3D CT-Inclusive Deep-Learning Model to Predict Mortality, ICU Admittance, and Intubation in COVID-19 Patients

Chest CT is a useful initial exam in patients with coronavirus disease 2019 (COVID-19) for assessing lung damage. AI-powered predictive models could be useful to better allocate resources in the midst of the pandemic. Our aim was to build a deep-learning (DL) model for COVID-19 outcome prediction inclusive of 3D chest CT images acquired at hospital admission. This retrospective multicentric study included 1051 patients (mean age 69, SD = 15) who presented to the emergency department of three different institutions between 20th March 2020 and 20th January 2021 with COVID-19 confirmed by real-time reverse transcriptase polymerase chain reaction (RT-PCR). Chest CT at hospital admission were evaluated by a 3D residual neural network algorithm. Training, internal validation, and external validation groups included 608, 153, and 290 patients, respectively. Images, clinical, and laboratory data were fed into different customizations of a dense neural network to choose the best performing architecture for the prediction of mortality, intubation, and intensive care unit (ICU) admission. The AI model tested on CT and clinical features displayed accuracy, sensitivity, specificity, and ROC-AUC, respectively, of 91.7%, 90.5%, 92.4%, and 95% for the prediction of patient's mortality; 91.3%, 91.5%, 89.8%, and 95% for intubation; and 89.6%, 90.2%, 86.5%, and 94% for ICU admission (internal validation) in the testing cohort. The performance was lower in the validation cohort for mortality (71.7%, 55.6%, 74.8%, 72%), intubation (72.6%, 74.7%, 45.7%, 64%), and ICU admission (74.7%, 77%, 46%, 70%) prediction. The addition of the available laboratory data led to an increase in sensitivity for patient's mortality (66%) and specificity for intubation and ICU admission (50%, 52%, respectively), while the other metrics maintained similar performance results. We present a deep-learning model to predict mortality, ICU admittance, and intubation in COVID-19 patients. KEY POINTS: • 3D CT-based deep learning model predicted the internal validation set with high accuracy, sensibility and specificity (> 90%) mortality, ICU admittance, and intubation in COVID-19 patients. • The model slightly increased prediction results when laboratory data were added to the analysis, despite data imbalance. However, the model accuracy dropped when CT images were not considered in the analysis, implying an important role of CT in predicting outcomes.

Diffusion Weighted Imaging in Neuro-Oncology: Diagnosis, Post-Treatment Changes, and Advanced Sequences-An Updated Review

DWI is an imaging technique commonly used for the assessment of acute ischemia, inflammatory disorders, and CNS neoplasia. It has several benefits since it is a quick, easily replicable sequence that is widely used on many standard scanners. In addition to its normal clinical purpose, DWI offers crucial functional and physiological information regarding brain neoplasia and the surrounding milieu. A narrative review of the literature was conducted based on the PubMed database with the purpose of investigating the potential role of DWI in the neuro-oncology field. A total of 179 articles were included in the study.

Synthetic Post-Contrast Imaging through Artificial Intelligence: Clinical Applications of Virtual and Augmented Contrast Media

Contrast media are widely diffused in biomedical imaging, due to their relevance in the diagnosis of numerous disorders. However, the risk of adverse reactions, the concern of potential damage to sensitive organs, and the recently described brain deposition of gadolinium salts, limit the use of contrast media in clinical practice. In recent years, the application of artificial intelligence (AI) techniques to biomedical imaging has led to the development of 'virtual' and 'augmented' contrasts. The idea behind these applications is to generate synthetic post-contrast images through AI computational modeling starting from the information available on other images acquired during the same scan. In these AI models, non-contrast images (virtual contrast) or low-dose post-contrast images (augmented contrast) are used as input data to generate synthetic post-contrast images, which are often undistinguishable from the native ones. In this review, we discuss the most recent advances of AI applications to biomedical imaging relative to synthetic contrast media.

Understanding Language Reorganization With Neuroimaging: How Language Adapts to Different Focal Lesions and Insights Into Clinical Applications

When the language-dominant hemisphere is damaged by a focal lesion, the brain may reorganize the language network through functional and structural changes known as adaptive plasticity. Adaptive plasticity is documented for triggers including ischemic, tumoral, and epileptic focal lesions, with effects in clinical practice. Many questions remain regarding language plasticity. Different lesions may induce different patterns of reorganization depending on pathologic features, location in the brain, and timing of onset. Neuroimaging provides insights into language plasticity due to its non-invasiveness, ability to image the whole brain, and large-scale implementation. This review provides an overview of language plasticity on MRI with insights for patient care. First, we describe the structural and functional language network as depicted by neuroimaging. Second, we explore language reorganization triggered by stroke, brain tumors, and epileptic lesions and analyze applications in clinical diagnosis and treatment planning. By comparing different focal lesions, we investigate determinants of language plasticity including lesion location and timing of onset, longitudinal evolution of reorganization, and the relationship between structural and functional changes.

Magnetic resonance imaging of pilocytic astrocytomas in adults with histopathologic correlation: a report of six consecutive cases

Pilocytic astrocytoma is a WHO grade I tumor usually diagnosed in pediatric patients, and rarely encountered in the adult population. Therefore, available information about the magnetic resonance imaging characteristics of adult pilocytic astrocytoma is scarce. We report on the MRI features and corresponding histopathologic findings of six consecutive aPA cases diagnosed. The tumors were encountered in both infra- and supratentorial compartments, and their MRI characteristics were quite heterogeneous. Features included the typical solid-cystic appearance located in the cerebellum as well as the relatively unusual multifocal and/or hemorrhagic features located intra-ventricularly. The aPA MRI characteristics are remarkably variable, and might mimic those of higher grade tumors in adult patients.

AI and High-Grade Glioma for Diagnosis and Outcome Prediction: Do All Machine Learning Models Perform Equally Well?

Radiomic models outperform clinical data for outcome prediction in high-grade gliomas (HGG). However, lack of parameter standardization limits clinical applications. Many machine learning (ML) radiomic models employ single classifiers rather than ensemble learning, which is known to boost performance, and comparative analyses are lacking in the literature. We aimed to compare ML classifiers to predict clinically relevant tasks for HGG: overall survival (OS), isocitrate dehydrogenase (IDH) mutation, O-6-methylguanine-DNA-methyltransferase (MGMT) promoter methylation, epidermal growth factor receptor vIII (EGFR) amplification, and Ki-67 expression, based on radiomic features from conventional and advanced magnetic resonance imaging (MRI). Our objective was to identify the best algorithm for each task. One hundred fifty-six adult patients with pathologic diagnosis of HGG were included. Three tumoral regions were manually segmented: contrast-enhancing tumor, necrosis, and non-enhancing tumor. Radiomic features were extracted with a custom version of Pyradiomics and selected through Boruta algorithm. A Grid Search algorithm was applied when computing ten times K-fold cross-validation (K=10) to get the highest mean and lowest spread of accuracy. Model performance was assessed as AUC-ROC curve mean values with 95% confidence intervals (CI). Extreme Gradient Boosting (xGB) obtained highest accuracy for OS (74,5%), Adaboost (AB) for IDH mutation (87.5%), MGMT methylation (70,8%), Ki-67 expression (86%), and EGFR amplification (81%). Ensemble classifiers showed the best performance across tasks. High-scoring radiomic features shed light on possible correlations between MRI and tumor histology.

COVID Mortality Prediction with Machine Learning Methods: A Systematic Review and Critical Appraisal

More than a year has passed since the report of the first case of coronavirus disease 2019 (COVID), and increasing deaths continue to occur. Minimizing the time required for resource allocation and clinical decision making, such as triage, choice of ventilation modes and admission to the intensive care unit is important. Machine learning techniques are acquiring an increasingly sought-after role in predicting the outcome of COVID patients. Particularly, the use of baseline machine learning techniques is rapidly developing in COVID mortality prediction, since a mortality prediction model could rapidly and effectively help clinical decision-making for COVID patients at imminent risk of death. Recent studies reviewed predictive models for SARS-CoV-2 diagnosis, severity, length of hospital stay, intensive care unit admission or mechanical ventilation modes outcomes; however, systematic reviews focused on prediction of COVID mortality outcome with machine learning methods are lacking in the literature. The present review looked into the studies that implemented machine learning, including deep learning, methods in COVID mortality prediction thus trying to present the existing published literature and to provide possible explanations of the best results that the studies obtained. The study also discussed challenging aspects of current studies, providing suggestions for future developments.

Glioblastoma radiomics to predict survival: Diffusion characteristics of surrounding nonenhancing tissue to select patients for extensive resection

BACKGROUND AND PURPOSE

Glioblastoma (GBM) is an aggressive primary CNS neoplasm with poor overall survival (OS) despite standard of care. On MRI, GBM is usually characterized by an enhancing portion (CET) (surgery target) and a nonenhancing surrounding (NET). Extent of resection is a long debated issue in GBM, with recent evidence suggesting that both CET and NET should be resected in <65 years old patients, regardless of other risk factors (i.e., molecular biomarkers). Our aim was to test a radiomic model for patient survival stratification in <65 years old patients, by analyzing MRI features of NET, to aid tumor resection.

METHODS

Sixty-eight <65 years old GBM patients, with extensive CET resection, were selected. Resection was evaluated by manually segmenting CET on volumetric T1-weighted MRI pre and postsurgery (within 72 h). All patients underwent the same treatment protocol including chemoradiation. NET radiomic features were extracted with a custom version of Pyradiomics. Feature selection was performed with principal component analysis (PCA) and its effect on survival tested with Cox regression model. Twelve months OS discrimination was tested by t-test followed by logistic regression. Statistical significance was set at p<0.05. The most relevant features were identified from the component matrix.

RESULTS

Five PCA components (PC1-5) explained 90% of the variance. PC5 resulted significant in the Cox model (p = 0.002; exp(B) = 0.686), at t-test (p = 0.002) and logistic regression analysis (p = 0.006). Apparent diffusion coefficient (ADC)-based features were the most significant for patient survival stratification.

CONCLUSIONS

ADC radiomic features on NET predict survival after standard therapy and could be used to improve patient selection for more extensive surgery.

Deep Learning Can Differentiate IDH-Mutant from IDH-Wild GBM

Isocitrate dehydrogenase (IDH) mutant and wildtype glioblastoma multiforme (GBM) often show overlapping features on magnetic resonance imaging (MRI), representing a diagnostic challenge. Deep learning showed promising results for IDH identification in mixed low/high grade glioma populations; however, a GBM-specific model is still lacking in the literature. Our aim was to develop a GBM-tailored deep-learning model for IDH prediction by applying convoluted neural networks (CNN) on multiparametric MRI. We selected 100 adult patients with pathologically demonstrated WHO grade IV gliomas and IDH testing. MRI sequences included: MPRAGE, T1, T2, FLAIR, rCBV and ADC. The model consisted of a 4-block 2D CNN, applied to each MRI sequence. Probability of IDH mutation was obtained from the last dense layer of a softmax activation function. Model performance was evaluated in the test cohort considering categorical cross-entropy loss (CCEL) and accuracy. Calculated performance was: rCBV (accuracy 83%, CCEL 0.64), T1 (accuracy 77%, CCEL 1.4), FLAIR (accuracy 77%, CCEL 1.98), T2 (accuracy 67%, CCEL 2.41), MPRAGE (accuracy 66%, CCEL 2.55). Lower performance was achieved on ADC maps. We present a GBM-specific deep-learning model for IDH mutation prediction, with a maximal accuracy of 83% on rCBV maps. Highest predictivity achieved on perfusion images possibly reflects the known link between IDH and neoangiogenesis through the hypoxia inducible factor.

Non-congenital viral infections of the central nervous system: from the immunocompetent to the immunocompromised child

Non-congenital viral infections of the central nervous system in children can represent a severe clinical condition that needs a prompt diagnosis and management. However, the aetiological diagnosis can be challenging because symptoms are often nonspecific and cerebrospinal fluid analysis is not always diagnostic. In this context, neuroimaging represents a helpful tool, even though radiologic patterns sometimes overlap. The purpose of this pictorial essay is to suggest a schematic representation of different radiologic patterns of non-congenital viral encephalomyelitis based on the predominant viral tropism and vulnerability of specific regions: cortical grey matter, deep grey matter, white matter, brainstem, cerebellum and spine.

Arterial Spin Labeling MRI in Carotid Stenosis: Arterial Transit Artifacts May Predict Symptoms

BackgroundStenosis of the internal carotid artery has a higher risk for stroke. Many investigations have focused on structure and plaque composition as signs of plaque vulnerability, but few studies have analyzed hemodynamic changes in the brain as a risk factor.PurposeTo use 3-T MRI methods including contrast material-enhanced MR angiography, carotid plaque imaging, and arterial spin labeling (ASL) to identify imaging parameters that best help distinguish between asymptomatic and symptomatic participants with carotid stenosis.Materials and MethodsParticipants with carotid stenosis from two ongoing prospective studies who underwent ASL and carotid plaque imaging with use of 3-T MRI in the same setting from 2014 to 2018 were studied. Participants were assessed clinically for recent symptoms (transient ischemic attack or stroke) and divided equally into symptomatic and nonsymptomatic groups. Reviewers were blinded to the symptomatic status and MRI scans were analyzed for the degree of stenosis, plaque surface structure, presence of intraplaque hemorrhage (IPH), circle of Willis collaterals, and the presence and severity of arterial transit artifacts (ATAs) at ASL imaging. MRI findings were correlated with symptomatic status by using t tests and the Fisher exact test.ResultsA total of 44 participants (mean age, 71 years ± 10 [standard deviation]; 31 men) were evaluated. ATAs were seen only in participants with greater than 70% stenosis (16 of 28 patients; P < .001) and were associated with absence of anterior communicating artery (13 of 16 patients; P = .003). There was no association between history of symptoms and degree of stenosis (27 patients with ≥70% stenosis and 17 patients with <70%; P = .54), IPH (12 patients with IPH and 32 patients without IPH; P = .31), and plaque surface structure (17 patients with irregular or ulcerated plaque and 27 with smooth plaque; P = .54). Participants with ATAs (n = 16) were more likely to be symptomatic than were those without ATAs (n = 28) (P = .004). Symptomatic status also was associated with the severity of ATAs (P = .002).ConclusionArterial transit artifacts were the only factor associated with recent ischemic symptoms in participants with carotid stenosis. The degree of stenosis, plaque ulceration, and intraplaque hemorrhage were not associated with symptomatic status.© RSNA, 2020Online supplemental material is available for this article.See also the editorial by Zaharchuk in this issue.

Central Nervous System involvement in tuberculosis: An MRI study considering differences between patients with and without Human Immunodeficiency Virus 1 infection

BACKGROUND

Magnetic resonance imaging (MRI) is largely used in the diagnosis of central nervous system involvement of tuberculosis (CNSTB), yet there is no MRI comparison study between HIV+ and HIV- patients with CNSTB. The aim of the present study was to identify MRI differences in CNSTB between HIV+ and HIV- patients and possibly find early characteristics that could raise the suspect of this disease.

METHODS

We included all patients admitted in our institution between 2011 and 2018 with confirmed diagnosis of CNSTB, and MRI performed in the first week. Patients with preexisting brain pathology or immunodeficiency not HIV related were excluded. We compared CNSTB MRI features between the two groups.

RESULTS

Sixty-nine patients were included (19 HIV+; 50 HIV-). Findings in HIV+ group: 6 lung TB, 5 hydrocephalus, 4 meningeal enhancement, 6 stroke, 2 hemorrhages, and 10 tuberculomas. HIV- group: 22 lung tuberculosis, 15 hydrocephalus, 21 meningeal enhancement, 5 stroke, 4 hemorrhages, 20 tuberculomas. The only statistically significant difference between the two groups was in the stroke occurrence, more frequent in the HIV+ group (P=.028), all involving the basal ganglia.

CONCLUSIONS

Stroke involving the basal ganglia best differentiates CNSTB patients who are HIV+ from those HIV-. This finding was not correlated with meningeal enhancement suggesting that small arteries involvement might precede it. Therefore, we think that HIV+ patients with a new onset of stroke should be evaluated for CNSTB. Follow-up MRI should also be planned since meningeal enhancement might appear in later stages of the disease.