An interactive atlas of three-dimensional syndromic facial morphology.

Craniofacial phenotyping is critical for both syndrome delineation and diagnosis because craniofacial abnormalities occur in 30% of characterized genetic syndromes. Clinical reports, textbooks, and available software tools typically provide two-dimensional, static images and illustrations of the characteristic phenotypes of genetic syndromes. In this work, we provide an interactive web application that provides three-dimensional, dynamic visualizations for the characteristic craniofacial effects of 95 syndromes. Users can visualize syndrome facial appearance estimates quantified from data and easily compare craniofacial phenotypes of different syndromes. Our application also provides a map of morphological similarity between a target syndrome and other syndromes. Finally, users can upload 3D facial scans of individuals and compare them to our syndrome atlas estimates. In summary, we provide an interactive reference for the craniofacial phenotypes of syndromes that allows for precise, individual-specific comparisons of dysmorphology.

Providing clinical context to the spatio-temporal analysis of 4D CT perfusion to predict acute ischemic stroke lesion outcomes.

Acute ischemic stroke is a leading cause of mortality and morbidity worldwide. Timely identification of the extent of a stroke is crucial for effective treatment, whereas spatio-temporal (4D) Computed Tomography Perfusion (CTP) imaging is playing a critical role in this process. Recently, the first deep learning-based methods that leverage the full spatio-temporal nature of perfusion imaging for predicting stroke lesion outcomes have been proposed. However, clinical information is typically not integrated into the learning process, which may be helpful to improve the tissue outcome prediction given the known influence of various factors (i.e., physiological, demographic, and treatment factors) on lesion growth. Cross-attention, a multimodal fusion strategy, has been successfully used to combine information from multiple sources, but it has yet to be applied to stroke lesion outcome prediction. Therefore, this work aimed to develop and evaluate a novel multimodal and spatio-temporal deep learning model that utilizes cross-attention to combine information from 4D CTP and clinical metadata simultaneously to predict stroke lesion outcomes. The proposed model was evaluated using a dataset of 70 acute ischemic stroke patients, demonstrating significantly improved volume estimates (mean error = 19 ml) compared to a baseline unimodal approach (mean error = 35 ml, p< 0.05). The proposed model allows generating attention maps and counterfactual outcome scenarios to investigate the relevance of clinical variables in predicting stroke lesion outcomes at a patient level, helping to provide a better understanding of the model's decision-making process.

Machine learning using multimodal clinical, electroencephalographic, and magnetic resonance imaging data can predict incident depression in adults with epilepsy: A pilot study.

OBJECTIVE: This study was undertaken to develop a multimodal machine learning (ML) approach for predicting incident depression in adults with epilepsy. METHODS: We randomly selected 200 patients from the Calgary Comprehensive Epilepsy Program registry and linked their registry-based clinical data to their first-available clinical electroencephalogram (EEG) and magnetic resonance imaging (MRI) study. We excluded patients with a clinical or Neurological Disorders Depression Inventory for Epilepsy (NDDI-E)-based diagnosis of major depression at baseline. The NDDI-E was used to detect incident depression over a median of 2.4 years of follow-up (interquartile range [IQR] = 1.5-3.3 years). A ReliefF algorithm was applied to clinical as well as quantitative EEG and MRI parameters for feature selection. Six ML algorithms were trained and tested using stratified threefold cross-validation. Multiple metrics were used to assess model performances. RESULTS: Of 200 patients, 150 had EEG and MRI data of sufficient quality for ML, of whom 59 were excluded due to prevalent depression. Therefore, 91 patients (41 women) were included, with a median age of 29 (IQR = 22-44) years. A total of 42 features were selected by ReliefF, none of which was a quantitative MRI or EEG variable. All models had a sensitivity > 80%, and five of six had an F1 score ≥ .72. A multilayer perceptron model had the highest F1 score (median = .74, IQR = .71-.78) and sensitivity (84.3%). Median area under the receiver operating characteristic curve and normalized Matthews correlation coefficient were .70 (IQR = .64-.78) and .57 (IQR = .50-.65), respectively. SIGNIFICANCE: Multimodal ML using baseline features can predict incident depression in this population. Our pilot models demonstrated high accuracy for depression prediction. However, overall performance and calibration can be improved. This model has promise for identifying those at risk for incident depression during follow-up, although efforts to refine it in larger populations along with external validation are required.

Expert elicitation of risk factors for progression to dementia in individuals with mild cognitive impairment.

INTRODUCTION: This study assesses experts' beliefs about important predictors of developing dementia in persons with mild cognitive impairment (MCI). METHODS: Structured expert elicitation, a methodology to quantify expert knowledge, was used to elicit the most important risk factors for developing dementia. We recruited 11 experts (6 neurologists, 3 geriatricians, and 2 psychiatrists). Ten experts fully participated in introductory meetings, two rounds of surveys, and discussion meetings. The data from these ten experts were utilized for this study. RESULTS: The expert elicitation identified age, CSF analysis, fluorodeoxyglucose-positron emission tomography (FDG-PET) findings, hippocampal atrophy, MoCA (or MMSE) score, parkinsonism, apathy, psychosis, informant report of cognitive symptoms, and global atrophy as the ten most important predictors of progressing to dementia in persons with MCI. DISCUSSION: Several dementia predictors are not routinely collected in existing registries, observational studies, or usual care. This might partially explain the low uptake of existing published dementia risk scores in clinical practice.

Multimodal biological brain age prediction using magnetic resonance imaging and angiography with the identification of predictive regions.

Biological brain age predicted using machine learning models based on high-resolution imaging data has been suggested as a potential biomarker for neurological and cerebrovascular diseases. In this work, we aimed to develop deep learning models to predict the biological brain age using structural magnetic resonance imaging and angiography datasets from a large database of 2074 adults (21-81 years). Since different imaging modalities can provide complementary information, combining them might allow to identify more complex aging patterns, with angiography data, for instance, showing vascular aging effects complementary to the atrophic brain tissue changes seen in T1-weighted MRI sequences. We used saliency maps to investigate the contribution of cortical, subcortical, and arterial structures to the prediction. Our results show that combining T1-weighted and angiography MR data led to a significantly improved brain age prediction accuracy, with a mean absolute error of 3.85 years comparing the predicted and chronological age. The most predictive brain regions included the lateral sulcus, the fourth ventricle, and the amygdala, while the brain arteries contributing the most to the prediction included the basilar artery, the middle cerebral artery M2 segments, and the left posterior cerebral artery. Our study proposes a framework for brain age prediction using multimodal imaging, which gives accurate predictions and allows identifying the most predictive regions for this task, which can serve as a surrogate for the brain regions that are most affected by aging.

Dementia risk prediction in individuals with mild cognitive impairment: a comparison of Cox regression and machine learning models.

BACKGROUND: Cox proportional hazards regression models and machine learning models are widely used for predicting the risk of dementia. Existing comparisons of these models have mostly been based on empirical datasets and have yielded mixed results. This study examines the accuracy of various machine learning and of the Cox regression models for predicting time-to-event outcomes using Monte Carlo simulation in people with mild cognitive impairment (MCI). METHODS: The predictive accuracy of nine time-to-event regression and machine learning models were investigated. These models include Cox regression, penalized Cox regression (with Ridge, LASSO, and elastic net penalties), survival trees, random survival forests, survival support vector machines, artificial neural networks, and extreme gradient boosting. Simulation data were generated using study design and data characteristics of a clinical registry and a large community-based registry of patients with MCI. The predictive performance of these models was evaluated based on three-fold cross-validation via Harrell's concordance index (c-index), integrated calibration index (ICI), and integrated brier score (IBS). RESULTS: Cox regression and machine learning model had comparable predictive accuracy across three different performance metrics and data-analytic conditions. The estimated c-index values for Cox regression, random survival forests, and extreme gradient boosting were 0.70, 0.69 and 0.70, respectively, when the data were generated from a Cox regression model in a large sample-size conditions. In contrast, the estimated c-index values for these models were 0.64, 0.64, and 0.65 when the data were generated from a random survival forest in a large sample size conditions. Both Cox regression and random survival forest had the lowest ICI values (0.12 for a large sample size and 0.18 for a small sample size) among all the investigated models regardless of sample size and data generating model. CONCLUSION: Cox regression models have comparable, and sometimes better predictive performance, than more complex machine learning models. We recommend that the choice among these models should be guided by important considerations for research hypotheses, model interpretability, and type of data.

Machine learning for precision medicine.

Precision medicine is an emerging approach to clinical research and patient care that focuses on understanding and treating disease by integrating multi-modal or multi-omics data from an individual to make patient-tailored decisions. With the large and complex datasets generated using precision medicine diagnostic approaches, novel techniques to process and understand these complex data were needed. At the same time, computer science has progressed rapidly to develop techniques that enable the storage, processing, and analysis of these complex datasets, a feat that traditional statistics and early computing technologies could not accomplish. Machine learning, a branch of artificial intelligence, is a computer science methodology that aims to identify complex patterns in data that can be used to make predictions or classifications on new unseen data or for advanced exploratory data analysis. Machine learning analysis of precision medicine's multi-modal data allows for broad analysis of large datasets and ultimately a greater understanding of human health and disease. This review focuses on machine learning utilization for precision medicine's "big data", in the context of genetics, genomics, and beyond.

Neural correlates of the impulse dyscontrol domain of mild behavioral impairment.

OBJECTIVES: Agitation and aggression are common in dementia and pre-dementia. The dementia risk syndrome mild behavioral impairment (MBI) includes these symptoms in the impulse dyscontrol domain. However, the neural circuitry associated with impulse dyscontrol in neurodegenerative disease is not well understood. The objective of this work was to investigate if regional micro- and macro-structural brain properties were associated with impulse dyscontrol symptoms in older adults with normal cognition, mild cognitive impairment, and Alzheimer's disease (AD). METHODS: Clinical, neuropsychiatric, and T1-weighted and diffusion-tensor magnetic resonance imaging (DTI) data from 80 individuals with and 123 individuals without impulse dyscontrol were obtained from the AD Neuroimaging Initiative. Linear mixed effect models were used to assess if impulse dyscontrol was related to regional DTI and volumetric parameters. RESULTS: Impulse dyscontrol was present in 17% of participants with NC, 43% with MCI, and 66% with AD. Impulse dyscontrol was associated with: (1) lower fractional anisotropy (FA), and greater mean, axial, and radial diffusivity in the fornix; (2) lesser FA and greater radial diffusivity in the superior fronto-occipital fasciculus; (3) greater axial diffusivity in the cingulum; (4) greater axial and radial diffusivity in the uncinate fasciculus; (5) gray matter atrophy, specifically, lower cortical thickness in the parahippocampal gyrus. CONCLUSION: Our findings provide evidence that well-established atrophy patterns of AD are prominent in the presence of impulse dyscontrol, even when disease status is controlled for, and possibly in advance of dementia. Our findings support the growing evidence for impulse dyscontrol symptoms as an early manifestation of AD.

An Analysis of the Vulnerability of Two Common Deep Learning-Based Medical Image Segmentation Techniques to Model Inversion Attacks.

Recent research in computer vision has shown that original images used for training of deep learning models can be reconstructed using so-called inversion attacks. However, the feasibility of this attack type has not been investigated for complex 3D medical images. Thus, the aim of this study was to examine the vulnerability of deep learning techniques used in medical imaging to model inversion attacks and investigate multiple quantitative metrics to evaluate the quality of the reconstructed images. For the development and evaluation of model inversion attacks, the public LPBA40 database consisting of 40 brain MRI scans with corresponding segmentations of the gyri and deep grey matter brain structures were used to train two popular deep convolutional neural networks, namely a U-Net and SegNet, and corresponding inversion decoders. Matthews correlation coefficient, the structural similarity index measure (SSIM), and the magnitude of the deformation field resulting from non-linear registration of the original and reconstructed images were used to evaluate the reconstruction accuracy. A comparison of the similarity metrics revealed that the SSIM is best suited to evaluate the reconstruction accuray, followed closely by the magnitude of the deformation field. The quantitative evaluation of the reconstructed images revealed SSIM scores of 0.73±0.12 and 0.61±0.12 for the U-Net and the SegNet, respectively. The qualitative evaluation showed that training images can be reconstructed with some degradation due to blurring but can be correctly matched to the original images in the majority of the cases. In conclusion, the results of this study indicate that it is possible to reconstruct patient data used for training of convolutional neural networks and that the SSIM is a good metric to assess the reconstruction accuracy.

Combined Atlas and Convolutional Neural Network-Based Segmentation of the Hippocampus from MRI According to the ADNI Harmonized Protocol.

Hippocampus atrophy is an early structural feature that can be measured from magnetic resonance imaging (MRI) to improve the diagnosis of neurological diseases. An accurate and robust standardized hippocampus segmentation method is required for reliable atrophy assessment. The aim of this work was to develop and evaluate an automatic segmentation tool (DeepHarp) for hippocampus delineation according to the ADNI harmonized hippocampal protocol (HarP). DeepHarp utilizes a two-step process. First, the approximate location of the hippocampus is identified in T1-weighted MRI datasets using an atlas-based approach, which is used to crop the images to a region-of-interest (ROI) containing the hippocampus. In the second step, a convolutional neural network trained using datasets with corresponding manual hippocampus annotations is used to segment the hippocampus from the cropped ROI. The proposed method was developed and validated using 107 datasets with manually segmented hippocampi according to the ADNI-HarP standard as well as 114 multi-center datasets of patients with Alzheimer's disease, mild cognitive impairment, cerebrovascular disease, and healthy controls. Twenty-three independent datasets manually segmented according to the ADNI-HarP protocol were used for testing to assess the accuracy, while an independent test-retest dataset was used to assess precision. The proposed DeepHarp method achieved a mean Dice similarity score of 0.88, which was significantly better than four other established hippocampus segmentation methods used for comparison. At the same time, the proposed method also achieved a high test-retest precision (mean Dice score: 0.95). In conclusion, DeepHarp can automatically segment the hippocampus from T1-weighted MRI datasets according to the ADNI-HarP protocol with high accuracy and robustness, which can aid atrophy measurements in a variety of pathologies.

Automated syndrome diagnosis by three-dimensional facial imaging.

PURPOSE: Deep phenotyping is an emerging trend in precision medicine for genetic disease. The shape of the face is affected in 30-40% of known genetic syndromes. Here, we determine whether syndromes can be diagnosed from 3D images of human faces. METHODS: We analyzed variation in three-dimensional (3D) facial images of 7057 subjects: 3327 with 396 different syndromes, 727 of their relatives, and 3003 unrelated, unaffected subjects. We developed and tested machine learning and parametric approaches to automated syndrome diagnosis using 3D facial images. RESULTS: Unrelated, unaffected subjects were correctly classified with 96% accuracy. Considering both syndromic and unrelated, unaffected subjects together, balanced accuracy was 73% and mean sensitivity 49%. Excluding unrelated, unaffected subjects substantially improved both balanced accuracy (78.1%) and sensitivity (56.9%) of syndrome diagnosis. The best predictors of classification accuracy were phenotypic severity and facial distinctiveness of syndromes. Surprisingly, unaffected relatives of syndromic subjects were frequently classified as syndromic, often to the syndrome of their affected relative. CONCLUSION: Deep phenotyping by quantitative 3D facial imaging has considerable potential to facilitate syndrome diagnosis. Furthermore, 3D facial imaging of "unaffected" relatives may identify unrecognized cases or may reveal novel examples of semidominant inheritance.

Building machine learning models without sharing patient data: A simulation-based analysis of distributed learning by ensembling.

The development of machine learning solutions in medicine is often hindered by difficulties associated with sharing patient data. Distributed learning aims to train machine learning models locally without requiring data sharing. However, the utility of distributed learning for rare diseases, with only a few training examples at each contributing local center, has not been investigated. The aim of this work was to simulate distributed learning models by ensembling with artificial neural networks (ANN), support vector machines (SVM), and random forests (RF) and evaluate them using four medical datasets. Distributed learning by ensembling locally trained agents improved performance compared to models trained using the data from a single institution, even in cases where only a very few training examples are available per local center. Distributed learning improved when more locally trained models were added to the ensemble. Local class imbalance reduced distributed SVM performance but did not impact distributed RF and ANN classification. Our results suggest that distributed learning by ensembling can be used to train machine learning models without sharing patient data and is suitable to use with small datasets.

Children with epilepsy demonstrate macro- and microstructural changes in the thalamus, putamen, and amygdala.

PURPOSE: Despite evidence for macrostructural alteration in epilepsy patients later in life, little is known about the underlying pathological or compensatory mechanisms at younger ages causing these alterations. The aim of this work was to investigate the impact of pediatric epilepsy on the central nervous system, including gray matter volume, cerebral blood flow, and water diffusion, compared with neurologically normal children. METHODS: Inter-ictal magnetic resonance imaging data was obtained from 30 children with epilepsy ages 1-16 (73% F, 27% M). An atlas-based approach was used to determine values for volume, cerebral blood flow, and apparent diffusion coefficient in the cerebral cortex, hippocampus, thalamus, caudate, putamen, globus pallidus, amygdala, and nucleus accumbens. These values were then compared with previously published values from 100 neurologically normal children using a MANCOVA analysis. RESULTS: Most brain volumes of children with epilepsy followed a pattern similar to typically developing children, except for significantly larger putamen and amygdala. Cerebral blood flow was also comparable between the groups, except for the putamen, which demonstrated decreased blood flow in children with epilepsy. Diffusion (apparent diffusion coefficient) showed a trend towards higher values in children with epilepsy, with significantly elevated diffusion within the thalamus in children with epilepsy compared with neurologically normal children. CONCLUSION: Children with epilepsy show statistically significant differences in volume, diffusion, and cerebral blood flow within their thalamus, putamen, and amygdala, suggesting that epilepsy is associated with structural changes of the central nervous system influencing brain development and potentially leading to poorer neurocognitive outcomes.

Interdatabase Variability in Cortical Thickness Measurements.

The phenomenon of cortical thinning with age has been well established; however, the measured rate of change varies between studies. The source of this variation could be image acquisition techniques including hardware and vendor specific differences. Databases are often consolidated to increase the number of subjects but underlying differences between these datasets could have undesired effects. We explore differences in cerebral cortex thinning between 4 databases, totaling 1382 subjects. We investigate several aspects of these databases, including: 1) differences between databases of cortical thinning rates versus age, 2) correlation of cortical thinning rates between regions for each database, and 3) regression bootstrapping to determine the effect of the number of subjects included. We also examined the effect of different databases on age prediction modeling. Cortical thinning rates were significantly different between databases in all 68 parcellated regions (ANCOVA, P < 0.001). Subtle differences were observed in correlation matrices and bootstrapping convergence. Age prediction modeling using a leave-one-out cross-validation approach showed varying prediction performance (0.64 < R2 < 0.82) between databases. When a database was used to calibrate the model and then applied to another database, prediction performance consistently decreased. We conclude that there are indeed differences in the measured cortical thinning rates between these large-scale databases.

The Association of Saliva Cytokines and Pediatric Sports-Related Concussion Outcomes.

OBJECTIVES: This study aimed to explore cytokine alterations following pediatric sports-related concussion (SRC) and whether a specific cytokine profile could predict symptom burden and time to return to sports (RTS). SETTING: Sports Medicine Clinic. PARTICIPANTS: Youth ice hockey participants (aged 12-17 years) were recruited prior to the 2013-2016 hockey season. DESIGN: Prospective exploratory cohort study. MAIN MEASURE: Following SRC, saliva samples were collected and a Sport Concussion Assessment Tool version 3 (SCAT3) was administered within 72 hours of injury and analyzed for cytokines. Additive regression of decision stumps was used to model symptom burden and length to RTS based on cytokine and clinical features. RRelieFF feature selection was used to determine the predictive value of each cytokine and clinical feature, as well as to identify the optimal cytokine profile for the symptom burden and RTS. RESULTS: Thirty-six participants provided samples post-SRC (81% male; age 14.4 ± 1.3 years). Of these, 10 features, sex, number of previous concussions, and 8 cytokines, were identified to lead to the best prediction of symptom severity (r = 0.505, P = .002), while 12 cytokines, age, and history of previous concussions predicted the number of symptoms best (r = 0.637, P < .001). The prediction of RTS led to the worst results, requiring 21 cytokines, age, sex, and number of previous concussions as features (r = -0.320, P = .076). CONCLUSIONS: In pediatric ice hockey participants following SRC, there is evidence of saliva cytokine profiles that are associated with increased symptom burden. However, further studies are needed.

Fully Automatic Landmarking of Syndromic 3D Facial Surface Scans Using 2D Images.

3D facial landmarks are known to be diagnostically relevant biometrics for many genetic syndromes. The objective of this study was to extend a state-of-the-art image-based 2D facial landmarking algorithm for the challenging task of 3D landmark identification on subjects with genetic syndromes, who often have moderate to severe facial dysmorphia. The automatic 3D facial landmarking algorithm presented here uses 2D image-based facial detection and landmarking models to identify 12 landmarks on 3D facial surface scans. The landmarking algorithm was evaluated using a test set of 444 facial scans with ground truth landmarks identified by two different human observers. Three hundred and sixty nine of the subjects in the test set had a genetic syndrome that is associated with facial dysmorphology. For comparison purposes, the manual landmarks were also used to initialize a non-linear surface-based registration of a non-syndromic atlas to each subject scan. Compared to the average intra- and inter-observer landmark distances of 1.1 mm and 1.5 mm respectively, the average distance between the manual landmark positions and those produced by the automatic image-based landmarking algorithm was 2.5 mm. The average error of the registration-based approach was 3.1 mm. Comparing the distributions of Procrustes distances from the mean for each landmarking approach showed that the surface registration algorithm produces a systemic bias towards the atlas shape. In summary, the image-based automatic landmarking approach performed well on this challenging test set, outperforming a semi-automatic surface registration approach, and producing landmark errors that are comparable to state-of-the-art 3D geometry-based facial landmarking algorithms evaluated on non-syndromic subjects.

ERASER.

Background and Purpose- Using a novel study design with virtual comparators based on predictive modeling, we investigated whether next-generation mechanical thrombectomy devices improve outcomes in patients with ischemic stroke. We hypothesized that this new study design shows that a next-generation mechanical thrombectomy system is superior to intravenous tPA (tissue-type plasminogen activator) therapy (IVT) alone. Methods- ERASER (Eric Acute Stroke Recanalization) was an investigator-initiated, prospective, multicenter, single-arm (virtual 2-arm) study that evaluated the effectiveness of a new recanalization device together with a specific intermediate catheter (Embolus Retriever with Interlinked Cages/SOFIA, Microvention) in stroke patients with internal carotid artery or middle cerebral artery occlusions. The primary end point was the volume of saved tissue. Volume of saved tissue was defined as the difference of actual infarct volume and brain volume predicted to develop infarction using a machine learning model based on data from intravenous tPA therapy patients. Results- Eighty-one patients were enrolled. The median patient age was 71 years (interquartile range, 61-77). National Institutes of Health Stroke Scale score was 14 (interquartile range, 12-18). The actual infarct volume was smaller than predicted by the intravenous tPA therapy model, with a median volume of saved tissue of 50 mL (interquartile range, 19-103; P<0.0001). Good clinical outcome (modified Rankin Scale, 0-2 at 90 days) was observed in 48 out of 69 (70%). The recanalization rate (Thrombolysis in Cerebral Infarction 2b/3) was 95%. Conclusions- ERASER is the first mechanical thrombectomy study with a primary end point based on predictive analytics enabling intraindividual virtual comparisons. The next-generation mechanical thrombectomy method resulted in smaller infarcts than predicted after intravenous tPA therapy alone and showed a high rate of good clinical outcome. The novel study design with virtual comparisons is promising for further application and testing in the neurovascular arena. Clinical Trial Registration- URL: https://www.clinicaltrials.gov . Unique identifier: NCT02534701.

Rapid solution of the Bloch-Torrey equation in anisotropic tissue: Application to dynamic susceptibility contrast MRI of cerebral white matter.

Blood vessel related magnetic resonance imaging (MRI) contrast provides a window into the brain's metabolism and function. Here, we show that the spin echo dynamic susceptibility contrast (DSC) MRI signal of the brain's white matter (WM) strongly depends on the angle between WM tracts and the main magnetic field. The apparent cerebral blood flow and volume are 20% larger in fibres perpendicular to the main magnetic field compared to parallel fibres. We present a rapid numerical framework for the solution of the Bloch-Torrey equation that allows us to explore the isotropic and anisotropic components of the vascular tree. By fitting the simulated spin echo DSC signal to the measured data, we show that half of the WM vascular volume is comprised of vessels running in parallel with WM fibre tracts. The WM blood volume corresponding to the best fit to the experimental data was 2.82%, which is close to the PET gold standard of 2.6%.

Attenuation Changes in ASPECTS Regions: A Surrogate for CT Perfusion-based Ischemic Core in Acute Ischemic Stroke.

Background Recent studies have proven the effectiveness of thrombectomy up to 24 hours after stroke onset for patients with specific criteria at advanced CT or MRI. Clinical implementation of treatment in this extended time window remains a challenge, as many stroke centers do not routinely use advanced imaging. Purpose To determine whether automated cerebral x-ray attenuation measurements at noncontrast CT provide information on the presence of CT perfusion-defined ischemic core as applied in late time windows for thrombectomy. Materials and Methods In this retrospective study, patients with middle cerebral artery stroke due to proximal occlusion from 2009 to 2017 were included. All patients underwent noncontrast CT and CT perfusion. Automated software was used to calculate relative Hounsfield unit (rHU) values for Alberta Stroke Program Early CT Score (ASPECTS) regions on noncontrast CT images as the ratio of x-ray attenuation between ischemic versus non-ischemic hemispheres. Sensitivity, specificity, and diagnostic performance of rHU and composite rHU-ASPECTS, a score incorporating rHU from all regions, were analyzed for the classification of regional ischemic core and late time window thrombectomy criteria at CT perfusion. Results Data in a total of 200 patients were evaluated (105 women [mean age, 74 years ± 14 {standard deviation}] and 95 men [mean age, 76 years ± 14]). There were 121 patients in the validation cohort and 79 patients in the independent test cohort. Compared among all examined regions, rHU values yielded the best classification of ischemic core for the caudate nucleus, the lentiform nucleus, and the insula (with areas under the receiver operating characteristic curve [AUCs] ranging from 0.70 to 0.77; P < .001 for each). The composite rHU-ASPECTS score allowed classification of CT perfusion imaging selection criteria of ischemic core sizes of less than 70 mL and target mismatch of greater than 1.8 with AUCs of 0.80 (P = .001; 75% sensitivity and 83% specificity) in the test cohort and 0.74 (P < .001; 58% sensitivity and 82% specificity) in the validation cohort. Conclusion Noncontrast CT x-ray attenuation measurements identify Alberta Stroke Program Early CT Score regions classified as ischemic core at CT perfusion. This approach may serve as a selection criteria surrogate for thrombectomy in late time windows. © RSNA, 2019 Online supplemental material is available for this article.

A longitudinal magnetic resonance imaging study of neurodegenerative and small vessel disease, and clinical cognitive trajectories in non demented patients with transient ischemic attack: the PREVENT study.

BACKGROUND: Late-life cognitive decline, caused by progressive neuronal loss leading to brain atrophy years before symptoms are detected, is expected to double in Canada over the next two decades. Cognitive impairment in late life is attributed to vascular and lifestyle related risk factors in mid-life in a substantial proportion of cases (50%), thereby providing an opportunity for effective prevention of cognitive decline if incipient disease is detected earlier. Patients presenting with transient ischemic attack (TIA) commonly display some degree of cognitive impairment and are at a 4-fold increased risk of dementia. In the Predementia Neuroimaging of Transient Ischemic Attack (PREVENT) study, we will address what disease processes (i.e., Alzheimer's vs. vascular disease) lead to neurodegeneration, brain atrophy, and cognitive decline, and whether imaging measurements of brain iron accumulation using quantitative susceptibility mapping predicts subsequent brain atrophy and cognitive decline. METHODS: A total of 440 subjects will be recruited for this study with 220 healthy subjects and 220 TIA patients. Early Alzheimer's pathology will be determined by cerebrospinal fluid samples (including tau, a marker of neuronal injury, and amyloid ß1-42) and by MR measurements of iron accumulation, a marker for Alzheimer's-related neurodegeneration. Small vessel disease will be identified by changes in white matter lesion volume. Predictors of advanced rates of cerebral and hippocampal atrophy at 1 and 3 years will include in vivo Alzheimer's disease pathology markers, and MRI measurements of brain iron accumulation and small vessel disease. Clinical and cognitive function will be assessed annually post-baseline for a period of 5-years using a clinical questionnaire and a battery of neuropsychological tests, respectively. DISCUSSION: The PREVENT study expects to demonstrate that TIA patients have increased early progressive rates of cerebral brain atrophy after TIA, before cognitive decline can be clinically detected. By developing and optimizing high-level machine learning models based on clinical data, image-based (quantitative susceptibility mapping, regional brain, and white matter lesion volumes) features, and cerebrospinal fluid biomarkers, PREVENT will provide a timely opportunity to identify individuals at greatest risk of late-life cognitive decline early in the course of disease, supporting future therapeutic strategies for the promotion of healthy aging.