Uncovering Brain Differences in Preschoolers and Young Adolescents with Autism Spectrum Disorder Using Deep Learning.

Identifying brain abnormalities in autism spectrum disorder (ASD) is critical for early diagnosis and intervention. To explore brain differences in ASD and typical development (TD) individuals by detecting structural features using T1-weighted magnetic resonance imaging (MRI), we developed a deep learning-based approach, three-dimensional (3D)-ResNet with inception (I-ResNet), to identify participants with ASD and TD and propose a gradient-based backtracking method to pinpoint image areas that I-ResNet uses more heavily for classification. The proposed method was implemented in a preschool dataset with 110 participants and a public autism brain imaging data exchange (ABIDE) dataset with 1099 participants. An extra epilepsy dataset with 200 participants with clear degeneration in the parahippocampal area was applied as a verification and an extension. Among the datasets, we detected nine brain areas that differed significantly between ASD and TD. From the ROC in PASD and ABIDE, the sensitivity was 0.88 and 0.86, specificity was 0.75 and 0.62, and area under the curve was 0.787 and 0.856. In a word, I-ResNet with gradient-based backtracking could identify brain differences between ASD and TD. This study provides an alternative computer-aided technique for helping physicians to diagnose and screen children with an potential risk of ASD with deep learning model.

Multiparametric liver MRI for predicting early recurrence of hepatocellular carcinoma after microwave ablation.

BACKGROUND: High early recurrence (ER) of hepatocellular carcinoma (HCC) after microwave ablation (MWA) represents a sign of aggressive behavior and severely worsens prognosis. The aim of this study was to estimate the outcome of HCC following MWA and develop a response algorithmic strategy based on multiparametric MRI and clinical variables. METHODS: In this retrospective study, we reviewed the records of 339 patients (mean age, 62 ± 12 years; 106 men) treated with percutaneous MWA for HCC between January 2014 and December 2017 that were evaluated by multiparametric MRI. These patients were randomly split into a development and an internal validation group (3:1). Logistic regression analysis was used to screen imaging features. Multivariate Cox regression analysis was then performed to determine predictors of ER (within 2 years) of MWA. The response algorithmic strategy to predict ER was developed and validated using these data sets. ER rates were also evaluated by Kaplan-Meier analysis. RESULTS: Based on logistic regression analyses, we established an image response algorithm integrating ill-defined margins, lack of capsule enhancement, pre-ablative ADC, ΔADC, and EADC to calculate recurrence scores and define the risk of ER. In a multivariate Cox regression model, the independent risk factors of ER (p < 0.05) were minimal ablative margin (MAM) (HR 0.57; 95% CI 0.35 - 0.95; p < 0.001), the recurrence score (HR: 9.25; 95% CI 4.25 - 16.56; p = 0.021), and tumor size (HR 6.21; 95% CI 1.25 - 10.82; p = 0.014). Combining MAM and tumor size, the recurrence score calculated by the response algorithmic strategy provided predictive accuracy of 93.5%, with sensitivity of 92.3% and specificity of 83.1%. Kaplan-Meier estimates of the rates of ER in the low-risk and high-risk groups were 6.8% (95% CI 4.0 - 9.6) and 30.5% (95% CI 23.6 - 37.4), respectively. CONCLUSION: A response algorithmic strategy based on multiparametric MRI and clinical variables was useful for predicting the ER of HCC after MWA.

3D Cube FLAIR plus HyperSense compressed sensing is superior to 2D T2WI FLAIR scanning regarding image quality, spatial resolution, detection rate for cortical microinfarcts.

3-dimention (3D) Cube isotropic volumetric magnetic resonance imaging (MRI) facilitates comprehensive recognition of microinfarcts while it takes long scanning time. HyperSense compressed sensing is an emerging technique for accelerating MRI acquisition to reduce scanning time, while its application along with 3D Cube MRI for microinfarcts is seldom reported. Therefore, this study aimed to investigate the efficiency of 3D Cube FLAIR plus HyperSense compressed sensing technique versus conventional 2-dimention (2D) FLAIR scanning in the detection of cortical microinfarcts (CMIs). Totally 59 patients with cerebrovascular disease were enrolled then scanned by 3D Cube FLAIR plus HyperSense compressed sensing and 2D T2WI FLAIR sequences. The image quality scores, signal-to-noise ratio (SNR) for gray matter (GM), SNR for white matter (WM), their contrast-to-noise ratio (WM-to-GM CNR), detected number of CMIs were evaluated. 3D Cube FLAIR plus HyperSense showed a dramatically increased scores of uniformity, artifact, degree of lesion displacement, and overall image quality compared to 2D T2WI FLAIR. Meanwhile, it exhibited similar SNRwm and SNRgm, but a higher WM-to-GM contrast-to-noise ratio compared with 2D T2WI FLAIR. Furthermore, the scanning time of 3D Cube FLAIR plus HyperSense and 2D T2WI FLAIR were both set as 2.5 minutes. Encouragingly, 244 CMIs were detected by 3D Cube FLAIR plus HyperSense, which was higher compared to 2D T2WI FLAIR (106 detected CMIs). 3D Cube FLAIR plus HyperSense compressed sensing is superior to 2D T2WI FLAIR scanning regarding image quality, spatial resolution, detection rate for CMIs; meanwhile, it does not increase the scanning time. These findings may contribute to early detection and treatment of stroke.

Application study of 3D LAVA-Flex on lumbar intervertebral disc degeneration.

BACKGROUND: Degeneration of the intervertebral discs are very common diseases, indicating the specific or malignant changes in intervertebral disc component, structure and function. Imaging examination is currently used to evaluate the severity of lumbar intervertebral disc degeneration. This study was designed to investigate the diagnostic value of 3D LAVA-Flex in lumbar intervertebral disc degeneration. MATERIAL AND METHODS: Sagittal 3D LAVA-Flex and T2WI scans were performed in 45 patients with lumbar intervertebral disc degeneration. On T2WI, the degenerated intervertebral disc in every patient was evaluated using Pfirrmann grade. Then, the patients were re-evaluated using 3D LAVA-Flex with considerations of the distinction of nucleus pulposus and annulus fibrosus, hypointense signal of intervertebral disc and height of intervertebral disc. The evaluation results were compared between 3D LAVA-Flex and T2WI. Virtual endoscopy was also performed to evaluate the degenerated intervertebral disc. RESULTS: The intermediate-intense signal of nucleus pulposus and complete ring-shaped hyperintense signal of annulus fibrosus were found and the distinction of nucleus pulposus and annulus fibrosus was clear in the normal intervertebral disc on 3D LAVA-Flex. The incidence of linear hypointensity of narrowed intervertebral space (65/91) was higher than that of normal intervertebral space (4/134) (P = 0.000). A good consistency was shown between the LAVA-Flex grade and T2WI-based Pfirrmann grade. Virtual endoscopy based on 3D LAVA-Flex could help clearly show the anatomic relationship between the degenerated disc and intervertebral foramen. CONCLUSIONS: 3D LAVA-Flex and T2WI show similar efficacy in evaluating lumbar intervertebral disc degeneration. 3D LAVA-Flex-based virtual endoscopy possesses great potential in the study of intervertebral disc abnormalities.

The effects of three different contrast agents (Gd-BOPTA, Gd-DTPA, and Gd-DOTA) on brachial plexus magnetic resonance imaging.

BACKGROUND: MRI is very important for guiding the diagnosis and treatment of brachial plexus diseases. The most used type of MRI brachial plexus imaging is the 3D Short Term Inversion Recovery (STIR) sequence with contrast agent. This study aimed to investigate the effect of three contrast agents; gadobenate dimeglumine (Gd-BOPTA), gadopentetate dimeglumine (Gd-DTPA), and Gadoteric Acid Meglumine (Gd-DOTA) on brachial plexus magnetic resonance imaging (MRI). METHODS: We recruited 60 patients with suspected brachial plexus injury randomly into three groups. MRI images were obtained from each patient. Prior to scanning, the first group was injected with GD-BOPTA, the second group with Gd-DTPA, and the third with Gd-DOTA. The amount of contrast agent was 0.1 mmol/kg according to the weight of each patient, the injection rate was 1.5 mL/s, and 20 mL saline was injected at the same rate with a high-pressure injector. Immediately after the injection of contrast agent and saline, a 3D Sampling perfection with application optimized contrasts using different flip angle evolutions (SPACE) STIR sequence was used for scanning. The Signal Intensity (SI) and Standard Deviation (SD) of Maximal intensity projection (MIP) images for regions outside the anatomy (ROI background) with area of 17 mm2 on both sides of the C6 peripheral nerves (ROI nerve), and tissue adjacent to the peripheral nerves (ROI tissue) were obtained. Signal to noise ratio (SNR) and contrast to noise ratio (CNR) were then calculated. RESULTS: The SNR was 40.66±25.27, 34.65±14.86, and 44.63±30.79 for Gd-BOPTA, Gd-DTPA, and Gd-DOTA, respectively and the CNR was 20.24±15.17, 16.07±7.50, and 20.84±15.53 for Gd-BOPTA, Gd-DTPA, and Gd-DOTA, respectively. In addition, there was no statistical difference in the SNR or CNR of brachial plexus nerves using the three contrast agents to enhance the 3D SPACE sequence χ2=1.877, P=0.391>0.05 and χ2=1.717, P=0.424, respectively. CONCLUSIONS: There were no significant differences in the efficacy of three contrast agents in imaging the brachial plexus.

Alterations of regional homogeneity in Parkinson's disease with mild cognitive impairment: a preliminary resting-state fMRI study.

PURPOSE: Mild cognitive impairment (MCI) is commonly observed in Parkinson's disease (PD), even in the early stages. However, the neural substrates of cognitive impairment in PD remain unclear. The aim of the current study was to investigate the change of local brain function in PD patients with MCI. METHODS: Fifty patients with PD, including 25 PD patients with MCI (PD-MCI) and 25 PD patients with normal cognition (PD-NC), and 25 age- and sex-matched healthy controls (HC) were enrolled. Conventional magnetic resonance imaging (MRI), 3D structural images, and resting state-functional MRI (rs-fMRI) were performed in all subjects. Regional homogeneity (ReHo) was measured based on the rs-fMRI images to investigate the altered local brain functions. RESULTS: Brain regions with decreased ReHo were located in the left posterior cerebellar lobe in PD sub-groups compared to the HC group, and the brain regions with increased ReHo were located in the limbic lobe (right precuneus/bilateral middle cingulate cortex) in PD-MCI compared with HC group. PD-MCI presented with increased ReHo in the bilateral precuneus/left superior parietal lobe and decreased ReHo in the left insula compared to PD-NC. ReHo values for the left precuneus were negatively related to neuropsychological scores, and ReHo values for the left insula were positively related to neuropsychological scores in PD subjects. CONCLUSION: The present study demonstrated abnormal spontaneous synchrony in the left insula and left precuneus in patients with PD-MCI compared to PD-NC, which might provide a novel insight into the diagnosis and clinical treatment of cognitive impairment in PD.

Whole-brain morphometric studies in alcohol addicts by voxel-based morphometry.

BACKGROUND: To investigate the morphological changes of the whole-brain in patients with alcohol addiction using voxel-based morphometry (VBM). METHODS: Twenty alcohol addiction individuals and 20 healthy control volunteers, matched in gender, age, handedness and education, were enrolled as the alcohol addiction group and control group, respectively, from June 2017 to February 2018. The two groups underwent magnetic resonance imaging (MRI) scans of the whole-brain using a 3.0T MRI system, and the MRI images were analyzed by VBM. Total brain volume analysis was performed to observe the changes of local brain volume. Independent samples t-test was used for comparison of whole-brain volumes between the two groups. RESULTS: Gray matter volume (GMV) was significantly more reduced in the left middle frontal gyrus, left superior frontal gyrus, left precuneate gyrus, left gyrus rectus, right medial superior frontal gyrus, and right orbital gyrus in patients with alcohol addiction than in healthy controls. White matter volume (WMV) in the bilateral superior frontal gyrus was also significantly more reduced in patients with alcohol addiction than in healthy controls. Whole-brain VBM results showed that white matter and brain parenchyma volumes (BPVs) were significantly smaller in patients with alcohol addiction than in healthy controls (both P<0.05). There was no significant difference in GMV between patients with alcohol addiction and healthy controls (P>0.05). CONCLUSIONS: There are dominant areas of brain atrophy in patients with alcohol addiction. The VBM has a potential application value in detecting subtle brain atrophy in patients with alcohol addiction and providing an imaging basis in the diagnosis of alcohol addiction.

Preliminary study on cervical spinal cord in patients with amyotrophic lateral sclerosis using MR diffusion tensor imaging.

RATIONALE AND OBJECTIVES: To investigate the conventional magnetic resonance (MR) findings of cervical spinal cord, to explore the possible changes on diffusion tensor imaging (DTI) in patients with amyotrophic lateral sclerosis (ALS), and to assess the correlations between the changes on DTI and clinical parameters in patients with ALS. MATERIALS AND METHODS: Conventional MR imaging (MRI) and DTI in 24 patients with ALS and 16 age-matched control subjects were obtained. On axial planes, regions of interest (ROIs) were marked in bilateral spinothalamic tracts (STs), posterior funiculus, and bilateral lateral corticospinal tracts (LCTs), respectively, at the levels of cervical 2-4 vertebral bodies. Fractional anisotropy (FA) and apparent diffusion coefficient (ADC) values of these ROIs were estimated. Independent sample t test and Pearson correlation analysis were used. RESULTS: In patients with ALS, no abnormal findings were noted in the cervical spinal cord on conventional MRI. FA values of bilateral LCTs decreased significantly compared to those of the control group (P < .05), and ADC values of bilateral LCTs were significantly greater than those of the control group (P < .05). FA and ADC values of bilateral LCTs showed no significant difference between patients with definite and probable ALS (P > .05). No significant correlation existed between abnormal DTI parameters (FA and ADC values of bilateral LCTs) and clinical parameters (P > .05). CONCLUSIONS: Subtle abnormalities in bilateral LCTs in the "normal-appearing" cervical spinal cord can be detected using quantitative DTI technique in patients with ALS.