There has been extensive interest in promoting gender equality within radiology, a predominately male field. In this study, our aim was to quantify gender representation in neuroradiology faculty ...rankings and determine any related factors that may contribute to any such disparity.
We evaluated the academic and administrative faculty members of neuroradiology divisions for all on-line listed programs in the US and Canada. After excluding programs that did not fulfill our selection criteria, we generated a short list of 85 US and 8 Canadian programs. We found 465 faculty members who met the inclusion criteria for our study. We used Elsevier's SCOPUS for gathering the data pertaining to the publications, H-index, citations, and tenure of the productivity of each faculty member.
Gender disparity was insignificant when analyzing academic ranks. There are more men working in neuroimaging relative to women (χ
= 0.46;
= .79). However, gender disparity was highly significant for leadership positions in neuroradiology (χ
= 6.76;
= .009). The median H-index was higher among male faculty members (17.5) versus female faculty members (9). Female faculty members have odds of 0.84 compared with male faculty members of having a higher H-index, adjusting for publications, citations, academic ranks, leadership ranks, and interaction between gender and publications and gender and citations (9).
Neuroradiology faculty members follow the same male predominance seen in many other specialties of medicine. In this study, issues such as mentoring, role models, opportunities to engage in leadership/research activities, funding opportunities, and mindfulness regarding research productivity are explored.
•It is the first time to propose a Genetic algorithm (GA)-based method to construct CNN structures for automatic medical image analysis.•We optimize the standard genetic algorithm and employ transfer ...learning to speed up the evolutionary process.•The proposed framework can flexibly treat various CNN components as genes in the evolutionary process to obtain the state-of-the-art performance.•Our proposed framework is extendable to other deep learning based medical applications, such as lesion detection and tumor segmentation.
Display omitted
Deep convolutional neural networks offer state-of-the-art performance for medical image analysis. However, their architectures are manually designed for particular problems. On the one hand, a manual designing process requires many trials to tune a large number of hyperparameters and is thus quite a time-consuming task. On the other hand, the fittest hyperparameters that can adapt to source data properties (e.g., sparsity, noisy features) are not able to be quickly identified for target data properties. For instance, the realistic noise in medical images is usually mixed and complicated, and sometimes unknown, leading to challenges in applying existing methods directly and creating effective denoising neural networks easily. In this paper, we present a Genetic Algorithm (GA)-based network evolution approach to search for the fittest genes to optimize network structures automatically. We expedite the evolutionary process through an experience-based greedy exploration strategy and transfer learning. Our evolutionary algorithm procedure has flexibility, which allows taking advantage of current state-of-the-art modules (e.g., residual blocks) to search for promising neural networks. We evaluate our framework on a classic medical image analysis task: denoising. The experimental results on computed tomography perfusion (CTP) image denoising demonstrate the capability of the method to select the fittest genes for building high-performance networks, named EvoNets. Our results outperform state-of-the-art methods consistently at various noise levels.
Evidence-based medicine has emerged as a valuable tool to guide clinical decision-making, by summarizing the best possible evidence for both diagnostic and treatment strategies. Imaging plays a ...critical role in the evaluation and treatment of patients with acute ischemic stroke, especially those who are being considered for thrombolytic or endovascular therapy. Time from stroke-symptom onset to treatment is a strong predictor of long-term functional outcome after stroke. Therefore, imaging and treatment decisions must occur rapidly in this setting, while minimizing unnecessary delays in treatment. The aim of this review was to summarize the best available evidence for the diagnostic and therapeutic management of patients with acute ischemic stroke.
Impairments in cerebrovascular reserve (CVR) have been variably associated with increased risk of ischemic events and may stratify stroke risk in patients with high-grade internal carotid artery ...stenosis or occlusion. The purpose of this study is to perform a systematic review and meta-analysis to summarize the association of CVR impairment and stroke risk.
We performed a literature search evaluating the association of impairments in CVR with future stroke or transient ischemic attack in patients with high-grade internal carotid artery stenosis or occlusion. We included studies with a minimum of 1-year patient follow-up with baseline CVR measures performed by any modality and primary outcome measures of stroke and/or transient ischemic attack. A meta-analysis with assessment of study heterogeneity and publication bias was performed. Results were presented in a forest plot and summarized using a random-effects model.
Thirteen studies met the inclusion criteria, representing a total of 1061 independent CVR tests in 991 unique patients with a mean follow-up of 32.7 months. We found a significant positive relationship between impairment of CVR and development of stroke with a pooled random effects OR of 3.86 (95% CI, 1.99-7.48). Subset analysis showed that this association between CVR impairment and future risk of stroke/transient ischemic attack remained significant regardless of ischemic outcome measure, symptomatic or asymptomatic disease, stenosis or occlusion, or CVR testing method.
CVR impairment is strongly associated with increased risk of ischemic events in carotid stenosis or occlusion and may be useful for stroke risk stratification.
Hematoma expansion (HE) is a modifiable risk factor and a potential treatment target in patients with intracerebral hemorrhage (ICH). We aimed to train and validate deep-learning models for ...high-confidence prediction of supratentorial ICH expansion, based on admission non-contrast head Computed Tomography (CT). Applying Monte Carlo dropout and entropy of deep-learning model predictions, we estimated the model uncertainty and identified patients at high risk of HE with high confidence. Using the receiver operating characteristics area under the curve (AUC), we compared the deep-learning model prediction performance with multivariable models based on visual markers of HE determined by expert reviewers. We randomly split a multicentric dataset of patients (4-to-1) into training/cross-validation (n = 634) versus test (n = 159) cohorts. We trained and tested separate models for prediction of ≥6 mL and ≥3 mL ICH expansion. The deep-learning models achieved an AUC = 0.81 for high-confidence prediction of HE
and AUC = 0.80 for prediction of HE
, which were higher than visual maker models AUC = 0.69 for HE
(p = 0.036) and AUC = 0.68 for HE
(p = 0.043). Our results show that fully automated deep-learning models can identify patients at risk of supratentorial ICH expansion based on admission non-contrast head CT, with high confidence, and more accurately than benchmark visual markers.
A major driver of individual variation in long-term outcomes following a large vessel occlusion (LVO) stroke is the degree of collateral arterial circulation. We aimed to develop and evaluate ...machine-learning models that quantify LVO collateral status using admission computed tomography angiography (CTA) radiomics.
We extracted 1116 radiomic features from the anterior circulation territories from admission CTAs of 600 patients experiencing an acute LVO stroke. We trained and validated multiple machine-learning models for the prediction of collateral status based on consensus from two neuroradiologists as ground truth. Models were first trained to predict (1) good vs. intermediate or poor, or (2) good vs. intermediate or poor collateral status. Then, model predictions were combined to determine a three-tier collateral score (good, intermediate, or poor). We used the receiver operating characteristics area under the curve (AUC) to evaluate prediction accuracy.
We included 499 patients in training and 101 in an independent test cohort. The best-performing models achieved an averaged cross-validation AUC of 0.80 ± 0.05 for poor vs. intermediate/good collateral and 0.69 ± 0.05 for good vs. intermediate/poor, and AUC = 0.77 (0.67-0.87) and AUC = 0.78 (0.70-0.90) in the independent test cohort, respectively. The collateral scores predicted by the radiomics model were correlated with (rho = 0.45,
= 0.002) and were independent predictors of 3-month clinical outcome (
= 0.018) in the independent test cohort.
Automated tools for the assessment of collateral status from admission CTA-such as the radiomics models described here-can generate clinically relevant and reproducible collateral scores to facilitate a timely treatment triage in patients experiencing an acute LVO stroke.
Abnormalities in apparent diffusion coefficient (ADC), fractional anisotropy (FA), and mean diffusivity (MD) values can be used to assess microstructural damage to white matter tracts and could ...represent a quantitative marker of chronic ischemia and thereby potentially serve as a stroke risk factor or a measure of existing subclinical ischemic disease burden. We performed a systematic review and 3 separate meta-analyses to evaluate the association between unilateral carotid steno-occlusion and ipsilateral ADC, FA, or MD abnormality.
A comprehensive literature search evaluating the association of carotid disease and quantitative white matter diffusion imaging was performed. The included studies examined patients for ADC, FA, and MD values ipsilateral and contralateral to the site of carotid artery disease. Three meta-analyses using standardized mean differences with assessment of study heterogeneity were performed.
Of the 2,920 manuscripts screened, 6 met eligibility for meta-analysis. Of the included manuscripts, 2 studied ADC values, 6 studied FA values, and 2 studied MD values. Our 3 meta-analyses showed standardized mean difference for ADC, FA, and MD values between cerebral hemispheres ipsilateral and contralateral to carotid artery disease site as 1.13 (95% CI: .79-1.47, P < .001), -.42 (95% CI: -.62 to -.21, P < .001), and .23 (95% CI: -.32 to -.77, P = .41), respectively. Measures of heterogeneity showed mild heterogeneity in the 3 meta-analyses.
Carotid artery disease is associated with significant ADC and FA value changes, suggesting that carotid disease is associated with quantifiable white matter microstructural damage.
Quantitative analysis of medical images can provide objective tools for diagnosis, prognostication, and disease monitoring. Radiomics refers to automated extraction of a large number of quantitative ...features from medical images for characterization of underlying pathologies. In this review, we will discuss the principles of radiomics, image preprocessing, feature extraction workflow, and statistical analysis. We will also address the limitations and future directions of radiomics.
Display omitted
► Dictionaries are learned from high-dose CTP data for low-dose CBF estimation. ► Temporal convolution model is combined with spatial dictionary mapping prior. ► Evaluation on in vivo ...aneurysmal subarachnoid hemorrhage and normal patients. ► Outperform existing methods in CBF estimation for low-dose CTP data. ► Improve the differentiation between ischemic and normal tissues in the brain.
Computed tomography perfusion (CTP) is an important functional imaging modality in the evaluation of cerebrovascular diseases, particularly in acute stroke and vasospasm. However, the post-processed parametric maps of blood flow tend to be noisy, especially in low-dose CTP, due to the noisy contrast enhancement profile and the oscillatory nature of the results generated by the current computational methods. In this paper, we propose a robust sparse perfusion deconvolution method (SPD) to estimate cerebral blood flow in CTP performed at low radiation dose. We first build a dictionary from high-dose perfusion maps using online dictionary learning and then perform deconvolution-based hemodynamic parameters estimation on the low-dose CTP data. Our method is validated on clinical data of patients with normal and pathological CBF maps. The results show that we achieve superior performance than existing methods, and potentially improve the differentiation between normal and ischemic tissue in the brain.