Objective To evaluate the performance of a convolutional neural network (CNN) model that can automatically detect and classify rib fractures, and output structured reports from computed tomography ...(CT) images. Materials and Methods This study included 1079 patients (median age, 55 years; men, 718) from three hospitals, between January 2011 and January 2019, who were divided into a monocentric training set (n = 876; median age, 55 years; men, 582), five multicenter/multiparameter validation sets (n = 173; median age, 59 years; men, 118) with different slice thicknesses and image pixels, and a normal control set (n = 30; median age, 53 years; men, 18). Three classifications (fresh, healing, and old fracture) combined with fracture location (corresponding CT layers) were detected automatically and delivered in a structured report. Precision, recall, and F1-score were selected as metrics to measure the optimum CNN model. Detection/diagnosis time, precision, and sensitivity were employed to compare the diagnostic efficiency of the structured report and that of experienced radiologists. Results A total of 25054 annotations (fresh fracture, 10089; healing fracture, 10922; old fracture, 4043) were labelled for training (18584) and validation (6470). The detection efficiency was higher for fresh fractures and healing fractures than for old fractures (F1-scores, 0.849, 0.856, 0.770, respectively, p = 0.023 for each), and the robustness of the model was good in the five multicenter/multiparameter validation sets (all mean F1-scores > 0.8 except validation set 5 512 × 512 pixels; F1-score = 0.757). The precision of the five radiologists improved from 80.3% to 91.1%, and the sensitivity increased from 62.4% to 86.3% with artificial intelligence-assisted diagnosis. On average, the diagnosis time of the radiologists was reduced by 73.9 seconds. Conclusion Our CNN model for automatic rib fracture detection could assist radiologists in improving diagnostic efficiency, reducing diagnosis time and radiologists' workload.
In abdomen computed tomography (CT), repeated radiation exposures are often inevitable for cancer patients who receive surgery or radiotherapy guided by CT images. Low-dose scans should thus be ...considered in order to avoid the harm of accumulative x-ray radiation. This work is aimed at improving abdomen tumor CT images from low-dose scans by using a fast dictionary learning (DL) based processing. Stemming from sparse representation theory, the proposed patch-based DL approach allows effective suppression of both mottled noise and streak artifacts. The experiments carried out on clinical data show that the proposed method brings encouraging improvements in abdomen low-dose CT images with tumors.
Background
Granger causality analysis (GCA) has been used to investigate the pathophysiology of migraine. Amygdala plays a key role in pain modulation of migraine attack. However, the detailed ...neuromechanism remained to be elucidated. We applied GCA to explore the amygdala-based directional effective connectivity in migraine without aura (MwoA) and to determine the relation with clinical characteristics.
Methods
Forty-five MwoA patients and forty age-, sex-, and years of education-matched healthy controls(HCs) underwent resting-state functional magnetic resonance imaging (fMRI). Bilateral amygdala were used as seed regions in GCA to investigate directional effective connectivity and relation with migraine duration or attack frequency.
Results
MwoA patients showed significantly decreased effective connectivity from right amygdala to right superior temporal gyrus, left superior temporal gyrus and right precentral gyrus compared with HCs. Furthermore, MwoA patients demonstrated significantly decreased effective connectivity from the left amygdala to the ipsilateral superior temporal gyrus. Also, MwoA patients showed enhanced effective connectivity from left inferior frontal gyrus to left amygdala. Effective connectivity outflow from right amygdala to right precentral gyrus was negatively correlated to disease duration.
Conclusions
Altered directional effective connectivity of amygdala demonstrated that neurolimbic pain networks contribute to multisensory integration abnormalities and deficits in pain modulation of MwoA patients.
Objectives
To evaluate the relationship between hemodynamics and vessel wall remodeling patterns in middle cerebral artery (MCA) stenosis based on high-resolution magnetic resonance imaging and ...computational fluid dynamics (CFD).
Methods
Forty consecutive patients with recent ischemic stroke or transient ischemic attack attributed to unilateral atherosclerotic MCA stenosis (50–99%) were prospectively recruited. All patients underwent a cross-sectional scan of the stenotic MCA vessel wall. The parameters of the vessel wall, the number of patients with acute infarction, translesional wall shear stress ratio (WSSR), wall shear stress in stenosis (WSSs), and translesional pressure ratio were obtained. The patients were divided into positive remodeling (PR) and negative remodeling (NR) groups. The differences in vessel wall parameters and hemodynamics were compared. Correlations between the parameters of the vessel wall and hemodynamics were calculated.
Results
Of the 40 patients, 16 had PR, 19 had NR, and the other 5 displayed non-remodeling. The PR group had a smaller lumen area (
p
= 0.004), larger plaque area (
p
< 0.001), normal wall index (
p
= 0.004), and higher WSSR (
p
= 0.004) and WSSs (
p
= 0.023) at the most narrowed site. The PR group had more enhanced plaques (12 vs 6,
p
= 0.03). The number of patients with acute stroke in the PR group was more than that in the NR group (11 vs 4,
p
= 0.01). The remodeling index (
r
= 0.376,
p
= 0.026) and plaque area (
r
= 0.407,
p
= 0.015) showed a positive correlation with WSSR, respectively.
Conclusions
Hemodynamics plays a role in atherosclerotic plaques and vessel wall remodeling. Individuals with greater hemodynamic values might be more prone to stroke.
Key Points
• Stenotic plaques in middle cerebral artery with positive remodeling have smaller lumen area and larger resp. higher plaque area, normal wall index, translesional wall shear stress ratio, and wall shear stress than negative remodeling.
• The remodeling index and plaque area are positively correlated with translesional wall shear stress ratio.
• Hemodynamic may help to understand the role of positive remodeling in the development of acute stroke.
Full text
Available for:
EMUNI, FIS, FZAB, GEOZS, GIS, IJS, IMTLJ, KILJ, KISLJ, MFDPS, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, SBMB, SBNM, UKNU, UL, UM, UPUK, VKSCE, VSZLJ, ZAGLJ
Background
Post-traumatic headache (PTH) is one of the most frequent symptoms following mild traumatic brain injury (mTBI). Neuroimaging studies implicate hypothalamic function connectivity (FC) ...disruption as an important factor in pain disorders. However, it is unknown whether there are alterations in the hypothalamus-based resting state FC within PTH following mTBI at the acute stage and its relationship with headache symptom measurement.
Methods
Forty-four mTBI patients with PTH, 27 mTBI patients without PTH and 43 healthy controls who were well matched for age, gender, and years of education were enrolled in this study. All participants underwent resting-state functional magnetic resonance imaging (fMRI) scanning as well as headache symptom measurement and cognitive assessment. Hypothalamic resting state networks were characterized by using a standard seed-based whole-brain correlation method. The bilateral hypothalamic FC was compared among the three groups. Furthermore, the correlations between hypothalamic resting state networks and headache frequency, headache intensity and MoCA scores was investigated in mTBI patients with PTH using Pearson rank correlation.
Results
Compared with mTBI patients without PTH, mTBI patients with PTH at the acute stage presented significantly decreased left hypothalamus-based FC with the right middle frontal gyrus (MFG) and right medial superior frontal gyrus (mSFG), and significantly decreased right hypothalamus-based FC with the right MFG. Decreased FC of the right MFG was significantly positively associated with headache frequency and headache intensity (
r
= 0.339,
p
= 0.024;
r
= 0.408,
p
= 0.006, respectively). Decreased FC of the right mSFG was significantly positively associated with headache frequency and headache intensity (
r
= 0.740,
p
< 0.0001;
r
= 0.655,
p
< 0.0001, respectively).
Conclusion
Our data provided evidence of disrupted hypothalamic FC in patients with acute mTBI with PTH, while abnormal FC significantly correlated with headache symptom measurement. Taken together, these changes may play an essential role in the neuropathological mechanism of mTBI patients with PTH.
Background
Migraine constitutes a global health burden, and its pathophysiology is not well-understood; research evaluating cerebral perfusion and altered blood flow between brain areas using ...non-invasive imaging techniques, such as arterial spin labeling, have been scarce. This study aimed to assess cerebral blood flow (CBF) and its connectivity of migraine.
Methods
This study enrolled 40 patients with episodic migraine without aura (MwoA), as well as 42 healthy patients as control (HC). Two groups of normalized CBF and CBF connectivity were compared, and the relationship between CBF variation and clinical scale assessment was further evaluated.
Results
In comparison to HC subjects, MwoA patients exhibited higher CBF in the right middle frontal orbital gyrus (ORBmid.R) and the right middle frontal gyrus, while that in Vermis_6 declined. The increased CBF of ORBmid.R was positively correlated with both the Visual Light Sensitivity Questionnaire-8 (VLSQ-8) and the monthly attack frequency score. In MwoA, significantly decreased CBF connectivity was detected between ORBmid.R and the left superior frontal gyrus, the right putamen, the right caudate, as well as the right angular gyrus. In addition, increased CBF connectivity was observed between the left calcarine cortex and ORBmid.R.
Conclusions
Our results indicate that migraine patients exhibit abnormalities in regional CBF and feature CBF connection defects at the resting state. The affected areas involve information perception, information integration, and emotional, pain and visual processing. Our findings might provide important clues for the pathophysiology of migraine.
Objectives
To develop and externally validate a machine learning (ML) model based on diffusion-weighted imaging (DWI) and fluid-attenuated inversion recovery (FLAIR) to identify the onset time of ...wake-up stroke from MRI.
Methods
DWI and FLAIR images of stroke patients within 24 h of clear symptom onset in our hospital (dataset 1,
n
= 410) and another hospital (dataset 2,
n
= 177) were included. Seven ML models based on dataset 1 were developed to estimate the stroke onset time for binary classification (≤ 4.5 h or > 4.5 h): Random Forest (RF), support vector machine with kernel (svmLinear) or radial basis function kernel (svmRadial), Bayesian (Bayes), K-nearest neighbor (KNN), adaptive boosting (AdaBoost), and neural network (NNET). ROC analysis and RSD were performed to evaluate the performance and stability of the ML models, respectively, and dataset 2 was externally validated to evaluate the model generalization ability using ROC analysis.
Results
svmRadial achieved the best performance with the highest AUC and accuracy (AUC: 0.896, accuracy: 0.878), and was the most stable (RSD% of AUC: 0.08, RSD% of accuracy: 0.06). The svmRadial model was then selected as the final model, and the AUC of the svmRadial model for predicting the onset time external validation was 0.895, with 0.825 accuracy.
Conclusions
The svmRadial model using DWI + FLAIR is the most stable and generalizable for identifying the onset time of wake-up stroke patients within 4.5 h of symptom onset.
Key Points
• Machining learning model helps clinicians to identify wake-up stroke patients within 4.5 h of symptom onset.
• A prospective study showed that svmRadial model based on DWI + FLAIR was the most stable in predicting the stroke onset time.
• External validation showed that svmRadial model has good generalization ability in predicting the stroke onset time.
Full text
Available for:
EMUNI, FIS, FZAB, GEOZS, GIS, IJS, IMTLJ, KILJ, KISLJ, MFDPS, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, SBMB, SBNM, UKNU, UL, UM, UPUK, VKSCE, VSZLJ, ZAGLJ
Background
Resting-state functional magnetic resonance imaging (fMRI) has confirmed disrupted visual network connectivity in migraine without aura (MwoA). The thalamus plays a pivotal role in a ...number of pain conditions, including migraine. However, the significance of altered thalamo-visual functional connectivity (FC) in migraine remains unknown. The goal of this study was to explore thalamo-visual FC integrity in patients with MwoA and investigate its clinical significance.
Methods
Resting-state fMRI data were acquired from 33 patients with MwoA and 22 well-matched healthy controls. After identifying the visual network by independent component analysis, we compared neural activation in the visual network and thalamo-visual FC and assessed whether these changes were linked to clinical characteristics. We used voxel-based morphometry to determine whether functional differences were dependent on structural differences.
Results
The visual network exhibited significant differences in regions (bilateral cunei, right lingual gyrus and left calcarine sulcus) by inter-group comparison. The patients with MwoA showed significantly increased FC between the left thalami and bilateral cunei and between the right thalamus and the contralateral calcarine sulcus and right cuneus. Furthermore, the neural activation of the left calcarine sulcus was positively correlated with visual analogue scale scores (
r
= 0.319,
p
= 0.043), and enhanced FC between the left thalamus and right cuneus in migraine patients was negatively correlated with Generalized Anxiety Disorder scores (
r
= − 0.617,
p
= 0.005).
Conclusion
Our data suggest that migraine distress is exacerbated by aberrant feedback projections to the visual network, playing a crucial role in migraine physiological mechanisms. The current study provides further insights into the complex scenario of migraine mechanisms.
Objective
To develop a convolutional neural network (CNN) model for the automatic detection and classification of rib fractures in actual clinical practice based on cross-modal data (clinical ...information and CT images).
Materials
In this retrospective study, CT images and clinical information (age, sex and medical history) from 1020 participants were collected and divided into a single-centre training set (
n
= 760; age: 55.8 ± 13.4 years; men: 500), a single-centre testing set (
n
= 134; age: 53.1 ± 14.3 years; men: 90), and two independent multicentre testing sets from two different hospitals (
n
= 62, age: 57.97 ± 11.88, men: 41;
n
= 64, age: 57.40 ± 13.36, men: 35). A Faster Region–based CNN (Faster R-CNN) model was applied to integrate CT images and clinical information. Then, a result merging technique was used to convert 2D inferences into 3D lesion results. The diagnostic performance was assessed on the basis of the receiver operating characteristic (ROC) curve, free-response ROC (fROC) curve, precision, recall (sensitivity), F1-score, and diagnosis time. The classification performance was evaluated in terms of the area under the ROC curve (AUC), sensitivity, and specificity.
Results
The CNN model showed improved performance on fresh, healing, and old fractures and yielded good classification performance for all three categories when both clinical information and CT images were used compared to the use of CT images alone. Compared with experienced radiologists, the CNN model achieved higher sensitivity (mean sensitivity: 0.95 > 0.77, 0.89 > 0.61 and 0.80 > 0.55), comparable precision (mean precision: 0.91 > 0.87, 0.84 > 0.77, and 0.95 > 0.70), and a shorter diagnosis time (average reduction of 126.15 s).
Conclusions
A CNN model combining CT images and clinical information can automatically detect and classify rib fractures with good performance and feasibility in actual clinical practice.
Key Points
•
The developed convolutional neural network (CNN) performed better in fresh, healing, and old fractures and yielded a good classification performance in three categories, if both (clinical information and CT images) were used compared to CT images alone.
•
The CNN model had a higher sensitivity and matched precision in three categories than experienced radiologists with a shorter diagnosis time in actual clinical practice.
Full text
Available for:
EMUNI, FIS, FZAB, GEOZS, GIS, IJS, IMTLJ, KILJ, KISLJ, MFDPS, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, SBMB, SBNM, UKNU, UL, UM, UPUK, VKSCE, VSZLJ, ZAGLJ
Background
Resting-state functional magnetic resonance imaging (Rs-fMRI) has confirmed sensorimotor network (SMN) dysfunction in migraine without aura (MwoA). However, the underlying mechanisms of ...SMN effective functional connectivity in MwoA remain unclear. We aimed to explore the association between clinical characteristics and effective functional connectivity in SMN, in interictal patients who have MwoA.
Methods
We used Rs-fMRI to acquire imaging data in 40 episodic patients with MwoA in the interictal phase and 34 healthy controls (HCs). Independent component analysis was used to profile the distribution of SMN and calculate the different SMN activity between the two groups. Subsequently, Granger causality analysis was used to analyze the effective functional connectivity between the SMN and other brain regions.
Results
Compared to the HCs, MwoA patients showed higher activity in the bilateral postcentral gyri (PoCG), but lower activity in the left midcingulate cortex (MCC). Moreover, MwoA patients showed decreased effective functional connectivity from the SMN to left middle temporal gyrus, right putamen, left insula and bilateral precuneus, but increased effective functional connectivity to the right paracentral lobule. There was also significant effective functional connectivity from the primary visual cortex, right cuneus and right putamen to the SMN. In the interictal period, there was positive correlation between the activity of the right PoCG and the frequency of headache. The disease duration was positively correlated with abnormal effective functional connectivity from the left PoCG to right precuneus. In addition, the headache impact scores were negatively correlated with abnormal effective functional connectivity from the left MCC to right paracentral lobule, as well as from the right precuneus to left PoCG.
Conclusions
These differential, resting-state functional activities of the SMN in episodic MwoA may contribute to the understanding of migraine-related intra- and internetwork imbalances associated with nociceptive regulation and chronification.