RESUMO
Chronic post-concussive symptoms are common after mild traumatic brain injury (mTBI) and are difficult to predict or treat. Thalamic functional integrity is particularly vulnerable in mTBI and may be related to long-term outcomes but requires further investigation. We compared structural MRI and resting state functional MRI in 108 patients with a Glasgow Coma Scale (GCS) of 13-15 and normal CT, and 76 controls. We examined whether acute changes in thalamic functional connectivity were early markers for persistent symptoms and explored neurochemical associations of our findings using PET data. Of the mTBI cohort, 47% showed incomplete recovery 6 months post-injury. Despite the absence of structural changes, we found acute thalamic hyperconnectivity in mTBI, with specific vulnerabilities of individual thalamic nuclei. Acute fMRI markers differentiated those with chronic post-concussive symptoms, with time- and outcome-dependent relationships in a sub-cohort followed longitudinally. Moreover, emotional and cognitive symptoms were associated with changes in thalamic functional connectivity to known serotonergic and noradrenergic targets, respectively. Our findings suggest that chronic symptoms can have a basis in early thalamic pathophysiology. This may aid identification of patients at risk of chronic post-concussive symptoms following mTBI, provide a basis for development of new therapies and facilitate precision medicine application of these therapies.
Assuntos
Concussão Encefálica , Lesões Encefálicas , Síndrome Pós-Concussão , Humanos , Concussão Encefálica/complicações , Concussão Encefálica/diagnóstico por imagem , Síndrome Pós-Concussão/diagnóstico por imagem , Tálamo/diagnóstico por imagem , Emoções , Imageamento por Ressonância Magnética , EncéfaloRESUMO
BACKGROUND: Structural disconnectivity was found to precede dementia. Global white matter abnormalities might also be associated with postoperative delirium (POD). METHODS: We recruited older patients (≥65 years) without dementia that were scheduled for major surgery. Diffusion kurtosis imaging metrics were obtained preoperatively, after 3 and 12 months postoperatively. We calculated fractional anisotropy (FA), mean diffusivity (MD), mean kurtosis (MK), and free water (FW). A structured and validated delirium assessment was performed twice daily. RESULTS: Of 325 patients, 53 patients developed POD (16.3%). Preoperative global MD (standardized beta 0.27 [95% confidence interval [CI] 0.21-0.32] p < 0.001) was higher in patients with POD. Preoperative global MK (-0.07 [95% CI -0.11 to (-0.04)] p < 0.001) and FA (0.07 [95% CI -0.10 to (-0.04)] p < 0.001) were lower. When correcting for baseline diffusion, postoperative MD was lower after 3 months (0.05 [95% CI -0.08 to (-0.03)] p < 0.001; n = 183) and higher after 12 months (0.28 [95% CI 0.20-0.35] p < 0.001; n = 45) among patients with POD. DISCUSSION: Preoperative structural disconnectivity was associated with POD. POD might lead to white matter depletion 3 and 12 months after surgery.
Assuntos
Demência , Delírio do Despertar , Substância Branca , Humanos , Idoso , Estudos de Coortes , Substância Branca/diagnóstico por imagem , Imagem de Tensor de Difusão/métodosRESUMO
There is substantial interest in the potential for traumatic brain injury to result in progressive neurological deterioration. While blood biomarkers such as glial fibrillary acid protein (GFAP) and neurofilament light have been widely explored in characterizing acute traumatic brain injury (TBI), their use in the chronic phase is limited. Given increasing evidence that these proteins may be markers of ongoing neurodegeneration in a range of diseases, we examined their relationship to imaging changes and functional outcome in the months to years following TBI. Two-hundred and three patients were recruited in two separate cohorts; 6 months post-injury (n = 165); and >5 years post-injury (n = 38; 12 of whom also provided data â¼8 months post-TBI). Subjects underwent blood biomarker sampling (n = 199) and MRI (n = 172; including diffusion tensor imaging). Data from patient cohorts were compared to 59 healthy volunteers and 21 non-brain injury trauma controls. Mean diffusivity and fractional anisotropy were calculated in cortical grey matter, deep grey matter and whole brain white matter. Accelerated brain ageing was calculated at a whole brain level as the predicted age difference defined using T1-weighted images, and at a voxel-based level as the annualized Jacobian determinants in white matter and grey matter, referenced to a population of 652 healthy control subjects. Serum neurofilament light concentrations were elevated in the early chronic phase. While GFAP values were within the normal range at â¼8 months, many patients showed a secondary and temporally distinct elevations up to >5 years after injury. Biomarker elevation at 6 months was significantly related to metrics of microstructural injury on diffusion tensor imaging. Biomarker levels at â¼8 months predicted white matter volume loss at >5 years, and annualized brain volume loss between â¼8 months and 5 years. Patients who worsened functionally between â¼8 months and >5 years showed higher than predicted brain age and elevated neurofilament light levels. GFAP and neurofilament light levels can remain elevated months to years after TBI, and show distinct temporal profiles. These elevations correlate closely with microstructural injury in both grey and white matter on contemporaneous quantitative diffusion tensor imaging. Neurofilament light elevations at â¼8 months may predict ongoing white matter and brain volume loss over >5 years of follow-up. If confirmed, these findings suggest that blood biomarker levels at late time points could be used to identify TBI survivors who are at high risk of progressive neurological damage.
Assuntos
Lesões Encefálicas Traumáticas , Lesões Encefálicas , Substância Branca , Biomarcadores , Lesões Encefálicas/complicações , Lesões Encefálicas Traumáticas/complicações , Lesões Encefálicas Traumáticas/diagnóstico por imagem , Imagem de Tensor de Difusão/métodos , Progressão da Doença , Proteína Glial Fibrilar Ácida/metabolismo , HumanosRESUMO
Metabolic derangements following traumatic brain injury are poorly characterized. In this single-centre observational cohort study we combined 18F-FDG and multi-tracer oxygen-15 PET to comprehensively characterize the extent and spatial pattern of metabolic derangements. Twenty-six patients requiring sedation and ventilation with intracranial pressure monitoring following head injury within a Neurosciences Critical Care Unit, and 47 healthy volunteers were recruited. Eighteen volunteers were excluded for age over 60 years (n = 11), movement-related artefact (n = 3) or physiological instability during imaging (n = 4). We measured cerebral blood flow, blood volume, oxygen extraction fraction, and 18F-FDG transport into the brain (K1) and its phosphorylation (k3). We calculated oxygen metabolism, 18F-FDG influx rate constant (Ki), glucose metabolism and the oxygen/glucose metabolic ratio. Lesion core, penumbra and peri-penumbra, and normal-appearing brain, ischaemic brain volume and k3 hotspot regions were compared with plasma and microdialysis glucose in patients. Twenty-six head injury patients, median age 40 years (22 male, four female) underwent 34 combined 18F-FDG and oxygen-15 PET at early, intermediate, and late time points (within 24 h, Days 2-5, and Days 6-12 post-injury; n = 12, 8, and 14, respectively), and were compared with 20 volunteers, median age 43 years (15 male, five female) who underwent oxygen-15, and nine volunteers, median age 56 years (three male, six female) who underwent 18F-FDG PET. Higher plasma glucose was associated with higher microdialysate glucose. Blood flow and K1 were decreased in the vicinity of lesions, and closely related when blood flow was <25 ml/100 ml/min. Within normal-appearing brain, K1 was maintained despite lower blood flow than volunteers. Glucose utilization was globally reduced in comparison with volunteers (P < 0.001). k3 was variable; highest within lesions with some patients showing increases with blood flow <25 ml/100 ml/min, but falling steeply with blood flow lower than 12 ml/100 ml/min. k3 hotspots were found distant from lesions, with k3 increases associated with lower plasma glucose (Rho -0.33, P < 0.001) and microdialysis glucose (Rho -0.73, P = 0.02). k3 hotspots showed similar K1 and glucose metabolism to volunteers despite lower blood flow and oxygen metabolism (P < 0.001, both comparisons); oxygen extraction fraction increases consistent with ischaemia were uncommon. We show that glucose delivery was dependent on plasma glucose and cerebral blood flow. Overall glucose utilization was low, but regional increases were associated with reductions in glucose availability, blood flow and oxygen metabolism in the absence of ischaemia. Clinical management should optimize blood flow and glucose delivery and could explore the use of alternative energy substrates.
Assuntos
Lesões Encefálicas Traumáticas/metabolismo , Circulação Cerebrovascular/fisiologia , Glucose/metabolismo , Adulto , Encéfalo/irrigação sanguínea , Encéfalo/metabolismo , Estudos de Coortes , Feminino , Humanos , Masculino , Pessoa de Meia-Idade , Tomografia por Emissão de PósitronsRESUMO
BACKGROUND: Magnetic resonance imaging (MRI) carries prognostic importance after traumatic brain injury (TBI), especially when computed tomography (CT) fails to fully explain the level of unconsciousness. However, in critically ill patients, the risk of deterioration during transfer needs to be balanced against the benefit of detecting prognostically relevant information on MRI. We therefore aimed to assess if day of injury serum protein biomarkers could identify critically ill TBI patients in whom the risks of transfer are compensated by the likelihood of detecting management-altering neuroimaging findings. METHODS: Data were obtained from the Collaborative European NeuroTrauma Effectiveness Research in Traumatic Brain Injury (CENTER-TBI) study. Eligibility criteria included: TBI patients aged ≥ 16 years, Glasgow Coma Score (GCS) < 13 or patient intubated with unrecorded pre-intubation GCS, CT with Marshall score < 3, serum biomarkers (GFAP, NFL, NSE, S100B, Tau, UCH-L1) sampled ≤ 24 h of injury, MRI < 30 days of injury. The degree of axonal injury on MRI was graded using the Adams-Gentry classification. The association between serum concentrations of biomarkers and Adams-Gentry stage was assessed and the optimum threshold concentration identified, assuming different minimum sensitivities for the detection of brainstem injury (Adams-Gentry stage 3). A cost-benefit analysis for the USA and UK health care settings was also performed. RESULTS: Among 65 included patients (30 moderate-severe, 35 unrecorded) axonal injury was detected in 54 (83%) and brainstem involvement in 33 (51%). In patients with moderate-severe TBI, brainstem injury was associated with higher concentrations of NSE, Tau, UCH-L1 and GFAP. If the clinician did not want to miss any brainstem injury, NSE could have avoided MRI transfers in up to 20% of patients. If a 94% sensitivity was accepted considering potential transfer-related complications, GFAP could have avoided 30% of transfers. There was no added net cost, with savings up to £99 (UK) or $612 (US). No associations between proteins and axonal injury were found in intubated patients without a recorded pre-intubation GCS. CONCLUSIONS: Serum protein biomarkers show potential to safely reduce the number of transfers to MRI in critically ill patients with moderate-severe TBI at no added cost.
Assuntos
Lesões Encefálicas Traumáticas , Estado Terminal , Humanos , Lesões Encefálicas Traumáticas/diagnóstico por imagem , Biomarcadores , Imageamento por Ressonância Magnética , Tomografia Computadorizada por Raios XRESUMO
BACKGROUND AND PURPOSE: The ISLES challenge (Ischemic Stroke Lesion Segmentation) enables globally diverse teams to compete to develop advanced tools for stroke lesion analysis with machine learning. Detection of irreversibly damaged tissue on computed tomography perfusion (CTP) is often necessary to determine eligibility for late-time-window thrombectomy. Therefore, the aim of ISLES-2018 was to segment infarcted tissue on CTP based on diffusion-weighted imaging as a reference standard. METHODS: The data, from 4 centers, consisted of 103 cases of acute anterior circulation large artery occlusion stroke who underwent diffusion-weighted imaging rapidly after CTP. Diffusion-weighted imaging lesion segmentation was performed manually and acted as a reference standard. The data were separated into 63 cases for training and 40 for testing, upon which quality metrics (dice score coefficient, Hausdorff distance, absolute lesion volume difference, etc) were computed to rank methods based on their overall performance. RESULTS: Twenty-four different teams participated in the challenge. Median time to CTP was 185 minutes (interquartile range, 180-238), the time between CTP and magnetic resonance imaging was 36 minutes (interquartile range, 25-79), and the median infarct lesion size was 15.2 mL (interquartile range, 5.7-45). The best performance for Dice score coefficient and absolute volume difference were 0.51 and 10.1 mL, respectively, from different teams. Based on the ranking criteria, the top team's algorithm demonstrated for average Dice score coefficient and average absolute volume difference 0.51 and 10.2 mL, respectively, outperforming the conventional threshold-based method (dice score coefficient, 0.3; volume difference, 15.3). Diverse algorithms were used, almost all based on deep learning, with top-ranked approaches making use of the raw perfusion data as well as methods to synthetically generate complementary information to boost prediction performance. CONCLUSIONS: Machine learning methods may predict infarcted tissue from CTP with improved accuracy compared with threshold-based methods used in clinical routine. This dataset will remain public and can be used to test improvement in algorithms over time.
Assuntos
Isquemia Encefálica/diagnóstico por imagem , Aprendizado de Máquina , Imagem de Perfusão/métodos , Acidente Vascular Cerebral/diagnóstico por imagem , Tomografia Computadorizada por Raios X/métodos , Idoso , Idoso de 80 Anos ou mais , Isquemia Encefálica/fisiopatologia , Infarto Cerebral/diagnóstico por imagem , Infarto Cerebral/fisiopatologia , Feminino , Humanos , Imageamento por Ressonância Magnética/métodos , Masculino , Pessoa de Meia-Idade , Valor Preditivo dos Testes , Acidente Vascular Cerebral/fisiopatologiaRESUMO
Background and Purpose- We evaluated deep learning algorithms' segmentation of acute ischemic lesions on heterogeneous multi-center clinical diffusion-weighted magnetic resonance imaging (MRI) data sets and explored the potential role of this tool for phenotyping acute ischemic stroke. Methods- Ischemic stroke data sets from the MRI-GENIE (MRI-Genetics Interface Exploration) repository consisting of 12 international genetic research centers were retrospectively analyzed using an automated deep learning segmentation algorithm consisting of an ensemble of 3-dimensional convolutional neural networks. Three ensembles were trained using data from the following: (1) 267 patients from an independent single-center cohort, (2) 267 patients from MRI-GENIE, and (3) mixture of (1) and (2). The algorithms' performances were compared against manual outlines from a separate 383 patient subset from MRI-GENIE. Univariable and multivariable logistic regression with respect to demographics, stroke subtypes, and vascular risk factors were performed to identify phenotypes associated with large acute diffusion-weighted MRI volumes and greater stroke severity in 2770 MRI-GENIE patients. Stroke topography was investigated. Results- The ensemble consisting of a mixture of MRI-GENIE and single-center convolutional neural networks performed best. Subset analysis comparing automated and manual lesion volumes in 383 patients found excellent correlation (ρ=0.92; P<0.0001). Median (interquartile range) diffusion-weighted MRI lesion volumes from 2770 patients were 3.7 cm3 (0.9-16.6 cm3). Patients with small artery occlusion stroke subtype had smaller lesion volumes ( P<0.0001) and different topography compared with other stroke subtypes. Conclusions- Automated accurate clinical diffusion-weighted MRI lesion segmentation using deep learning algorithms trained with multi-center and diverse data is feasible. Both lesion volume and topography can provide insight into stroke subtypes with sufficient sample size from big heterogeneous multi-center clinical imaging phenotype data sets.
Assuntos
Isquemia Encefálica/diagnóstico por imagem , Imagem de Difusão por Ressonância Magnética/métodos , Acidente Vascular Cerebral/diagnóstico por imagem , Adulto , Idoso , Idoso de 80 Anos ou mais , Algoritmos , Big Data , Isquemia Encefálica/epidemiologia , Feminino , Humanos , Processamento de Imagem Assistida por Computador , Aprendizado de Máquina , Masculino , Pessoa de Meia-Idade , Redes Neurais de Computação , Variações Dependentes do Observador , Fenótipo , Estudos Retrospectivos , Fatores de Risco , Fatores Socioeconômicos , Acidente Vascular Cerebral/epidemiologiaRESUMO
Purpose: Diffusion tensor imaging (DTI) is a magnetic resonance imaging technique that provides unique information about white matter microstructure in the brain but is susceptible to confounding effects introduced by scanner or acquisition differences. ComBat is a leading approach for addressing these site biases. However, despite its frequent use for harmonization, ComBat's robustness toward site dissimilarities and overall cohort size have not yet been evaluated in terms of DTI. Approach: As a baseline, we match N=358 participants from two sites to create a "silver standard" that simulates a cohort for multi-site harmonization. Across sites, we harmonize mean fractional anisotropy and mean diffusivity, calculated using participant DTI data, for the regions of interest defined by the JHU EVE-Type III atlas. We bootstrap 10 iterations at 19 levels of total sample size, 10 levels of sample size imbalance between sites, and 6 levels of mean age difference between sites to quantify (i) ßAGE, the linear regression coefficient of the relationship between FA and age; (ii) γ/f*, the ComBat-estimated site-shift; and (iii) δ/f*, the ComBat-estimated site-scaling. We characterize the reliability of ComBat by evaluating the root mean squared error in these three metrics and examine if there is a correlation between the reliability of ComBat and a violation of assumptions. Results: ComBat remains well behaved for ßAGE when N>162 and when the mean age difference is less than 4 years. The assumptions of the ComBat model regarding the normality of residual distributions are not violated as the model becomes unstable. Conclusion: Prior to harmonization of DTI data with ComBat, the input cohort should be examined for size and covariate distributions of each site. Direct assessment of residual distributions is less informative on stability than bootstrap analysis. We caution use ComBat of in situations that do not conform to the above thresholds.
RESUMO
Deep learning has allowed for remarkable progress in many medical scenarios. Deep learning prediction models often require 105-107 examples. It is currently unknown whether deep learning can also enhance predictions of symptoms post-stroke in real-world samples of stroke patients that are often several magnitudes smaller. Such stroke outcome predictions however could be particularly instrumental in guiding acute clinical and rehabilitation care decisions. We here compared the capacities of classically used linear and novel deep learning algorithms in their prediction of stroke severity. Our analyses relied on a total of 1430 patients assembled from the MRI-Genetics Interface Exploration collaboration and a Massachusetts General Hospital-based study. The outcome of interest was National Institutes of Health Stroke Scale-based stroke severity in the acute phase after ischaemic stroke onset, which we predict by means of MRI-derived lesion location. We automatically derived lesion segmentations from diffusion-weighted clinical MRI scans, performed spatial normalization and included a principal component analysis step, retaining 95% of the variance of the original data. We then repeatedly separated a train, validation and test set to investigate the effects of sample size; we subsampled the train set to 100, 300 and 900 and trained the algorithms to predict the stroke severity score for each sample size with regularized linear regression and an eight-layered neural network. We selected hyperparameters on the validation set. We evaluated model performance based on the explained variance (R2) in the test set. While linear regression performed significantly better for a sample size of 100 patients, deep learning started to significantly outperform linear regression when trained on 900 patients. Average prediction performance improved by â¼20% when increasing the sample size 9× [maximum for 100 patients: 0.279 ± 0.005 (R2, 95% confidence interval), 900 patients: 0.337 ± 0.006]. In summary, for sample sizes of 900 patients, deep learning showed a higher prediction performance than typically employed linear methods. These findings suggest the existence of non-linear relationships between lesion location and stroke severity that can be utilized for an improved prediction performance for larger sample sizes.
RESUMO
Importance: The chronic neuronal burden of traumatic brain injury (TBI) is not fully characterized by routine imaging, limiting understanding of the role of neuronal substrates in adverse outcomes. Objective: To determine whether tissues that appear healthy on routine imaging can be investigated for selective neuronal loss using [11C]flumazenil (FMZ) positron emission tomography (PET) and to examine whether this neuronal loss is associated with long-term outcomes. Design, Setting, and Participants: In this cross-sectional study, data were collected prospectively from 2 centers (University of Cambridge in the UK and Weill Cornell Medicine in the US) between September 1, 2004, and May 31, 2021. Patients with TBI (>6 months postinjury) were compared with healthy control participants (all aged >18 years). Individuals with neurological disease, benzodiazepine use, or contraindication to magnetic resonance imaging were excluded. Data were retrospectively collated with nonconsecutive recruitment, owing to convenience and scanner or PET ligand availability. Data were analyzed between February 1 and September 30, 2023. Exposure: Flumazenil voxelwise binding potential relative to nondisplaceable binding potential (BPND). Main Outcomes and Measures: Selective neuronal loss identified with FMZ PET was compared between groups on voxelwise and regional scales, and its association with functional, cognitive, and psychological outcomes was examined using Glasgow Outcome Scale (GOS) scores, measures of sustained executive attention (animal and sustained fluency), and 36-Item Short Form Health Survey (SF-36) scores. Diffusion tensor imaging was used to assess structural connectivity of regions of cortical damage, and its association with thalamic selective neuronal loss. Results: In this study, 24 patients with chronic TBI (mean [SD] age, 39.2 [12.3] years; 18 men [75.0%]) and 33 healthy control participants (mean [SD] age, 47.6 [20.5] years; 23 men [69.7%]) underwent FMZ PET. Patients with TBI had a median time of 29 (range, 7-95) months from injury to scan. They displayed selective neuronal loss in thalamic nuclei, over and above gross volume loss in the left thalamus, and bilateral central, mediodorsal, ventral-lateral dorsal, anterior, and ventral anterior thalamic nuclei, across a wide range of injury severities. Neuronal loss was associated with worse functional outcome using GOS scores (left thalamus, left ventral anterior, and bilateral central, mediodorsal, and anterior nuclei), worse cognitive outcome on measures of sustained executive attention (left thalamus, bilateral central, and right mediodorsal nuclei), and worse emotional outcome using SF-36 scores (right central thalamic nucleus). Chronic thalamic neuronal loss partially mirrored the location of primary cortical contusions, which may indicate secondary injury mechanisms of transneuronal degeneration. Conclusions and Relevance: The findings of this study suggest that selective thalamic vulnerability may have chronic neuronal consequences with relevance to long-term outcome, suggesting the evolving and potentially lifelong thalamic neuronal consequences of TBI. FMZ PET is a more sensitive marker of the burden of neuronal injury than routine imaging; therefore, it could inform outcome prognostication and may lead to the development of individualized precision medicine approaches.
Assuntos
Lesões Encefálicas Traumáticas , Tomografia por Emissão de Pósitrons , Tálamo , Humanos , Masculino , Feminino , Adulto , Lesões Encefálicas Traumáticas/diagnóstico por imagem , Lesões Encefálicas Traumáticas/patologia , Lesões Encefálicas Traumáticas/complicações , Estudos Transversais , Pessoa de Meia-Idade , Tomografia por Emissão de Pósitrons/métodos , Tálamo/diagnóstico por imagem , Tálamo/patologia , Flumazenil/análogos & derivados , Neurônios/patologiaRESUMO
BACKGROUND: It has been rightfully emphasized that the use of AI for clinical decision making could amplify health disparities. An algorithm may encode protected characteristics, and then use this information for making predictions due to undesirable correlations in the (historical) training data. It remains unclear how we can establish whether such information is actually used. Besides the scarcity of data from underserved populations, very little is known about how dataset biases manifest in predictive models and how this may result in disparate performance. This article aims to shed some light on these issues by exploring methodology for subgroup analysis in image-based disease detection models. METHODS: We utilize two publicly available chest X-ray datasets, CheXpert and MIMIC-CXR, to study performance disparities across race and biological sex in deep learning models. We explore test set resampling, transfer learning, multitask learning, and model inspection to assess the relationship between the encoding of protected characteristics and disease detection performance across subgroups. FINDINGS: We confirm subgroup disparities in terms of shifted true and false positive rates which are partially removed after correcting for population and prevalence shifts in the test sets. We find that transfer learning alone is insufficient for establishing whether specific patient information is used for making predictions. The proposed combination of test-set resampling, multitask learning, and model inspection reveals valuable insights about the way protected characteristics are encoded in the feature representations of deep neural networks. INTERPRETATION: Subgroup analysis is key for identifying performance disparities of AI models, but statistical differences across subgroups need to be taken into account when analyzing potential biases in disease detection. The proposed methodology provides a comprehensive framework for subgroup analysis enabling further research into the underlying causes of disparities. FUNDING: European Research Council Horizon 2020, UK Research and Innovation.
Assuntos
Aprendizado Profundo , Humanos , Raios X , Redes Neurais de Computação , Algoritmos , RadiografiaRESUMO
Purpose: To analyze a recently published chest radiography foundation model for the presence of biases that could lead to subgroup performance disparities across biologic sex and race. Materials and Methods: This Health Insurance Portability and Accountability Act-compliant retrospective study used 127 118 chest radiographs from 42 884 patients (mean age, 63 years ± 17 [SD]; 23 623 male, 19 261 female) from the CheXpert dataset that were collected between October 2002 and July 2017. To determine the presence of bias in features generated by a chest radiography foundation model and baseline deep learning model, dimensionality reduction methods together with two-sample Kolmogorov-Smirnov tests were used to detect distribution shifts across sex and race. A comprehensive disease detection performance analysis was then performed to associate any biases in the features to specific disparities in classification performance across patient subgroups. Results: Ten of 12 pairwise comparisons across biologic sex and race showed statistically significant differences in the studied foundation model, compared with four significant tests in the baseline model. Significant differences were found between male and female (P < .001) and Asian and Black (P < .001) patients in the feature projections that primarily capture disease. Compared with average model performance across all subgroups, classification performance on the "no finding" label decreased between 6.8% and 7.8% for female patients, and performance in detecting "pleural effusion" decreased between 10.7% and 11.6% for Black patients. Conclusion: The studied chest radiography foundation model demonstrated racial and sex-related bias, which led to disparate performance across patient subgroups; thus, this model may be unsafe for clinical applications.Keywords: Conventional Radiography, Computer Application-Detection/Diagnosis, Chest Radiography, Bias, Foundation Models Supplemental material is available for this article. Published under a CC BY 4.0 license.See also commentary by Czum and Parr in this issue.
RESUMO
Rationale and objectives: To develop a method for automatic localisation of brain lesions on head CT, suitable for both population-level analysis and lesion management in a clinical setting. Materials and methods: Lesions were located by mapping a bespoke CT brain atlas to the patient's head CT in which lesions had been previously segmented. The atlas mapping was achieved through robust intensity-based registration enabling the calculation of per-region lesion volumes. Quality control (QC) metrics were derived for automatic detection of failure cases. The CT brain template was built using 182 non-lesioned CT scans and an iterative template construction strategy. Individual brain regions in the CT template were defined via non-linear registration of an existing MRI-based brain atlas.Evaluation was performed on a multi-centre traumatic brain injury dataset (TBI) (n = 839 scans), including visual inspection by a trained expert. Two population-level analyses are presented as proof-of-concept: a spatial assessment of lesion prevalence, and an exploration of the distribution of lesion volume per brain region, stratified by clinical outcome. Results: 95.7% of the lesion localisation results were rated by a trained expert as suitable for approximate anatomical correspondence between lesions and brain regions, and 72.5% for more quantitatively accurate estimates of regional lesion load. The classification performance of the automatic QC showed an AUC of 0.84 when compared to binarised visual inspection scores. The localisation method has been integrated into the publicly available Brain Lesion Analysis and Segmentation Tool for CT (BLAST-CT). Conclusion: Automatic lesion localisation with reliable QC metrics is feasible and can be used for patient-level quantitative analysis of TBI, as well as for large-scale population analysis due to its computational efficiency (<2 min/scan on GPU).
RESUMO
The prediction of functional outcome after mild traumatic brain injury (mTBI) is challenging. Conventional magnetic resonance imaging (MRI) does not do a good job of explaining the variance in outcome, as many patients with incomplete recovery will have normal-appearing clinical neuroimaging. More advanced quantitative techniques such as diffusion MRI (dMRI), can detect microstructural changes not otherwise visible, and so may offer a way to improve outcome prediction. In this study, we explore the potential of linear support vector classifiers (linearSVCs) to identify dMRI biomarkers that can predict recovery after mTBI. Simultaneously, the harmonization of fractional anisotropy (FA) and mean diffusivity (MD) via ComBat was evaluated and compared for the classification performances of the linearSVCs. We included dMRI scans of 179 mTBI patients and 85 controls from the Collaborative European NeuroTrauma Effectiveness Research in Traumatic Brain Injury (CENTER-TBI), a multi-center prospective cohort study, up to 21 days post-injury. Patients were dichotomized according to their Extended Glasgow Outcome Scale (GOSE) scores at 6 months into complete (n = 92; GOSE = 8) and incomplete (n = 87; GOSE <8) recovery. FA and MD maps were registered to a common space and harmonized via the ComBat algorithm. LinearSVCs were applied to distinguish: (1) mTBI patients from controls and (2) mTBI patients with complete from those with incomplete recovery. The linearSVCs were trained on (1) age and sex only, (2) non-harmonized, (3) two-category-harmonized ComBat, and (4) three-category-harmonized ComBat FA and MD images combined with age and sex. White matter FA and MD voxels and regions of interest (ROIs) within the John Hopkins University (JHU) atlas were examined. Recursive feature elimination was used to identify the 10% most discriminative voxels or the 10 most discriminative ROIs for each implementation. mTBI patients displayed significantly higher MD and lower FA values than controls for the discriminative voxels and ROIs. For the analysis between mTBI patients and controls, the three-category-harmonized ComBat FA and MD voxel-wise linearSVC provided significantly higher classification scores (81.4% accuracy, 93.3% sensitivity, 80.3% F1-score, and 0.88 area under the curve [AUC], p < 0.05) compared with the classification based on age and sex only and the ROI approaches (accuracies: 59.8% and 64.8%, respectively). Similar to the analysis between mTBI patients and controls, the three-category-harmonized ComBat FA and MD maps voxelwise approach yields statistically significant prediction scores between mTBI patients with complete and those with incomplete recovery (71.8% specificity, 66.2% F1-score and 0.71 AUC, p < 0.05), which provided a modest increase in the classification score (accuracy: 66.4%) compared with the classification based on age and sex only and ROI-wise approaches (accuracy: 61.4% and 64.7%, respectively). This study showed that ComBat harmonized FA and MD may provide additional information for diagnosis and prognosis of mTBI in a multi-modal machine learning approach. These findings demonstrate that dMRI may assist in the early detection of patients at risk of incomplete recovery from mTBI.
Assuntos
Concussão Encefálica , Lesões Encefálicas Traumáticas , Humanos , Concussão Encefálica/diagnóstico , Imagem de Tensor de Difusão/métodos , Máquina de Vetores de Suporte , Estudos Prospectivos , Prognóstico , Anisotropia , Encéfalo/patologiaRESUMO
OBJECTIVES: The UK Biobank (UKBB) and German National Cohort (NAKO) are among the largest cohort studies, capturing a wide range of health-related data from the general population, including comprehensive magnetic resonance imaging (MRI) examinations. The purpose of this study was to demonstrate how MRI data from these large-scale studies can be jointly analyzed and to derive comprehensive quantitative image-based phenotypes across the general adult population. MATERIALS AND METHODS: Image-derived features of abdominal organs (volumes of liver, spleen, kidneys, and pancreas; volumes of kidney hilum adipose tissue; and fat fractions of liver and pancreas) were extracted from T1-weighted Dixon MRI data of 17,996 participants of UKBB and NAKO based on quality-controlled deep learning generated organ segmentations. To enable valid cross-study analysis, we first analyzed the data generating process using methods of causal discovery. We subsequently harmonized data from UKBB and NAKO using the ComBat approach for batch effect correction. We finally performed quantile regression on harmonized data across studies providing quantitative models for the variation of image-derived features stratified for sex and dependent on age, height, and weight. RESULTS: Data from 8791 UKBB participants (49.9% female; age, 63 ± 7.5 years) and 9205 NAKO participants (49.1% female, age: 51.8 ± 11.4 years) were analyzed. Analysis of the data generating process revealed direct effects of age, sex, height, weight, and the data source (UKBB vs NAKO) on image-derived features. Correction of data source-related effects resulted in markedly improved alignment of image-derived features between UKBB and NAKO. Cross-study analysis on harmonized data revealed comprehensive quantitative models for the phenotypic variation of abdominal organs across the general adult population. CONCLUSIONS: Cross-study analysis of MRI data from UKBB and NAKO as proposed in this work can be helpful for future joint data analyses across cohorts linking genetic, environmental, and behavioral risk factors to MRI-derived phenotypes and provide reference values for clinical diagnostics.
Assuntos
Bancos de Espécimes Biológicos , Imageamento por Ressonância Magnética , Humanos , Feminino , Masculino , Imageamento por Ressonância Magnética/métodos , Estudos de Coortes , Abdome/diagnóstico por imagem , Reino UnidoRESUMO
There are many ways to acquire and process diffusion MRI (dMRI) data for group studies, but it is unknown which maximizes the sensitivity to white matter (WM) pathology. Inspired by this question, we analyzed data acquired for diffusion tensor imaging (DTI) and diffusion kurtosis imaging (DKI) at 3T (3T-DTI and 3T-DKI) and DTI at 7T in patients with systemic lupus erythematosus (SLE) and healthy controls (HC). Parameter estimates in 72 WM tracts were obtained using TractSeg. The impact on the sensitivity to WM pathology was evaluated for the diffusion protocol, the magnetic field strength, and the processing pipeline. Sensitivity was quantified in terms of Cohen's d for group comparison. Results showed that the choice of diffusion protocol had the largest impact on the effect size. The effect size in fractional anisotropy (FA) across all WM tracts was 0.26 higher when derived by DTI than by DKI and 0.20 higher in 3T compared with 7T. The difference due to the diffusion protocol was larger than the difference due to magnetic field strength for the majority of diffusion parameters. In contrast, the difference between including or excluding different processing steps was near negligible, except for the correction of distortions from eddy currents and motion which had a clearly positive impact. For example, effect sizes increased on average by 0.07 by including motion and eddy correction for FA derived from 3T-DTI. Effect sizes were slightly reduced by the incorporation of denoising and Gibbs-ringing removal (on average by 0.011 and 0.005, respectively). Smoothing prior to diffusion model fitting generally reduced effect sizes. In summary, 3T-DTI in combination with eddy current and motion correction yielded the highest sensitivity to WM pathology in patients with SLE. However, our results also indicated that the 3T-DKI and 7T-DTI protocols used here may be adjusted to increase effect sizes.
RESUMO
Background: The growth in multi-center neuroimaging studies generated a need for methods that mitigate the differences in hardware and acquisition protocols across sites i.e., scanner effects. ComBat harmonization methods have shown promise but have not yet been tested on all the data types commonly studied with magnetic resonance imaging (MRI). This study aimed to validate neuroCombat, longCombat and gamCombat on both structural and diffusion metrics in both cross-sectional and longitudinal data. Methods: We used a travelling subject design whereby 73 healthy volunteers contributed 161 scans across two sites and four machines using one T1 and five diffusion MRI protocols. Scanner was defined as a composite of site, machine and protocol. A common pipeline extracted two structural metrics (volumes and cortical thickness) and two diffusion tensor imaging metrics (mean diffusivity and fractional anisotropy) for seven regions of interest including gray and (except for cortical thickness) white matter regions. Results: Structural data exhibited no significant scanner effect and therefore did not benefit from harmonization in our particular cohort. Indeed, attempting harmonization obscured the true biological effect for some regions of interest. Diffusion data contained marked scanner effects and was successfully harmonized by all methods, resulting in smaller scanner effects and better detection of true biological effects. LongCombat less effectively reduced the scanner effect for cross-sectional white matter data but had a slightly lower probability of incorrectly finding group differences in simulations, compared to neuroCombat and gamCombat. False positive rates for all methods and all metrics did not significantly exceed 5%. Conclusions: Statistical harmonization of structural data is not always necessary and harmonization in the absence of a scanner effect may be harmful. Harmonization of diffusion MRI data is highly recommended with neuroCombat, longCombat and gamCombat performing well in cross-sectional and longitudinal settings.
RESUMO
INTRODUCTION: Whole-body MRI (WB-MRI) is recommended by the National Institute of Clinical Excellence as the first-line imaging tool for diagnosis of multiple myeloma. Reporting WB-MRI scans requires expertise to interpret and can be challenging for radiologists who need to meet rapid turn-around requirements. Automated computational tools based on machine learning (ML) could assist the radiologist in terms of sensitivity and reading speed and would facilitate improved accuracy, productivity and cost-effectiveness. The MALIMAR study aims to develop and validate a ML algorithm to increase the diagnostic accuracy and reading speed of radiological interpretation of WB-MRI compared with standard methods. METHODS AND ANALYSIS: This phase II/III imaging trial will perform retrospective analysis of previously obtained clinical radiology MRI scans and scans from healthy volunteers obtained prospectively to implement training and validation of an ML algorithm. The study will comprise three project phases using approximately 633 scans to (1) train the ML algorithm to identify active disease, (2) clinically validate the ML algorithm and (3) determine change in disease status following treatment via a quantification of burden of disease in patients with myeloma. Phase 1 will primarily train the ML algorithm to detect active myeloma against an expert assessment ('reference standard'). Phase 2 will use the ML output in the setting of radiology reader study to assess the difference in sensitivity when using ML-assisted reading or human-alone reading. Phase 3 will assess the agreement between experienced readers (with and without ML) and the reference standard in scoring both overall burden of disease before and after treatment, and response. ETHICS AND DISSEMINATION: MALIMAR has ethical approval from South Central-Oxford C Research Ethics Committee (REC Reference: 17/SC/0630). IRAS Project ID: 233501. CPMS Portfolio adoption (CPMS ID: 36766). Participants gave informed consent to participate in the study before taking part. MALIMAR is funded by National Institute for Healthcare Research Efficacy and Mechanism Evaluation funding (NIHR EME Project ID: 16/68/34). Findings will be made available through peer-reviewed publications and conference dissemination. TRIAL REGISTRATION NUMBER: NCT03574454.
Assuntos
Aprendizado de Máquina , Imageamento por Ressonância Magnética , Mieloma Múltiplo , Imagem Corporal Total , Clorobenzenos , Ensaios Clínicos Fase II como Assunto , Ensaios Clínicos Fase III como Assunto , Estudos Transversais , Testes Diagnósticos de Rotina , Humanos , Imageamento por Ressonância Magnética/métodos , Mieloma Múltiplo/diagnóstico por imagem , Mieloma Múltiplo/terapia , Estudos Retrospectivos , Sulfetos , Imagem Corporal Total/métodosRESUMO
BACKGROUND: The thalamus seems to be important in the development of postoperative delirium (POD) as previously revealed by volumetric and diffusion magnetic resonance imaging. In this observational cohort study, we aimed to further investigate the impact of the microstructural integrity of the thalamus and thalamic nuclei on the incidence of POD by applying diffusion kurtosis imaging (DKI). METHODS: Older patients without dementia (≥65 years) who were scheduled for major elective surgery received preoperative DKI at two study centres. The DKI metrics fractional anisotropy (FA), mean diffusivity (MD), mean kurtosis (MK) and free water (FW) were calculated for the thalamus and - as secondary outcome - for eight predefined thalamic nuclei and regions. Low FA and MK and, conversely, high MD and FW, indicate aspects of microstructural abnormality. To assess patients' POD status, the Nursing Delirium Screening Scale (Nu-DESC), Richmond Agitation Sedation Scale (RASS), Confusion Assessment Method (CAM) and Confusion Assessment Method for the Intensive Care Unit score (CAM-ICU) and chart review were applied twice a day after surgery for the duration of seven days or until discharge. For each metric and each nucleus, logistic regression was performed to assess the risk of POD. RESULTS: This analysis included the diffusion scans of 325 patients, of whom 53 (16.3 %) developed POD. Independently of age, sex and study centre, thalamic MD was statistically significantly associated with POD [OR 1.65 per SD increment (95 %CI 1.17 - 2.34) p = 0.004]. FA (p = 0.84), MK (p = 0.41) and FW (p = 0.06) were not significantly associated with POD in the examined sample. Exploration of thalamic nuclei also indicated that only the MD in certain areas of the thalamus was associated with POD. MD was increased in bilateral hemispheres, pulvinar nuclei, mediodorsal nuclei and the left anterior nucleus. CONCLUSIONS: Microstructural abnormalities of the thalamus and thalamic nuclei, as reflected by increased MD, appear to predispose to POD. These findings affirm the thalamus as a region of interest in POD research.
Assuntos
Delírio do Despertar , Humanos , Idoso , Estudos de Coortes , Estudos Prospectivos , Imagem de Tensor de Difusão/métodos , Núcleos Talâmicos , Tálamo/diagnóstico por imagemRESUMO
Large epidemiological studies such as the UK Biobank (UKBB) or German National Cohort (NAKO) provide unprecedented health-related data of the general population aiming to better understand determinants of health and disease. As part of these studies, Magnetic Resonance Imaging (MRI) is performed in a subset of participants allowing for phenotypical and functional characterization of different organ systems. Due to the large amount of imaging data, automated image analysis is required, which can be performed using deep learning methods, e. g. for automated organ segmentation. In this paper we describe a computational pipeline for automated segmentation of abdominal organs on MRI data from 20,000 participants of UKBB and NAKO and provide results of the quality control process. We found that approx. 90% of data sets showed no relevant segmentation errors while relevant errors occurred in a varying proportion of data sets depending on the organ of interest. Image-derived features based on automated organ segmentations showed relevant deviations of varying degree in the presence of segmentation errors. These results show that large-scale, deep learning-based abdominal organ segmentation on MRI data is feasible with overall high accuracy, but visual quality control remains an important step ensuring the validity of down-stream analyses in large epidemiological imaging studies.