Automated unruptured cerebral aneurysms detection in TOF MR angiography images using dual-channel SE-3D UNet: a multi-center research.

OBJECTIVES: Time of flight magnetic resonance angiography (TOF-MRA) is the primary non-invasive screening method for cerebral aneurysms. We aimed to develop a computer-aided aneurysm detection method to improve the diagnostic efficiency and accuracy, especially decrease the false positive rate. METHODS: This is a retrospective multicenter study. The dataset contained 1160 TOF-MRA examinations composed of unruptured aneurysms (n = 1096) and normal controls (n = 166) from six hospitals. A total of 1037 examinations acquired from 2013 to 2019 were used as training set; 123 examinations acquired from 2020 to 2021 were used as external test set. We proposed an equalized augmentation strategy based on aneurysm location and constructed a detection model based on dual channel SE-3D UNet. The model was trained with a 5-fold cross-validation in the training set, then tested on the external test set. RESULTS: The proposed method achieved 82.46% sensitivity on patient-level, 73.85% sensitivity on lesion-level, and 0.88 false positives per case in the external test set. The performance did not show significant differences in subgroups according to the aneurysm site (except ACA), aneurysm size (except smaller than 3 mm), or MRI scanners. The performance preceded the basic SE-3D UNet by increasing 15.79% patient-level sensitivity and decreasing 4.19 FPs/case. CONCLUSIONS: The proposed automated aneurysm detection method achieved acceptable sensitivity while controlling fairly low false positives per case. It might provide a useful auxiliary tool of cerebral aneurysms MRA screening. KEY POINTS: • The need for automated cerebral aneurysms detecting is growing. • The strategy of equalized augmentation based on aneurysm location and dual-channel input could improve the model performance. • The retrospective multi-center study showed that the proposed automated cerebral aneurysms detection using dual-channel SE-3D UNet could achieve acceptable sensitivity while controlling a low false positive rate.

A MRI study of the default mode network subsystems in relapsing-remitting multiple sclerosis patients with preserved cognitive function / 中华放射学杂志

Objective:To investigate the changes in gray matter volume of the subsystems as well as intra-subsystem and inter-subsystem functional connectivity in the default mode network (DMN) of relapsing-remitting multiple sclerosis (RRMS) patients with preserved cognitive function.Methods:In this prospective study, thirty-seven RRMS patients with preserved cognitive function who were admitted to Huashan Hospital of Fudan University from April 2020 to January 2021 (RRMS group) and 43 healthy volunteers (HC group) were recruited. Patients in the RRMS group received the cognitive assessment using a clinical cognitive functioning scale. Three-dimensional T 1WI and resting-state functional MRI were performed to obtain the brain structural and functional data. The DMN was divided into three subsystems: CORE, dorsal medial prefrontal cortex (DMPFC), and medial temporal lobe (MTL). The gray matter volume of the three subsystems were extracted from the gray matter volume map generated by spatial normalization; 24 regions of interest (ROIs) of the DMN were defined based on Yeo′s 17 networks, and their functional connectivity values were calculated to derive the mean intra-subsystem and inter-subsystem functional connectivity values. Differences in gray matter volume and functional connectivity between the RRMS and HC groups were compared using independent sample t-tests; Spearman′s partial correlation was used to analyze the correlation between subsystems′ gray matter volume and functional connectivity, as well as between subsystems′ functional connectivity and clinical scale scores. Results:Compared to the HC group, the gray matter volume of the three subsystems of the DMN were considerably reduced in the RRMS group ( P<0.05). The functional connectivity within and between the three subsystems were not statistically significantly different between the HC and RRMS groups ( P>0.05). Based on the ROI analysis, patients with RRMS the brain regions with significantly reduced DMN intra-subsystem functional connectivity values were mainly located in the left dorsomedial prefrontal cortex of the DMPFC, the right lateral temporal cortex of the DMPFC, and the left medial temporal cortex of the MTL, as compared with the HC group ( P<0.01). The gray matter volume of DMPFC was positively correlated with the functional connectivity within DMPFC in the control group ( r=0.326, P=0.040). In the RRMS group, the gray matter volume of CORE was positively correlated with the functional connectivity between CORE and DMPFC ( r=0.363, P=0.038), and the functional connectivity within CORE was positively correlated with scores on the memory and executive screening scale ( r=0.430, P=0.036). Conclusions:RRMS patients with preserved cognitive function exhibit gray matter atrophy in all three DMN subsystems. There is no correlation between the structure and function of the DMPFC subsystem. The functional connectivity within CORE subsystem may reflect memory and execution status; DMPFC and CORE may be critical encephalic regions for neurodegeneration and brain functional changes in RRMS patients with preserved cognitive function.