Diffusion-/perfusion-weighted imaging fusion to automatically identify stroke within 4.5 h.

OBJECTIVES: We aimed to develop machine learning (ML) models based on diffusion- and perfusion-weighted imaging fusion (DP fusion) for identifying stroke within 4.5 h, to compare them with DWI- and/or PWI-based ML models, and to construct an automatic segmentation-classification model and compare with manual labeling methods. METHODS: ML models were developed from multimodal MRI datasets of acute stroke patients within 24 h of clear symptom onset from two centers. The processes included manual segmentation, registration, DP fusion, feature extraction, and model establishment (logistic regression (LR) and support vector machine (SVM)). A segmentation-classification model (X-Net) was proposed for automatically identifying stroke within 4.5 h. The area under the receiver operating characteristic curve (AUC), sensitivity, Dice coefficients, decision curve analysis, and calibration curves were used to evaluate model performance. RESULTS: A total of 418 patients (≤ 4.5 h: 214; > 4.5 h: 204) were evaluated. The DP fusion model achieved the highest AUC in identifying the onset time in the training (LR: 0.95; SVM: 0.92) and test sets (LR: 0.91; SVM: 0.90). The DP fusion-LR model displayed consistent positive and greater net benefits than other models across a broad range of risk thresholds. The calibration curve demonstrated the good calibration of the DP fusion-LR model (average absolute error: 0.049). The X-Net model obtained the highest Dice coefficients (DWI: 0.81; Tmax: 0.83) and achieved similar performance to manual labeling (AUC: 0.84). CONCLUSIONS: The automatic segmentation-classification models based on DWI and PWI fusion images had high performance in identifying stroke within 4.5 h. CLINICAL RELEVANCE STATEMENT: Perfusion-weighted imaging (PWI) fusion images had high performance in identifying stroke within 4.5 h. The automatic segmentation-classification models based on DWI and PWI fusion images could provide clinicians with decision-making guidance for acute stroke patients with unknown onset time. KEY POINTS: • The diffusion/perfusion-weighted imaging fusion model had the best performance in identifying stroke within 4.5 h. • The X-Net model had the highest Dice and achieved performance close to manual labeling in segmenting lesions of acute stroke. • The automatic segmentation-classification model based on DP fusion images performed well in identifying stroke within 4.5 h.

Detection and Semiquantitative Analysis of Cardiomegaly, Pneumothorax, and Pleural Effusion on Chest Radiographs.

PURPOSE: To develop and evaluate deep learning models for the detection and semiquantitative analysis of cardiomegaly, pneumothorax, and pleural effusion on chest radiographs. MATERIALS AND METHODS: In this retrospective study, models were trained for lesion detection or for lung segmentation. The first dataset for lesion detection consisted of 2838 chest radiographs from 2638 patients (obtained between November 2018 and January 2020) containing findings positive for cardiomegaly, pneumothorax, and pleural effusion that were used in developing Mask region-based convolutional neural networks plus Point-based Rendering models. Separate detection models were trained for each disease. The second dataset was from two public datasets, which included 704 chest radiographs for training and testing a U-Net for lung segmentation. Based on accurate detection and segmentation, semiquantitative indexes were calculated for cardiomegaly (cardiothoracic ratio), pneumothorax (lung compression degree), and pleural effusion (grade of pleural effusion). Detection performance was evaluated by average precision (AP) and free-response receiver operating characteristic (FROC) curve score with the intersection over union greater than 75% (AP75; FROC score75). Segmentation performance was evaluated by Dice similarity coefficient. RESULTS: The detection models achieved high accuracy for detecting cardiomegaly (AP75, 98.0%; FROC score75, 0.985), pneumothorax (AP75, 71.2%; FROC score75, 0.728), and pleural effusion (AP75, 78.2%; FROC score75, 0.802), and they also weakened boundary aliasing. The segmentation effect of the lung field (Dice, 0.960), cardiomegaly (Dice, 0.935), pneumothorax (Dice, 0.827), and pleural effusion (Dice, 0.826) was good, which provided important support for semiquantitative analysis. CONCLUSION: The developed models could detect cardiomegaly, pneumothorax, and pleural effusion, and semiquantitative indexes could be calculated from segmentations.Keywords: Computer-Aided Diagnosis (CAD), Thorax, CardiacSupplemental material is available for this article.© RSNA, 2021.

MRI predictors of intracranial hemorrhage in acute ischemic stroke after endovascular thrombectomy therapy.

OBJECTIVES: To investigate the predictors for the occurrence of intracranial hemorrhage (ICH) after endovascular thrombectomy (EVT) therapy in acute ischemic stroke (AIS) patients. METHODS: Patients with AIS who underwent EVT and bridging therapy were enrolled retrospectively. ICH was evaluated on follow-up noncontrast CT or MRI. Diffusion weighted imaging (DWI) volume, perfusion weighted imaging (PWI) volume, DWI-PWI mismatch (DPM) volume and other clinical data were collected for 135 AIS patients. Multivariate logistic regression analysis was used to predict ICH after therapy in AIS patients. RESULTS: The DWI volume in patients undergoing EVT with ICH was significantly larger than that in patients without ICH (50.61±47.43 vs 26.65±29.51; t=-2.416, P=0.020). For patients treated with bridging therapy, patients with ICH had larger DWI volume (26.32±29.66 vs 13.04±20.14; t=-2.013, P=0.037) and PWI volume (174.21±75.12 vs 129.87±60.29; t=-2.618, P=0.011) than patients without ICH. More patients with ICH were attempted for >3 passes with retriever during EVT than patients without ICH (EVT: 51.72% vs 26.19%; χ2 =5.131, P=0.028; bridging therapy: 48.15% vs 21.62%; χ2 =4.982, P=0.033). Multivariable logistic regression analysis demonstrated that DWI volume (OR, 1.017 (95% CI, 1.002-1.033); P=0.022) and >3 passes with the retriever (OR, 0.327 (95% CI, 0.114-0.936); P=0.037) were independently associated with ICH after EVT in AIS patients. DWI volume (OR, 1.024 (95% CI, 1.011-1.048); P=0.046), PWI volume (OR, 1.010 (95% CI, 1.002-1.018); P=0.016) and >3 passes with the retriever (OR, 0.281 (95% CI, 0.089-0.887); P=0.030) were independently associated with ICH after bridging therapy in AIS patients. CONCLUSIONS: DWI volume, PWI volume and >3 passes with the retriever were able to predict the ICH in patients with AIS after EVT therapy.

Diffusion-weighted imaging (DWI) ischemic volume is related to FLAIR hyperintensity-DWI mismatch and functional outcome after endovascular therapy.

BACKGROUND: We assessed whether diffusion-weighted imaging (DWI) volume was associated with fluid-attenuated inversion recovery vascular hyperintensities (FVH)-DWI mismatch and functional outcome in patients with acute stroke who received endovascular therapy (EVT). METHODS: Fifty-three acute stroke patients who received EVT were enrolled. FVH-DWI mismatch, DWI volume on admission, DWI volume on follow-up, DWI volume growth, the functional outcome at 3 months (mRS) and other clinical data were collected. Receiver operating characteristic (ROC) analysis was performed to evaluate the value of DWI volume in predicting functional outcome after stroke. RESULTS: The FVH-DWI mismatch group had a smaller DWI volume on admission (13.86±19.58 vs. 65.07±52.21; t=-4.301, P=0.000), a smaller DWI volume on follow-up (29.88±33.52 vs. 112.43±87.19; t=-4.143, P=0.000), and a lower DWI volume growth (16.02±19.90 vs. 47.36±40.06; t=-3.326, P=0.003) than those of the no FVH-DWI mismatch group. The good functional outcome group had a smaller DWI volume on admission (13.30±13.26 vs. 68.56±54.28; t=-5.611, P=0.000), a smaller DWI volume on follow-up (27.65±18.80 vs. 120.25±90.37; t=-5.720, P=0.000), lower DWI volume growth (14.35±15.06 vs. 51.69±41.17; t=-4.737, P=0.001) and a higher FVH-DWI mismatch ratio (75.76% vs. 35%; t=8.647; P=0.004) than those of the poor functional outcome group. ROC analysis showed that the sensitivity and specificity of DWI volume on admission for predicting functional outcome were 65% and 96.97%, respectively (the optimal cut-off value: 33.50 mL); DWI volume on follow-up was 48.6 mL, with a sensitivity and specificity of 80% and 87.88%, respectively; DWI volume growth was 22.25 mL, with a sensitivity and specificity of 70% and 87.88%, respectively. CONCLUSIONS: DWI volume and DWI volume growth can provide the prognostic information of acute stroke patients after thrombectomy.

FLAIR hyperintensities-DWI mismatch in acute stroke: associations with DWI volume and functional outcome.

The value of fluid-attenuated inversion recovery vascular hyperintensity (FVH)-diffusion-weighted imaging (DWI)mismatch in predicting outcome with acute stroke is unclear. We assessed the association between FVH-DWI mismatch and functional outcome in patients with acute stroke patients receiving endovascular therapy. FVH-DWI mismatch, DWI volume, the functional outcome at 3 months (modified Rankin scale; mRS) and other clinical data were collected for 59 acute stroke patients receiving MRI within 6 h before endovascular therapy. Statistical analysis was performed to predict functional outcome after stroke. Patients with FVH-DWI mismatch (n = 39) had a higher FVH score (p = 0.038), smaller DWI volume (p = 0.007), higher American Society of Interventional and Therapeutic Neuroradiology (ASITN) (p = 0.043), higher interval time of ASITN (p = 0.038), and better functional outcome (p = 0.010) than patients with no FVH-DWI mismatch (n = 20). The good functional outcome group (n = 32) had a higher FVH score (p = 0.028), smaller DWI volume (p = 0.003) and higher ASITN grade (p = 0.043) than the poor functional outcome group (n = 27). Multivariable logistic analysis demonstrated that DWI volume (OR, 1.031; 95% CI, 1.005-1.058; P = 0.021) and FVH-DWI mismatch (OR, 14.311; 95% CI, 2.670-76.703; P = 0.002) were independently associated with functional outcome. Assessing DWI volume and FVH-DWI mismatch in acute stroke patients might be useful for predicting functional outcome after stroke.

Functional connectivity disruption of the substantia nigra associated with cognitive impairment in acute mild traumatic brain injury.

PURPOSE: Mild traumatic brain injury is known to have frequent cognitive impairment. Accumulating evidence is pointing to the malfunctioning of the substantia nigra (SN) as an important factor for head trauma. However, it remains unknown whether changes in the SN-based resting state functional connectivity following mTBI at acute stage and its relationship with cognitive function. MATERIALS AND METHODS: 58 patients with mTBI and 30 age-, gender-, and years of education-matched healthy controls were enrolled in the current study. All of participants received resting state functional magnetic resonance imaging as well as neuropsychological assessment. The resting state functional MR imaging data were analyzed by using a standard seed-based whole-brain correlation method to characterize SN resting state networks. Student t tests were used to perform comparisons. The association between SN resting state networks and performance on neuropsychological measures was also investigated in patients with mTBI by using Pearson rank correlation. RESULTS: Patients with mTBI at acute stage exhibited reduced left SN-based functional connectivity with right insula and caudate and increased left SN-based functional connectivity with left precuneus and left middle occipital gyrus, and reduced right SN-based functional connectivity with left insula. Increased functional connectivity of left precuneus was negatively associated with neurocognitive functions as well (r = -0.266; P = 0.049). CONCLUSION: The present study indicated that patients with acute mTBI suffer from disruption in their SN resting state networks. Moreover, abnormal functional connectivity significantly correlated with cognitive function. Taking together, these results may better improve our understanding of the neuropathological mechanism underlying the neurocognitive symptoms associated with acute mTBI.

FLAIR vascular hyperintensity in acute stroke is associated with collateralization and functional outcome.

PURPOSE: Fluid-attenuated inversion recovery vascular hyperintensity (FVH) is frequently found in stroke patients after intracranial arterial occlusion, but the prognostic value of FVH findings is unclear. We assessed whether FVH is associated with cerebral collateral status and functional outcome in patients with acute stroke patients receiving endovascular therapy. METHODS: FVH score, American Society of Interventional and Therapeutic Neuroradiology (ASITN) grade, the functional outcome at 3 months (modified Rankin Scale (mRS)), and other clinical data were collected for 37 acute stroke patients with large vessel occlusion (LVO) receiving MRI before and after endovascular therapy. Statistical analysis was performed to predict functional outcome after stroke. RESULTS: The good functional outcome group (n = 16) had a higher FVH1 (FVH before therapy) score (4.63 ± 1.20 vs 3.14 ± 1.15; p = 0.001) and ASITN grade (3.31 ± 0.48 vs 2.00 ± 1.22; p < 0.001) and a lower FVH2 (FVH after therapy) score than the poor functional outcome group (n = 21; 0.125 ± 0.50 vs 1.44 ± 2.16; p = 0.030). mRS at 3 months was negatively correlated with FVH1 (r = - 0.525, p = 0.001) and the ASITN grade (r = - 0.478, p = 0.003) and positively correlated with FVH2 (r = 0.376, p = 0.034). FVH1 (OR, 0.085; 95% CI, 0.013-0.577; p = 0.012) and FVH2 (OR, 2.724; 95% CI, 1.061-6.996; p = 0.037) were independently associated with functional outcome in multivariable logistic regression analysis. CONCLUSIONS: Assessing FVH before and after therapy in acute stroke patients with LVO might be useful for predicting functional outcome after stroke. KEY POINTS: • Fluid-attenuated inversion recovery vascular hyperintensity is a circular or serpentine brightening in the brain parenchyma or cortical surface bordering the subarachnoid space on MR imaging. • A prospective study showed that fluid-attenuated inversion recovery vascular hyperintensity is associated with cerebral collateral circulation and prognosis. • Fluid-attenuated inversion recovery vascular hyperintensity helps clinicians to predict the prognosis of patients with acute stroke.

Geometric factors affecting capillary discharge jet length in atmospheric pressure air.

The capillary discharge triggered by a pulse source can produce a certain length of plasma jet. In this paper, the physical process of the capillary discharge jet is analyzed, and it is pointed out that the capillary plasma-jet length is significantly affected by the expansion pressure caused by the arc discharge in the capillary chamber. The greater the pressure in the capillary chamber is, the longer the jet length. The experimental setup of the capillary discharge is established in atmospheric pressure air. The influence of the surface distance along the capillary wall, the diameter of the capillary cathode, and the length of the cathode tip on the plasma-jet length is studied under a specific trigger pulse. The experimental results show that the greater the deposited energy density in the capillary chamber is, the longer the plasma jet length. As the surface distance increases, the energy deposited in the arc channel increases first and then tends to be saturated. There is an optimum surface distance to maximize the energy density in the capillary chamber, while the plasma jet length is the longest.