AmyloidPETNet: Classification of Amyloid Positivity in Brain PET Imaging Using End-to-End Deep Learning.

Background Visual assessment of amyloid PET scans relies on the availability of radiologist expertise, whereas quantification of amyloid burden typically involves MRI for processing and analysis, which can be computationally expensive. Purpose To develop a deep learning model to classify minimally processed brain PET scans as amyloid positive or negative, evaluate its performance on independent data sets and different tracers, and compare it with human visual reads. Materials and Methods This retrospective study used 8476 PET scans (6722 patients) obtained from late 2004 to early 2023 that were analyzed across five different data sets. A deep learning model, AmyloidPETNet, was trained on 1538 scans from 766 patients, validated on 205 scans from 95 patients, and internally tested on 184 scans from 95 patients in the Alzheimer's Disease Neuroimaging Initiative (ADNI) fluorine 18 (18F) florbetapir (FBP) data set. It was tested on ADNI scans using different tracers and scans from independent data sets. Scan amyloid positivity was based on mean cortical standardized uptake value ratio cutoffs. To compare with model performance, each scan from both the Centiloid Project and a subset of the Anti-Amyloid Treatment in Asymptomatic Alzheimer's Disease (A4) study were visually interpreted with a confidence level (low, intermediate, high) of amyloid positivity/negativity. The area under the receiver operating characteristic curve (AUC) and other performance metrics were calculated, and Cohen κ was used to measure physician-model agreement. Results The model achieved an AUC of 0.97 (95% CI: 0.95, 0.99) on test ADNI 18F-FBP scans, which generalized well to 18F-FBP scans from the Open Access Series of Imaging Studies (AUC, 0.95; 95% CI: 0.93, 0.97) and the A4 study (AUC, 0.98; 95% CI: 0.98, 0.98). Model performance was high when applied to data sets with different tracers (AUC ≥ 0.97). Other performance metrics provided converging evidence. Physician-model agreement ranged from fair (Cohen κ = 0.39; 95% CI: 0.16, 0.60) on a sample of mostly equivocal cases from the A4 study to almost perfect (Cohen κ = 0.93; 95% CI: 0.86, 1.0) on the Centiloid Project. Conclusion The developed model was capable of automatically and accurately classifying brain PET scans as amyloid positive or negative without relying on experienced readers or requiring structural MRI. Clinical trial registration no. NCT00106899 © RSNA, 2024 Supplemental material is available for this article. See also the editorial by Bryan and Forghani in this issue.

Evaluating the relationship between magnetic resonance image quality metrics and deep learning-based segmentation accuracy of brain tumors.

BACKGROUND: Magnetic resonance imaging (MRI) scans are known to suffer from a variety of acquisition artifacts as well as equipment-based variations that impact image appearance and segmentation performance. It is still unclear whether a direct relationship exists between magnetic resonance (MR) image quality metrics (IQMs) (e.g., signal-to-noise, contrast-to-noise) and segmentation accuracy. PURPOSE: Deep learning (DL) approaches have shown significant promise for automated segmentation of brain tumors on MRI but depend on the quality of input training images. We sought to evaluate the relationship between IQMs of input training images and DL-based brain tumor segmentation accuracy toward developing more generalizable models for multi-institutional data. METHODS: We trained a 3D DenseNet model on the BraTS 2020 cohorts for segmentation of tumor subregions enhancing tumor (ET), peritumoral edematous, and necrotic and non-ET on MRI; with performance quantified via a 5-fold cross-validated Dice coefficient. MRI scans were evaluated through the open-source quality control tool MRQy, to yield 13 IQMs per scan. The Pearson correlation coefficient was computed between whole tumor (WT) dice values and IQM measures in the training cohorts to identify quality measures most correlated with segmentation performance. Each selected IQM was used to group MRI scans as "better" quality (BQ) or "worse" quality (WQ), via relative thresholding. Segmentation performance was re-evaluated for the DenseNet model when (i) training on BQ MRI images with validation on WQ images, as well as (ii) training on WQ images, and validation on BQ images. Trends were further validated on independent test sets derived from the BraTS 2021 training cohorts. RESULTS: For this study, multimodal MRI scans from the BraTS 2020 training cohorts were used to train the segmentation model and validated on independent test sets derived from the BraTS 2021 cohort. Among the selected IQMs, models trained on BQ images based on inhomogeneity measurements (coefficient of variance, coefficient of joint variation, coefficient of variation of the foreground patch) and the models trained on WQ images based on noise measurement peak signal-to-noise ratio (SNR) yielded significantly improved tumor segmentation accuracy compared to their inverse models. CONCLUSIONS: Our results suggest that a significant correlation may exist between specific MR IQMs and DenseNet-based brain tumor segmentation performance. The selection of MRI scans for model training based on IQMs may yield more accurate and generalizable models in unseen validation.

Genetic and Clinical Correlates of AI-Based Brain Aging Patterns in Cognitively Unimpaired Individuals.

Importance: Brain aging elicits complex neuroanatomical changes influenced by multiple age-related pathologies. Understanding the heterogeneity of structural brain changes in aging may provide insights into preclinical stages of neurodegenerative diseases. Objective: To derive subgroups with common patterns of variation in participants without diagnosed cognitive impairment (WODCI) in a data-driven manner and relate them to genetics, biomedical measures, and cognitive decline trajectories. Design, Setting, and Participants: Data acquisition for this cohort study was performed from 1999 to 2020. Data consolidation and harmonization were conducted from July 2017 to July 2021. Age-specific subgroups of structural brain measures were modeled in 4 decade-long intervals spanning ages 45 to 85 years using a deep learning, semisupervised clustering method leveraging generative adversarial networks. Data were analyzed from July 2021 to February 2023 and were drawn from the Imaging-Based Coordinate System for Aging and Neurodegenerative Diseases (iSTAGING) international consortium. Individuals WODCI at baseline spanning ages 45 to 85 years were included, with greater than 50 000 data time points. Exposures: Individuals WODCI at baseline scan. Main Outcomes and Measures: Three subgroups, consistent across decades, were identified within the WODCI population. Associations with genetics, cardiovascular risk factors (CVRFs), amyloid ß (Aß), and future cognitive decline were assessed. Results: In a sample of 27 402 individuals (mean [SD] age, 63.0 [8.3] years; 15 146 female [55%]) WODCI, 3 subgroups were identified in contrast with the reference group: a typical aging subgroup, A1, with a specific pattern of modest atrophy and white matter hyperintensity (WMH) load, and 2 accelerated aging subgroups, A2 and A3, with characteristics that were more distinct at age 65 years and older. A2 was associated with hypertension, WMH, and vascular disease-related genetic variants and was enriched for Aß positivity (ages ≥65 years) and apolipoprotein E (APOE) ε4 carriers. A3 showed severe, widespread atrophy, moderate presence of CVRFs, and greater cognitive decline. Genetic variants associated with A1 were protective for WMH (rs7209235: mean [SD] B = -0.07 [0.01]; P value = 2.31 × 10-9) and Alzheimer disease (rs72932727: mean [SD] B = 0.1 [0.02]; P value = 6.49 × 10-9), whereas the converse was observed for A2 (rs7209235: mean [SD] B = 0.1 [0.01]; P value = 1.73 × 10-15 and rs72932727: mean [SD] B = -0.09 [0.02]; P value = 4.05 × 10-7, respectively); variants in A3 were associated with regional atrophy (rs167684: mean [SD] B = 0.08 [0.01]; P value = 7.22 × 10-12) and white matter integrity measures (rs1636250: mean [SD] B = 0.06 [0.01]; P value = 4.90 × 10-7). Conclusions and Relevance: The 3 subgroups showed distinct associations with CVRFs, genetics, and subsequent cognitive decline. These subgroups likely reflect multiple underlying neuropathologic processes and affect susceptibility to Alzheimer disease, paving pathways toward patient stratification at early asymptomatic stages and promoting precision medicine in clinical trials and health care.

Gene-SGAN: discovering disease subtypes with imaging and genetic signatures via multi-view weakly-supervised deep clustering.

Disease heterogeneity has been a critical challenge for precision diagnosis and treatment, especially in neurologic and neuropsychiatric diseases. Many diseases can display multiple distinct brain phenotypes across individuals, potentially reflecting disease subtypes that can be captured using MRI and machine learning methods. However, biological interpretability and treatment relevance are limited if the derived subtypes are not associated with genetic drivers or susceptibility factors. Herein, we describe Gene-SGAN - a multi-view, weakly-supervised deep clustering method - which dissects disease heterogeneity by jointly considering phenotypic and genetic data, thereby conferring genetic correlations to the disease subtypes and associated endophenotypic signatures. We first validate the generalizability, interpretability, and robustness of Gene-SGAN in semi-synthetic experiments. We then demonstrate its application to real multi-site datasets from 28,858 individuals, deriving subtypes of Alzheimer's disease and brain endophenotypes associated with hypertension, from MRI and single nucleotide polymorphism data. Derived brain phenotypes displayed significant differences in neuroanatomical patterns, genetic determinants, biological and clinical biomarkers, indicating potentially distinct underlying neuropathologic processes, genetic drivers, and susceptibility factors. Overall, Gene-SGAN is broadly applicable to disease subtyping and endophenotype discovery, and is herein tested on disease-related, genetically-associated neuroimaging phenotypes.

Optimal approaches to analyzing functional MRI data in glioma patients.

BACKGROUND: Resting-state fMRI is increasingly used to study the effects of gliomas on the functional organization of the brain. A variety of preprocessing techniques and functional connectivity analyses are represented in the literature. However, there so far has been no systematic comparison of how alternative methods impact observed results. NEW METHOD: We first surveyed current literature and identified alternative analytical approaches commonly used in the field. Following, we systematically compared alternative approaches to atlas registration, parcellation scheme, and choice of graph-theoretical measure as regards differentiating glioma patients (N = 59) from age-matched reference subjects (N = 163). RESULTS: Our results suggest that non-linear, as opposed to affine registration, improves structural match to an atlas, as well as measures of functional connectivity. Functionally- as opposed to anatomically-derived parcellation schemes maximized the contrast between glioma patients and reference subjects. We also demonstrate that graph-theoretic measures strongly depend on parcellation granularity, parcellation scheme, and graph density. COMPARISON WITH EXISTING METHODS AND CONCLUSIONS: Our current work primarily focuses on technical optimization of rs-fMRI analysis in glioma patients and, therefore, is fundamentally different from the bulk of papers discussing glioma-induced functional network changes. We report that the evaluation of glioma-induced alterations in the functional connectome strongly depends on analytical approaches including atlas registration, choice of parcellation scheme, and graph-theoretical measures.

Genomic loci influence patterns of structural covariance in the human brain.

Normal and pathologic neurobiological processes influence brain morphology in coordinated ways that give rise to patterns of structural covariance (PSC) across brain regions and individuals during brain aging and diseases. The genetic underpinnings of these patterns remain largely unknown. We apply a stochastic multivariate factorization method to a diverse population of 50,699 individuals (12 studies and 130 sites) and derive data-driven, multi-scale PSCs of regional brain size. PSCs were significantly correlated with 915 genomic loci in the discovery set, 617 of which are newly identified, and 72% were independently replicated. Key pathways influencing PSCs involve reelin signaling, apoptosis, neurogenesis, and appendage development, while pathways of breast cancer indicate potential interplays between brain metastasis and PSCs associated with neurodegeneration and dementia. Using support vector machines, multi-scale PSCs effectively derive imaging signatures of several brain diseases. Our results elucidate genetic and biological underpinnings that influence structural covariance patterns in the human brain.

Association of partial T2-FLAIR mismatch sign and isocitrate dehydrogenase mutation in WHO grade 4 gliomas: results from the ReSPOND consortium.

PURPOSE: While the T2-FLAIR mismatch sign is highly specific for isocitrate dehydrogenase (IDH)-mutant, 1p/19q-noncodeleted astrocytomas among lower-grade gliomas, its utility in WHO grade 4 gliomas is not well-studied. We derived the partial T2-FLAIR mismatch sign as an imaging biomarker for IDH mutation in WHO grade 4 gliomas. METHODS: Preoperative MRI scans of adult WHO grade 4 glioma patients (n = 2165) from the multi-institutional ReSPOND (Radiomics Signatures for PrecisiON Diagnostics) consortium were analyzed. Diagnostic performance of the partial T2-FLAIR mismatch sign was evaluated. Subset analyses were performed to assess associations of imaging markers with overall survival (OS). RESULTS: One hundred twenty-one (5.6%) of 2165 grade 4 gliomas were IDH-mutant. Partial T2-FLAIR mismatch was present in 40 (1.8%) cases, 32 of which were IDH-mutant, yielding 26.4% sensitivity, 99.6% specificity, 80.0% positive predictive value, and 95.8% negative predictive value. Multivariate logistic regression demonstrated IDH mutation was significantly associated with partial T2-FLAIR mismatch (odds ratio [OR] 5.715, 95% CI [1.896, 17.221], p = 0.002), younger age (OR 0.911 [0.895, 0.927], p < 0.001), tumor centered in frontal lobe (OR 3.842, [2.361, 6.251], p < 0.001), absence of multicentricity (OR 0.173, [0.049, 0.612], p = 0.007), and presence of cystic (OR 6.596, [3.023, 14.391], p < 0.001) or non-enhancing solid components (OR 6.069, [3.371, 10.928], p < 0.001). Multivariate Cox analysis demonstrated cystic components (p = 0.024) and non-enhancing solid components (p = 0.003) were associated with longer OS, while older age (p < 0.001), frontal lobe center (p = 0.008), multifocality (p < 0.001), and multicentricity (p < 0.001) were associated with shorter OS. CONCLUSION: Partial T2-FLAIR mismatch sign is highly specific for IDH mutation in WHO grade 4 gliomas.

MRI-based classification of IDH mutation and 1p/19q codeletion status of gliomas using a 2.5D hybrid multi-task convolutional neural network.

Background: IDH mutation and 1p/19q codeletion status are important prognostic markers for glioma that are currently determined using invasive procedures. Our goal was to develop artificial intelligence-based methods to noninvasively determine molecular alterations from MRI. Methods: Pre-operative MRI scans of 2648 glioma patients were collected from Washington University School of Medicine (WUSM; n = 835) and publicly available Brain Tumor Segmentation (BraTS; n = 378), LGG 1p/19q (n = 159), Ivy Glioblastoma Atlas Project (Ivy GAP; n = 41), The Cancer Genome Atlas (TCGA; n = 461), and the Erasmus Glioma Database (EGD; n = 774) datasets. A 2.5D hybrid convolutional neural network was proposed to simultaneously localize glioma and classify its molecular status by leveraging MRI imaging features and prior knowledge features from clinical records and tumor location. The models were trained on 223 and 348 cases for IDH and 1p/19q tasks, respectively, and tested on one internal (TCGA) and two external (WUSM and EGD) test sets. Results: For IDH, the best-performing model achieved areas under the receiver operating characteristic (AUROC) of 0.925, 0.874, 0.933 and areas under the precision-recall curves (AUPRC) of 0.899, 0.702, 0.853 on the internal, WUSM, and EGD test sets, respectively. For 1p/19q, the best model achieved AUROCs of 0.782, 0.754, 0.842, and AUPRCs of 0.588, 0.713, 0.782, on those three data-splits, respectively. Conclusions: The high accuracy of the model on unseen data showcases its generalization capabilities and suggests its potential to perform "virtual biopsy" for tailoring treatment planning and overall clinical management of gliomas.

Structural gray matter alterations in glioblastoma and high-grade glioma-A potential biomarker of survival.

Background: Patients with glioblastoma (GBM) and high-grade glioma (HGG, World Health Organization [WHO] grade IV glioma) have a poor prognosis. Consequently, there is an unmet clinical need for accessible and noninvasively acquired predictive biomarkers of overall survival in patients. This study evaluated morphological changes in the brain separated from the tumor invasion site (ie, contralateral hemisphere). Specifically, we examined the prognostic value of widespread alterations of cortical thickness (CT) in GBM/HGG patients. Methods: We used FreeSurfer, applied with high-resolution T1-weighted MRI, to examine CT, evaluated prior to standard treatment with surgery and chemoradiation in patients (GBM/HGG, N = 162, mean age 61.3 years) and 127 healthy controls (HC; 61.9 years mean age). We then compared CT in patients to HC and studied patients' associated changes in CT as a potential biomarker of overall survival. Results: Compared to HC cases, patients had thinner gray matter in the contralesional hemisphere at the time of tumor diagnosis. patients had significant cortical thinning in parietal, temporal, and occipital lobes. Fourteen cortical parcels showed reduced CT, whereas in 5, it was thicker in patients' cases. Notably, CT in the contralesional hemisphere, various lobes, and parcels was predictive of overall survival. A machine learning classification algorithm showed that CT could differentiate short- and long-term survival patients with an accuracy of 83.3%. Conclusions: These findings identify previously unnoticed structural changes in the cortex located in the hemisphere contralateral to the primary tumor mass. Observed changes in CT may have prognostic value, which could influence care and treatment planning for individual patients.

Integrative Imaging Informatics for Cancer Research: Workflow Automation for Neuro-Oncology (I3CR-WANO).

PURPOSE: Efforts to use growing volumes of clinical imaging data to generate tumor evaluations continue to require significant manual data wrangling, owing to data heterogeneity. Here, we propose an artificial intelligence-based solution for the aggregation and processing of multisequence neuro-oncology MRI data to extract quantitative tumor measurements. MATERIALS AND METHODS: Our end-to-end framework (1) classifies MRI sequences using an ensemble classifier, (2) preprocesses the data in a reproducible manner, (3) delineates tumor tissue subtypes using convolutional neural networks, and (4) extracts diverse radiomic features. Moreover, it is robust to missing sequences and adopts an expert-in-the-loop approach in which the segmentation results may be manually refined by radiologists. After the implementation of the framework in Docker containers, it was applied to two retrospective glioma data sets collected from the Washington University School of Medicine (WUSM; n = 384) and The University of Texas MD Anderson Cancer Center (MDA; n = 30), comprising preoperative MRI scans from patients with pathologically confirmed gliomas. RESULTS: The scan-type classifier yielded an accuracy of >99%, correctly identifying sequences from 380 of 384 and 30 of 30 sessions from the WUSM and MDA data sets, respectively. Segmentation performance was quantified using the Dice Similarity Coefficient between the predicted and expert-refined tumor masks. The mean Dice scores were 0.882 (±0.244) and 0.977 (±0.04) for whole-tumor segmentation for WUSM and MDA, respectively. CONCLUSION: This streamlined framework automatically curated, processed, and segmented raw MRI data of patients with varying grades of gliomas, enabling the curation of large-scale neuro-oncology data sets and demonstrating high potential for integration as an assistive tool in clinical practice.

Gene-SGAN: a method for discovering disease subtypes with imaging and genetic signatures via multi-view weakly-supervised deep clustering.

Disease heterogeneity has been a critical challenge for precision diagnosis and treatment, especially in neurologic and neuropsychiatric diseases. Many diseases can display multiple distinct brain phenotypes across individuals, potentially reflecting disease subtypes that can be captured using MRI and machine learning methods. However, biological interpretability and treatment relevance are limited if the derived subtypes are not associated with genetic drivers or susceptibility factors. Herein, we describe Gene-SGAN - a multi-view, weakly-supervised deep clustering method - which dissects disease heterogeneity by jointly considering phenotypic and genetic data, thereby conferring genetic correlations to the disease subtypes and associated endophenotypic signatures. We first validate the generalizability, interpretability, and robustness of Gene-SGAN in semi-synthetic experiments. We then demonstrate its application to real multi-site datasets from 28,858 individuals, deriving subtypes of Alzheimer's disease and brain endophenotypes associated with hypertension, from MRI and SNP data. Derived brain phenotypes displayed significant differences in neuroanatomical patterns, genetic determinants, biological and clinical biomarkers, indicating potentially distinct underlying neuropathologic processes, genetic drivers, and susceptibility factors. Overall, Gene-SGAN is broadly applicable to disease subtyping and endophenotype discovery, and is herein tested on disease-related, genetically-driven neuroimaging phenotypes.

Five dominant dimensions of brain aging are identified via deep learning: associations with clinical, lifestyle, and genetic measures.

Brain aging is a complex process influenced by various lifestyle, environmental, and genetic factors, as well as by age-related and often co-existing pathologies. MRI and, more recently, AI methods have been instrumental in understanding the neuroanatomical changes that occur during aging in large and diverse populations. However, the multiplicity and mutual overlap of both pathologic processes and affected brain regions make it difficult to precisely characterize the underlying neurodegenerative profile of an individual from an MRI scan. Herein, we leverage a state-of-the art deep representation learning method, Surreal-GAN, and present both methodological advances and extensive experimental results that allow us to elucidate the heterogeneity of brain aging in a large and diverse cohort of 49,482 individuals from 11 studies. Five dominant patterns of neurodegeneration were identified and quantified for each individual by their respective (herein referred to as) R-indices. Significant associations between R-indices and distinct biomedical, lifestyle, and genetic factors provide insights into the etiology of observed variances. Furthermore, baseline R-indices showed predictive value for disease progression and mortality. These five R-indices contribute to MRI-based precision diagnostics, prognostication, and may inform stratification into clinical trials.

Covariance-based vs. correlation-based functional connectivity dissociates healthy aging from Alzheimer disease.

Prior studies of aging and Alzheimer disease have evaluated resting state functional connectivity (FC) using either seed-based correlation (SBC) or independent component analysis (ICA), with a focus on particular functional systems. SBC and ICA both are insensitive to differences in signal amplitude. At the same time, accumulating evidence indicates that the amplitude of spontaneous BOLD signal fluctuations is physiologically meaningful. We systematically compared covariance-based FC, which is sensitive to amplitude, vs. correlation-based FC, which is not, in affected individuals and controls drawn from two cohorts of participants including autosomal dominant Alzheimer disease (ADAD), late onset Alzheimer disease (LOAD), and age-matched controls. Functional connectivity was computed over 222 regions of interest and group differences were evaluated in terms of components projected onto a space of lower dimension. Our principal observations are: (1) Aging is associated with global loss of resting state fMRI signal amplitude that is approximately uniform across resting state networks. (2) Thus, covariance FC measures decrease with age whereas correlation FC is relatively preserved in healthy aging. (3) In contrast, symptomatic ADAD and LOAD both lead to loss of spontaneous activity amplitude as well as severely degraded correlation structure. These results demonstrate a double dissociation between age vs. Alzheimer disease and the amplitude vs. correlation structure of resting state BOLD signals. Modeling results suggest that the AD-associated loss of correlation structure is attributable to a relative increase in the fraction of locally restricted as opposed to widely shared variance.

CSF Tau phosphorylation at Thr205 is associated with loss of white matter integrity in autosomal dominant Alzheimer disease.

BACKGROUND: Hyperphosphorylation of tau leads to conformational changes that destabilize microtubules and hinder axonal transport in Alzheimer's disease (AD). However, it remains unknown whether white matter (WM) decline due to AD is associated with specific Tau phosphorylation site(s). METHODS: In autosomal dominant AD (ADAD) mutation carriers (MC) and non-carriers (NC) we compared cerebrospinal fluid (CSF) phosphorylation at tau sites (pT217, pT181, pS202, and pT205) and total tau with WM measures, as derived from diffusion tensor imaging (DTI), and cognition. A WM composite metric, derived from a principal component analysis, was used to identify spatial decline seen in ADAD. RESULTS: The WM composite explained over 70% of the variance in MC. WM regions that strongly contributed to the spatial topography were located in callosal and cingulate regions. Loss of integrity within the WM composite was strongly associated with AD progression in MC as defined by the estimated years to onset (EYO) and cognitive decline. A linear regression demonstrated that amyloid, gray matter atrophy and phosphorylation at CSF tau site pT205 each uniquely explained a reduction in the WM composite within MC that was independent of vascular changes (white matter hyperintensities), and age. Hyperphosphorylation of CSF tau at other sites and total tau did not significantly predict WM composite loss. CONCLUSIONS: We identified a site-specific relationship between CSF phosphorylated tau and WM decline within MC. The presence of both amyloid deposition and Tau phosphorylation at pT205 were associated with WM composite loss. These findings highlight a primary AD-specific mechanism for WM dysfunction that is tightly coupled to symptom manifestation and cognitive decline.

Accelerating Prediction of Malignant Cerebral Edema After Ischemic Stroke with Automated Image Analysis and Explainable Neural Networks.

BACKGROUND: Malignant cerebral edema is a devastating complication of stroke, resulting in deterioration and death if hemicraniectomy is not performed prior to herniation. Current approaches for predicting this relatively rare complication often require advanced imaging and still suffer from suboptimal performance. We performed a pilot study to evaluate whether neural networks incorporating data extracted from routine computed tomography (CT) imaging could enhance prediction of edema in a large diverse stroke cohort. METHODS: An automated imaging pipeline retrospectively extracted volumetric data, including cerebrospinal fluid (CSF) volumes and the hemispheric CSF volume ratio, from baseline and 24 h CT scans performed in participants of an international stroke cohort study. Fully connected and long short-term memory (LSTM) neural networks were trained using serial clinical and imaging data to predict those who would require hemicraniectomy or die with midline shift. The performance of these models was tested, in comparison with regression models and the Enhanced Detection of Edema in Malignant Anterior Circulation Stroke (EDEMA) score, using cross-validation to construct precision-recall curves. RESULTS: Twenty of 598 patients developed malignant edema (12 required surgery, 8 died). The regression model provided 95% recall but only 32% precision (area under the precision-recall curve [AUPRC] 0.74), similar to the EDEMA score (precision 28%, AUPRC 0.66). The fully connected network did not perform better (precision 33%, AUPRC 0.71), but the LSTM model provided 100% recall and 87% precision (AUPRC 0.97) in the overall cohort and the subgroup with a National Institutes of Health Stroke Scale (NIHSS) score ≥ 8 (p = 0.0001 vs. regression and fully connected models). Features providing the most predictive importance were the hemispheric CSF ratio and NIHSS score measured at 24 h. CONCLUSIONS: An LSTM neural network incorporating volumetric data extracted from routine CT scans identified all cases of malignant cerebral edema by 24 h after stroke, with significantly fewer false positives than a fully connected neural network, regression model, and the validated EDEMA score. This preliminary work requires prospective validation but provides proof of principle that a deep learning framework could assist in selecting patients for surgery prior to deterioration.

Different rates of cognitive decline in autosomal dominant and late-onset Alzheimer disease.

As prevention trials advance with autosomal dominant Alzheimer disease (ADAD) participants, understanding the similarities and differences between ADAD and "sporadic" late-onset AD (LOAD) is critical to determine generalizability of findings between these cohorts. Cognitive trajectories of ADAD mutation carriers (MCs) and autopsy-confirmed LOAD individuals were compared to address this question. Longitudinal rates of change on cognitive measures were compared in ADAD MCs (n = 310) and autopsy-confirmed LOAD participants (n = 163) before and after symptom onset (estimated/observed). LOAD participants declined more rapidly in the presymptomatic (preclinical) period and performed more poorly at symptom onset than ADAD participants on a cognitive composite. After symptom onset, however, the younger ADAD MCs declined more rapidly. The similar but not identical cognitive trajectories (declining but at different rates) for ADAD and LOAD suggest common AD pathologies but with some differences.

Developing methods to detect and diagnose chronic traumatic encephalopathy during life: rationale, design, and methodology for the DIAGNOSE CTE Research Project.

BACKGROUND: Chronic traumatic encephalopathy (CTE) is a neurodegenerative disease that has been neuropathologically diagnosed in brain donors exposed to repetitive head impacts, including boxers and American football, soccer, ice hockey, and rugby players. CTE cannot yet be diagnosed during life. In December 2015, the National Institute of Neurological Disorders and Stroke awarded a seven-year grant (U01NS093334) to fund the "Diagnostics, Imaging, and Genetics Network for the Objective Study and Evaluation of Chronic Traumatic Encephalopathy (DIAGNOSE CTE) Research Project." The objectives of this multicenter project are to: develop in vivo fluid and neuroimaging biomarkers for CTE; characterize its clinical presentation; refine and validate clinical research diagnostic criteria (i.e., traumatic encephalopathy syndrome [TES]); examine repetitive head impact exposure, genetic, and other risk factors; and provide shared resources of anonymized data and biological samples to the research community. In this paper, we provide a detailed overview of the rationale, design, and methods for the DIAGNOSE CTE Research Project. METHODS: The targeted sample and sample size was 240 male participants, ages 45-74, including 120 former professional football players, 60 former collegiate football players, and 60 asymptomatic participants without a history of head trauma or participation in organized contact sports. Participants were evaluated at one of four U.S. sites and underwent the following baseline procedures: neurological and neuropsychological examinations; tau and amyloid positron emission tomography; magnetic resonance imaging and spectroscopy; lumbar puncture; blood and saliva collection; and standardized self-report measures of neuropsychiatric, cognitive, and daily functioning. Study partners completed similar informant-report measures. Follow-up evaluations were intended to be in-person and at 3 years post-baseline. Multidisciplinary diagnostic consensus conferences are held, and the reliability and validity of TES diagnostic criteria are examined. RESULTS: Participant enrollment and all baseline evaluations were completed in February 2020. Three-year follow-up evaluations began in October 2019. However, in-person evaluation ceased with the COVID-19 pandemic, and resumed as remote, 4-year follow-up evaluations (including telephone-, online-, and videoconference-based cognitive, neuropsychiatric, and neurologic examinations, as well as in-home blood draw) in February 2021. CONCLUSIONS: Findings from the DIAGNOSE CTE Research Project should facilitate detection and diagnosis of CTE during life, and thereby accelerate research on risk factors, mechanisms, epidemiology, treatment, and prevention of CTE. TRIAL REGISTRATION: NCT02798185.

The Stroke Neuro-Imaging Phenotype Repository: An Open Data Science Platform for Stroke Research.

Stroke is one of the leading causes of death and disability worldwide. Reducing this disease burden through drug discovery and evaluation of stroke patient outcomes requires broader characterization of stroke pathophysiology, yet the underlying biologic and genetic factors contributing to outcomes are largely unknown. Remedying this critical knowledge gap requires deeper phenotyping, including large-scale integration of demographic, clinical, genomic, and imaging features. Such big data approaches will be facilitated by developing and running processing pipelines to extract stroke-related phenotypes at large scale. Millions of stroke patients undergo routine brain imaging each year, capturing a rich set of data on stroke-related injury and outcomes. The Stroke Neuroimaging Phenotype Repository (SNIPR) was developed as a multi-center centralized imaging repository of clinical computed tomography (CT) and magnetic resonance imaging (MRI) scans from stroke patients worldwide, based on the open source XNAT imaging informatics platform. The aims of this repository are to: (i) store, manage, process, and facilitate sharing of high-value stroke imaging data sets, (ii) implement containerized automated computational methods to extract image characteristics and disease-specific features from contributed images, (iii) facilitate integration of imaging, genomic, and clinical data to perform large-scale analysis of complications after stroke; and (iv) develop SNIPR as a collaborative platform aimed at both data scientists and clinical investigators. Currently, SNIPR hosts research projects encompassing ischemic and hemorrhagic stroke, with data from 2,246 subjects, and 6,149 imaging sessions from Washington University's clinical image archive as well as contributions from collaborators in different countries, including Finland, Poland, and Spain. Moreover, we have extended the XNAT data model to include relevant clinical features, including subject demographics, stroke severity (NIH Stroke Scale), stroke subtype (using TOAST classification), and outcome [modified Rankin Scale (mRS)]. Image processing pipelines are deployed on SNIPR using containerized modules, which facilitate replicability at a large scale. The first such pipeline identifies axial brain CT scans from DICOM header data and image data using a meta deep learning scan classifier, registers serial scans to an atlas, segments tissue compartments, and calculates CSF volume. The resulting volume can be used to quantify the progression of cerebral edema after ischemic stroke. SNIPR thus enables the development and validation of pipelines to automatically extract imaging phenotypes and couple them with clinical data with the overarching aim of enabling a broad understanding of stroke progression and outcomes.

Machine Learning Analytics of Resting-State Functional Connectivity Predicts Survival Outcomes of Glioblastoma Multiforme Patients.

Glioblastoma multiforme (GBM) is the most frequently occurring brain malignancy. Due to its poor prognosis with currently available treatments, there is a pressing need for easily accessible, non-invasive techniques to help inform pre-treatment planning, patient counseling, and improve outcomes. In this study we determined the feasibility of resting-state functional connectivity (rsFC) to classify GBM patients into short-term and long-term survival groups with respect to reported median survival (14.6 months). We used a support vector machine with rsFC between regions of interest as predictive features. We employed a novel hybrid feature selection method whereby features were first filtered using correlations between rsFC and OS, and then using the established method of recursive feature elimination (RFE) to select the optimal feature subset. Leave-one-subject-out cross-validation evaluated the performance of models. Classification between short- and long-term survival accuracy was 71.9%. Sensitivity and specificity were 77.1 and 65.5%, respectively. The area under the receiver operating characteristic curve was 0.752 (95% CI, 0.62-0.88). These findings suggest that highly specific features of rsFC may predict GBM survival. Taken together, the findings of this study support that resting-state fMRI and machine learning analytics could enable a radiomic biomarker for GBM, augmenting care and planning for individual patients.

Resting-State Functional Connectivity Disruption as a Pathological Biomarker in Autosomal Dominant Alzheimer Disease.

Aim: Identify a global resting-state functional connectivity (gFC) signature in mutation carriers (MC) from the Dominantly Inherited Alzheimer Network (DIAN). Assess the gFC with regard to amyloid (A), tau (T), and neurodegeneration (N) biomarkers, and estimated years to symptom onset (EYO). Introduction: Cross-sectional measures were assessed in MC (n = 171) and mutation noncarrier (NC) (n = 70) participants. A functional connectivity (FC) matrix that encompassed multiple resting-state networks was computed for each participant. Methods: A global FC was compiled as a single index indicating FC strength. The gFC signature was modeled as a nonlinear function of EYO. The gFC was linearly associated with other biomarkers used for assessing the AT(N) framework, including cerebrospinal fluid (CSF), positron emission tomography (PET) molecular biomarkers, and structural magnetic resonance imaging. Results: The gFC was reduced in MC compared with NC participants. When MC participants were differentiated by clinical dementia rating (CDR), the gFC was significantly decreased in MC CDR >0 (demented) compared with either MC CDR 0 (cognitively normal) or NC participants. The gFC varied nonlinearly with EYO and initially decreased at EYO = -24 years, followed by a stable period followed by a further decline near EYO = 0 years. Irrespective of EYO, a lower gFC associated with values of amyloid PET, CSF Aß1-42, CSF p-tau, CSF t-tau, 18F-fluorodeoxyglucose, and hippocampal volume. Conclusions: The gFC correlated with biomarkers used for defining the AT(N) framework. A biphasic change in the gFC suggested early changes associated with CSF amyloid and later changes associated with hippocampal volume.