Reliability and Value of 3D Sequential QUantitative T1 -T2 -T2 * MAppings (SQUMA) MR Multi-Parametric Imaging in Characterizing Carotid Artery Atherosclerosis.

BACKGROUND: T1 , T2 , and T2 * mappings are seldom performed in a single examination, and their values in evaluating symptomatic atherosclerosis are lacking. PURPOSE: To perform three-dimensional (3D) quantitative T1 , T2 , and T2 * mappings (SQUMA) multi-parametric imaging for carotid vessel wall and evaluate its reliability and value in assessing carotid atherosclerosis. STUDY TYPE: Prospective. SUBJECTS: Eight healthy subjects and 20 patients with symptomatic carotid atherosclerosis. FIELD STRENGTH/SEQUENCE: 3 T, SQUMA imaging T1 -, T2 -, and T2 *-mapping, multi-contrast vessel wall imaging including T1 - and T2 -weighted, time-of-flight, and SNAP sequences. ASSESSMENT: SQUMA was acquired in all subjects and multi-contrast images were acquired in healthy subjects. T1 , T2 , and T2 * values and lumen area (LA), wall area (WA), mean wall thickness (MeanWT), and normalized wall index (NWI) of carotid arteries were measured. SQUMA and multi-contrast measurements were compared in healthy subjects and differences in SQUMA measurements between healthy subjects and patients were assessed. The discriminative value of SQUMA measurements for symptomatic vessel was determined. STATISTICAL TESTS: Paired t or Wilcoxon signed-rank test, independent t or Mann-Whitney U test, area under the receiver operating characteristic curve (AUC), intraclass correlation coefficients, and Bland-Altman plots. Statistically significant level, P < 0.05. RESULTS: There were no significant differences in LA (P = 0.340), WA (P = 0.317), MeanWT (P = 0.088), and NWI (P = 0.091) of carotid arteries between SQUMA and multi-contrast vessel wall images. The values of T2 (50.9 ± 2.9 msec vs. 44.5 ± 4.2 msec), T2 * (28.2 ± 4.3 msec vs. 24.7 ± 2.6 msec), WA (23.7 ± 4.6 mm2 vs. 36.2 ± 7.7 mm2 ), MeanWT (0.99 ± 0.05 mm vs. 1.50 ± 0.28 mm), and NWI (40.7 ± 3.0% vs. 53.8 ± 5.4%) of carotid arteries in healthy subjects were significantly different from those in atherosclerotic patients. The combination of quantitative T1 , T2 , and T2 * values and MeanWT showed greatest AUC (0.81; 95% CI: 0.65-0.92) in discriminating symptomatic vessels. DATA CONCLUSION: Carotid MR 3D quantitative multi-parametric imaging of SQUMA enables acquisition of T1 , T2 , and T2 * maps, reliably measuring carotid morphology and discriminating carotid symptomatic atherosclerosis. LEVEL OF EVIDENCE: 2 TECHNICAL EFFICACY: Stage 2.

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.