Identification of early invisible acute ischemic stroke in non-contrast computed tomography using two-stage deep-learning model.
Although non-contrast computed tomography (NCCT) is the recommended examination for the suspected acute ischemic stroke (AIS), it cannot detect significant changes in the early infarction. We aimed to develop a deep-learning model to identify early invisible AIS in NCCT and evaluate its diagnostic performance and capacity for assisting radiologists in decision making. : In this multi-center, multi-manufacturer retrospective study, 1136 patients with suspected AIS but invisible lesions in NCCT were collected from two geographically distant institutions between May 2012 to May 2021. The AIS lesions were confirmed based on the follow-up diffusion-weighted imaging and clinical diagnosis. The deep-learning model was comprised of two deep convolutional neural networks to locate and classify. The performance of the model and radiologists was evaluated by the area under the receiver operator characteristic curve (AUC), sensitivity, specificity, and accuracy values with 95% confidence intervals. Delong's test was used to compare the AUC values, and a chi-squared test was used to evaluate the rate differences. 986 patients (728 AIS, median age, 55 years, interquartile range [IQR]: 47-65 years; 664 males) were assigned to the training and internal validation cohorts. 150 patients (74 AIS, median age, 63 years, IQR: 53-75 years; 100 males) were included as an external validation cohort. The AUCs of the model were 83.61% (sensitivity, 68.99%; specificity, 98.22%; and accuracy, 89.87%) and 76.32% (sensitivity, 62.99%; specificity, 89.65%; and accuracy, 88.61%) for the internal and external validation cohorts based on the slices. The AUC of the model was much higher than that of two experienced radiologists (65.52% and 59.48% in the internal validation cohort; 64.01% and 64.39% in external validation cohort; all < 0.001). The accuracy of two radiologists increased from 62.00% and 58.67% to 92.00% and 84.67% when assisted by the model for patients in the external validation cohort. : This deep-learning model represents a breakthrough in solving the challenge that early invisible AIS lesions cannot be detected by NCCT. The model we developed in this study can screen early AIS and save more time. The radiologists assisted with the model can provide more effective guidance in making patients' treatment plan in clinic.
Interhemispheric functional connectivity asymmetry is distinctly affected in left and right mesial temporal lobe epilepsy.
Brain and behavior
INTRODUCTION:The differences of functional connectivity (FC) and functional asymmetry between left and right mesial temporal lobe epilepsy with hippocampal sclerosis (LMTLE and RMTLE) have not been completely clarified yet. The purpose of the present study is to investigate the FC changes and the FC asymmetric patterns of MTLE, and to compare the differences in FC and functional asymmetry between LMTLE and RMTLE. METHODS:In total, 12 LMTLE, 11 RMTLE patients, and 23 healthy controls (HC) were included. Region of interest (ROI)-based analysis was used to evaluate FC. The right functional connectivity (rFC) and left functional connectivity (lFC) of each ROI were calculated. Asymmetry index (AI) was calculated based on the following formula: . Paired t-test and univariate analysis of variance were used to analyze FC asymmetry. Linear correlation analysis was performed between significant FC changes and lateralized ROIs and epilepsy onset age and duration. RESULTS:LMTLE and RMTLE patients showed different patterns of alteration in FC and functional asymmetry when compared with controls. RMTLE presented more extensive FC abnormalities than LMTLE. Regions in ipsilateral temporal lobe presented as central regions of abnormalities in both patient groups. In addition, the asymmetric characteristics of FC were reduced in MTLE compared with HC, with even more pronounced reduction for RMTLE group. Meanwhile, ROIs presented FC AI differences among the three groups were mostly involving left temporal lobe (L_hippo, L_amyg, L_TP, L_aMTG, and L_pTFusC). No correlation was found between significant FC changes and lateralized ROIs and epilepsy onset age and duration. CONCLUSION:The FC and asymmetric features of MTLE are altered and involve both the temporal lobe and extra-temporal lobe. Furthermore, the altered FC and asymmetric features were distinctly affected in LMTLE and RMTLE compared to controls.
Survival prediction of high-grade glioma patients with diffusion kurtosis imaging.
Zhang Ju,Jiang Jingjing,Zhao Lingyun,Zhang Jiaxuan,Shen Nanxi,Li Shihui,Guo Linying,Su Changliang,Jiang Rifeng,Zhu Wenzhen
American journal of translational research
PURPOSE:To evaluate the prognostic value of diffusion kurtosis imaging (DKI) for survival prediction of patients with high-grade glioma (HGG). MATERIALS AND METHODS:DKI was performed for fifty-eight patients with pathologically proven HGG by using a 3-T scanner. The mean kurtosis (MK), mean diffusivity (MD) and fractional anisotropy (FA) values in the solid part of the tumor were measured and normalized. Univariate Cox regression analysis was used to evaluate the association between overall survival (OS) and sex, age, Karnofsky performance status (KPS), tumor grade, Ki-67 labeling index (LI), extent of resection, use of chemoradiotherapy, MK, MD, and FA. Multivariate Cox regression analysis including sex, age, KPS, extent of resection, use of chemoradiotherapy, MK, MD, and FA was subsequently performed. Spearman's correlation coefficient for OS and the area under receiver operating characteristic curve (AUC) for predicting 2-year survival were calculated for each DKI parameter and further compared. RESULTS:In univariate Cox regression analyses, OS was significantly associated with the tumor grade, Ki-67 LI, extent of resection, use of chemoradiotherapy, MK, and MD (P < 0.05 for all). Multivariate Cox regression analyses indicated that MK, MD (hazard ratio = 1.582 and 0.828, respectively, for each 0.1 increase in the normalized value), extent of resection and use of chemoradiotherapy were independent predictors of OS. The absolute value of the correlation coefficient for OS and AUC for predicting 2-year survival by MK (rho = -0.565, AUC = 0.841) were higher than those by MD (rho = 0.492, AUC = 0.772), but the difference was not significant. CONCLUSION:DKI is a promising tool to predict the survival of HGG patients. MK and MD are independent predictors. MK is potentially better associated with OS than MD, which may lead to a more accurate evaluation of HGG patient survival.
Correlation of autopsy pathological findings and imaging features from 9 fatal cases of COVID-19 pneumonia.
ABSTRACT:We aimed to investigate the relationship of radiological features and the corresponding pulmonary pathology of patients with Coronavirus Disease (COVID-19) pneumonia.In this multicenter study, serial chest CT and radiographic images from 9 patients (51-85 years old, 56% male) were reviewed and analyzed. Postmortem lungs were sampled and studied from these autopsies, with a special focus on several corresponding sites based on imaging features.The predominant pattern of pulmonary injury in these 9 cases was diffuse alveolar damage (DAD) and interstitial inflammation. Moreover, acute fibrinous exudates, organization, inflammatory cell infiltration, hyaline membranes, pulmonary edema, pneumocyte hyperplasia, and fibrosis were all observed. The histopathology features varied according to the site and severity of each lesion. In most of the 9 cases, opacities started from a subpleural area and peripheral structures were more severely damaged based on gross views and pathological examinations. Fibrosis could occur in early stages of infection and this was supported by radiological and pathological findings. The radiological features of COVID-19 pneumonia, at the critically ill stage, were diffuse ground-glass opacities with consolidation, interstitial thickening, and fibrous stripes, which was based in the fibrous tissue proliferation in the alveolar and interlobular septa, and filled alveoli with organizing exudation. Fungal and bacterial co-infections were also observed in 6 cases.Typical imaging features can be correlated with underlying pathological findings. Combining assessments of imaging features with pathological findings therefore can enhance our understanding of the histopathological mechanism of COVID-19 pneumonia, and facilitate early radiological diagnosis and prognosis estimation of COVID-19 pneumonia, which has important implications for the development of clinical targeted treatments and research related to COVID-19 pneumonia.
Proton exchange rate of chemical exchange saturation transfer MRI constructed from direct saturation-removed omega plots to improve the assessment of patients with ischemic stroke.
Quantitative imaging in medicine and surgery
Background:Proton exchange rate ( ) magnetic resonance imaging (MRI) has recently been developed, with preliminary results demonstrating its potential for evaluating reactive oxygen species. This prospective cohort study investigated the in different stroke stages and its correlation with stroke severity and prognosis. Methods:In all, 96 ischemic stroke patients were included in the study. Patients were divided into 3 groups based on stroke phase (acute, subacute, and chronic). A spin echo-echo planar imaging sequence with presaturation powers of 1.5, 2.5, and 3.5 µT was implemented to obtain Z-spectra, and maps were constructed from direct saturation-removed omega plots. Relative (r ) and the relative apparent diffusion coefficient (rADC) were calculated as the ratio of or ADC in the infarcts to values in contralateral tissue, respectively. Correlations between both and r and National Institute of Health Stroke Scale (NIHSS) scores were evaluated. Receiver operating characteristic (ROC) analysis was used to evaluate the performance of , r , rADC, and lesion volume for predicting acute stroke outcome. Results:The was significantly higher in ischemic lesions than in contralateral tissue at all stages. In addition, the of acute lesions was higher than that of subacute and chronic lesions [mean (± SD) 935.1±81.5 881.4±55.7 and 866.9±76.7 s, respectively; P<0.05 and P<0.01, respectively]. The difference in between subacute and chronic lesions was not significant. In acute stroke, there was a limited correlation between a lesion's and NIHSS score (R=0.16; P=0.01) and between r and NIHSS score (R=0.28; P=0.001). Acute stroke patients with poor prognosis had significantly higher lesion and r than did those with good prognosis ( : 991.1±78.2 893.1±55.1 s, P<0.001; r : 1.28±0.09 1.15±0.06, P<0.001). In ROC analyses, and r showed favorable predictive performance for acute stroke outcome, with areas under the curve (AUC) of 0.837 and 0.880, respectively, which were slightly but not significantly higher than the AUCs for lesion volume (0.730) and rADC (0.673). Conclusions:This study indicates that MRI is promising for the diagnosis and management of ischemic stroke because it can reflect the oxidative stress of lesions and predict prognosis.
Clinical and High-Resolution CT Features of the COVID-19 Infection: Comparison of the Initial and Follow-up Changes.
OBJECTIVES:In late December 2019, an outbreak of coronavirus disease (COVID-19) in Wuhan, China was caused by a novel coronavirus, newly named severe acute respiratory syndrome coronavirus 2. We aimed to quantify the severity of COVID-19 infection on high-resolution chest computed tomography (CT) and to determine its relationship with clinical parameters. MATERIALS AND METHODS:From January 11, 2020, to February 5, 2020, the clinical, laboratory, and high-resolution CT features of 42 patients (26-75 years, 25 males) with COVID-19 were analyzed. The initial and follow-up CT, obtained a mean of 4.5 days and 11.6 days from the illness onset were retrospectively assessed for the severity and progression of pneumonia. Correlations among clinical parameters, initial CT features, and progression of opacifications were evaluated with Spearman correlation and linear regression analysis. RESULTS:Thirty-five patients (83%) exhibited a progressive process according to CT features during the early stage from onset. Follow-up CT findings showed progressive opacifications, consolidation, interstitial thickening, fibrous strips, and air bronchograms, compared with initial CT (all P < 0.05). Before regular treatments, there was a moderate correlation between the days from onset and sum score of opacifications (R = 0.68, P < 0.01). The C-reactive protein, erythrocyte sedimentation rate, and lactate dehydrogenase showed significantly positive correlation with the severity of pneumonia assessed on initial CT (Rrange, 0.36-0.75; P < 0.05). The highest temperature and the severity of opacifications assessed on initial CT were significantly related to the progression of opacifications on follow-up CT (P = 0.001-0.04). CONCLUSIONS:Patients with the COVID-19 infection usually presented with typical ground glass opacities and other CT features, which showed significant correlations with some clinical and laboratory measurements. Follow-up CT images often demonstrated progressions during the early stage from illness onset.
Feasibility of a Clinical-Radiomics Model to Predict the Outcomes of Acute Ischemic Stroke.
Korean journal of radiology
OBJECTIVE:To develop a model incorporating radiomic features and clinical factors to accurately predict acute ischemic stroke (AIS) outcomes. MATERIALS AND METHODS:Data from 522 AIS patients (382 male [73.2%]; mean age ± standard deviation, 58.9 ± 11.5 years) were randomly divided into the training (n = 311) and validation cohorts (n = 211). According to the modified Rankin Scale (mRS) at 6 months after hospital discharge, prognosis was dichotomized into good (mRS ≤ 2) and poor (mRS > 2); 1310 radiomics features were extracted from diffusion-weighted imaging and apparent diffusion coefficient maps. The minimum redundancy maximum relevance algorithm and the least absolute shrinkage and selection operator logistic regression method were implemented to select the features and establish a radiomics model. Univariable and multivariable logistic regression analyses were performed to identify the clinical factors and construct a clinical model. Ultimately, a multivariable logistic regression analysis incorporating independent clinical factors and radiomics score was implemented to establish the final combined prediction model using a backward step-down selection procedure, and a clinical-radiomics nomogram was developed. The models were evaluated using calibration, receiver operating characteristic (ROC), and decision curve analyses. RESULTS:Age, sex, stroke history, diabetes, baseline mRS, baseline National Institutes of Health Stroke Scale score, and radiomics score were independent predictors of AIS outcomes. The area under the ROC curve of the clinical-radiomics model was 0.868 (95% confidence interval, 0.825-0.910) in the training cohort and 0.890 (0.844-0.936) in the validation cohort, which was significantly larger than that of the clinical or radiomics models. The clinical radiomics nomogram was well calibrated ( > 0.05). The decision curve analysis indicated its clinical usefulness. CONCLUSION:The clinical-radiomics model outperformed individual clinical or radiomics models and achieved satisfactory performance in predicting AIS outcomes.