzheimer’s disease (AD), dementia with Lewy bodies (DLB), and frontotemporal lobar degeneration (FTLD). Detection and differential diagnosis in the early stages of dementia remains challenging. Fueled by AD Neuroimaging Initiatives (ADNI) (Data used in preparation of this article were obtained from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database. As such, the investigators within ADNI contributed to the design and implementation of ADNI and/or provided data but did not participate in analysis or writing of this report.), a number of neuroimaging biomarkers for AD have been proposed, yet it remains to be seen whether these markers are also sensitive to other types of dementia. We assessed AD-related metabolic patterns in 27 patients with diverse forms of dementia (five had probable/possible AD while others had atypical cases) and 20 non-demented individuals. All participants had positron emission tomography (PET) scans on file. We used a pre-trained machine learning-based AD designation (MAD) framework to investigate the AD-related metabolic pattern among the participants under study. The MAD algorithm showed a sensitivity of 0.67 and specificity of 0.90 for distinguishing dementia patients from non-dementia participants. A total of 18/27 dementia patients and 2/20 non-dementia patients were identified as having AD-like patterns of metabolism. These results highlight that many underlying causes of dementia have similar hypometabolic pattern as AD and this similarity is an interesting avenue for future research.

Transfer Learning for Alzheimer’s Disease through Neuroimaging Biomarkers: A Systematic Review

Alzheimer’s disease (AD) is a remarkable challenge for healthcare in the 21st century. Since 2017, deep learning models with transfer learning approaches have been gaining recognition in AD detection, and progression prediction by using neuroimaging biomarkers. This paper presents a systematic review of the current state of early AD detection by using deep learning models with transfer learning and neuroimaging biomarkers. Five databases were used and the results before screening report 215 studies published between 2010 and 2020. After screening, 13 studies met the inclusion criteria. We noted that the maximum accuracy achieved to date for AD classification is 98.20% by using the combination of 3D convolutional networks and local transfer learning, and that for the prognostic prediction of AD is 87.78% by using pre-trained 3D convolutional network-based architectures. The results show that transfer learning helps researchers in developing a more accurate system for the early diagnosis of AD. However, there is a need to consider some points in future research, such as improving the accuracy of the prognostic prediction of AD, exploring additional biomarkers such as tau-PET and amyloid-PET to understand highly discriminative feature representation to separate similar brain patterns, managing the size of the datasets due to the limited availability.

Is comprehensiveness critical? Comparing short and long format cognitive assessments in preclinical Alzheimer disease

Background Comprehensive testing of cognitive functioning is standard practice in studies of Alzheimer disease (AD). Short-form tests like the Montreal Cognitive Assessment (MoCA) use a “sampling” of measures, administering key items in a shortened format to efficiently assess cognition while reducing time requirements, participant burden, and administrative costs. We compared the MoCA to a commonly used long-form cognitive battery in predicting AD symptom onset and sensitivity to AD neuroimaging biomarkers. Methods Survival, area under the receiver operating characteristic (ROC) curve (AUC), and multiple regression analyses compared the MoCA and long-form measures in predicting time to symptom onset in cognitively normal older adults (n = 6230) from the National Alzheimer’s Coordinating Center (NACC) cohort who had, on average, 2.3 ± 1.2 annual assessments. Multiple regression models in a separate sample (n = 416) from the Charles F. and Joanne Knight Alzheimer Disease Research Center (Knight ADRC) compared the sensitivity of the MoCA and long-form measures to neuroimaging biomarkers including amyloid PET, tau PET, and cortical thickness. Results Hazard ratios suggested that both the MoCA and the long-form measures are similarly and modestly efficacious in predicting symptomatic conversion, although model comparison analyses indicated that the long-form measures slightly outperformed the MoCA (HRs > 1.57). AUC analyses indicated no difference between the measures in predicting conversion (DeLong’s test, Z = 1.48, p = 0.13). Sensitivity to AD neuroimaging biomarkers was similar for the two measures though there were only modest associations with tau PET (rs = − 0.13, ps < 0.02) and cortical thickness in cognitively normal participants (rs = 0.15–0.16, ps < 0.007). Conclusions Both test formats showed weak associations with symptom onset, AUC analyses indicated low diagnostic accuracy, and biomarker correlations were modest in cognitively normal participants. Alternative assessment approaches are needed to improve how clinicians and researchers monitor cognitive changes and disease progression prior to symptom onset.

An interpretable multiparametric radiomics model for the diagnosis of schizophrenia using magnetic resonance imaging of the corpus callosum

There is a growing need to develop novel strategies for the diagnosis of schizophrenia using neuroimaging biomarkers. We investigated the robustness of the diagnostic model for schizophrenia using radiomic features from T1-weighted and diffusion tensor images of the corpus callosum (CC). A total of 165 participants [86 schizophrenia and 79 healthy controls (HCs)] were allocated to training (N = 115) and test (N = 50) sets. Radiomic features of the CC subregions were extracted from T1-weighted, apparent diffusion coefficient (ADC), and fractional anisotropy (FA) images (N = 1605). Following feature selection, various combinations of classifiers were trained, and Bayesian optimization was adopted in the best performing classifier. Discrimination, calibration, and clinical utility of the model were assessed. An online calculator was constructed to offer the probability of having schizophrenia. SHapley Additive exPlanations (SHAP) was applied to explore the interpretability of the model. We identified 30 radiomic features to differentiate participants with schizophrenia from HCs. The Bayesian optimized model achieved the highest performance, with an area under the curve (AUC), accuracy, sensitivity, and specificity of 0.89 (95% confidence interval: 0.81–0.98), 80.0, 83.3, and 76.9%, respectively, in the test set. The final model offers clinical probability in an online calculator. The model explanation by SHAP suggested that second-order features from the posterior CC were highly associated with the risk of schizophrenia. The multiparametric radiomics model focusing on the CC shows its robustness for the diagnosis of schizophrenia. Radiomic features could be a potential source of biomarkers that support the biomarker-based diagnosis of schizophrenia and improve the understanding of its neurobiology.