BACKGROUND AND PURPOSE: Periventricular white matter injury is the common cause of spastic cerebral palsy. However, the early diagnosis of spastic cerebral palsy still remains a challenge. Our aim was to investigate whether infants with periventricular white matter injury with bilateral spastic cerebral palsy have unique lesions different from those in infants without cerebral palsy and to evaluate the efficiency of DTI in the early diagnosis of spastic cerebral palsy. MATERIALS AND METHODS: Infants with periventricular white matter injury and controls underwent MR imaging at 6–18 months of age. Fractional anisotropy was calculated from DTI. Cerebral palsy was diagnosed by 24–30 months of age. Subjects were divided into 3 groups: infants with periventricular white matter injury with bilateral spastic cerebral palsy, infants with periventricular white matter injury without cerebral palsy, and controls. Tract-Based Spatial Statistics and Automated Fiber Quantification were used to investigate intergroup differences. Receiver operating characteristic curves were used to assess the diagnostic accuracy of spastic cerebral palsy. Correlations between motor function scores and fractional anisotropy were evaluated along white matter tracts. RESULTS: There were 20, 19, and 33 subjects in periventricular white matter injury with spastic cerebral palsy, periventricular white matter injury without cerebral palsy, and control groups, respectively. Decreased fractional anisotropy in the corticospinal tract was only observed in infants with periventricular white matter injury with spastic cerebral palsy, whereas decreased fractional anisotropy in the posterior thalamic radiation and genu and splenium of the corpus callosum was seen in both periventricular white matter injury subgroups. Fractional anisotropy in the corticospinal tract at the internal capsule level was effective in differentiating infants with periventricular white matter injury with spastic cerebral palsy from those without cerebral palsy by a threshold of 0.53, and it had strong correlations with motor function scores. CONCLUSIONS: Corticospinal tract lesions play a crucial role in motor impairment related to spastic cerebral palsy in infants with periventricular white matter injury. Fractional anisotropy in the corticospinal tract at the internal capsule level could aid in the early diagnosis of spastic cerebral palsy with high diagnostic accuracy.

Background: Site-specific variations are challenges for pooling analyses in multi-center studies. This work aims to propose an inter-site harmonization method based on dual generative adversarial networks (GANs) for diffusion tensor imaging (DTI) derived metrics on neonatal brains. Results: DTI-derived metrics (fractional anisotropy, FA; mean diffusivity, MD) are obtained on age-matched neonates without magnetic resonance imaging (MRI) abnormalities: 42 neonates from site 1 and 42 neonates from site 2. Significant inter-site differences of FA can be observed. The proposed harmonization approach and three conventional methods (the global-wise scaling, the voxel-wise scaling, and the ComBat) are performed on DTI-derived metrics from two sites. During the tract-based spatial statistics, inter-site differences can be removed by the proposed dual GANs method, the voxel-wise scaling, and the ComBat. Among these methods, the proposed method holds the lowest median values in absolute errors and root mean square errors. During the pooling analysis of two sites, Pearson correlation coefficients between FA and the postmenstrual age after harmonization are larger than those before harmonization. The effect sizes (Cohen’s d between males and females) are also maintained by the harmonization procedure. Conclusions: The proposed dual GANs-based harmonization method is effective to harmonize neonatal DTI-derived metrics from different sites. Results in this study further suggest that the GANs-based harmonization is a feasible pre-processing method for pooling analyses in multi-center studies.

Background: Simple febrile seizures (SFS) and epilepsy are common seizures in childhood. However, the mechanism underlying SFS is uncertain, and the presence of obvious variances in white matter (WM) integrity and glymphatic function between SFS and epilepsy remain unclear. Therefore, this study aimed to investigate the differences in WM integrity and glymphatic function between SFS and epilepsy.Material and Methods: We retrospectively included 26 children with SFS, 33 children with epilepsy, and 28 controls aged 6–60 months who underwent magnetic resonance imaging (MRI). Tract-based spatial statistics (TBSS) were used to compare the diffusion tensor imaging (DTI) metrics of WM among the above-mentioned groups. T2-weighted imaging (T2WI) was used to segment the visible Virchow-Robin space (VRS) through a custom-designed automated method. VRS counts and volume were quantified and compared among the SFS, epilepsy, and control groups. Correlations of the VRS metrics and seizure duration and VRS metrics and the time interval between seizure onset and MRI scan were also investigated.Results: In comparison with controls, children with SFS showed no significant changes in fractional anisotropy (FA), axial diffusivity (AD), or radial diffusivity (RD) in the WM (P > 0.05). Decreased FA, unchanged AD, and increased RD were observed in the epilepsy group in comparison with the SFS and control groups (P < 0.05). Meanwhile, VRS counts were higher in the SFS and epilepsy groups than in the control group (VRS_SFS, 442.42 ± 74.58, VRS_epilepsy, 629.94 ± 106.55, VRS_control, 354.14 ± 106.58; P < 0.001), and similar results were found for VRS volume (VRS_SFS, 6,228.18 ± 570.74 mm³, VRS_epilepsy, 9,684.84 ± 7,292.66mm³, VRS_control, 4,007.22 ± 118.86 mm³; P < 0.001). However, VRS metrics were lower in the SFS group than in the epilepsy group (P < 0.001). In both SFS and epilepsy, VRS metrics positively correlated with seizure duration and negatively correlated with the course after seizure onset.Conclusion: SFS may not be associated with WM microstructural disruption; however, epilepsy is related to WM alterations. Seizures are associated with glymphatic dysfunction in either SFS or epilepsy.

Background: Chest radiography is the standard investigation for identifying rib fractures. The application of artificial intelligence (AI) for detecting rib fractures on chest radiographs is limited by image quality control and multilesion screening. To our knowledge, few studies have developed and verified the performance of an AI model for detecting rib fractures by using multi-center radiographs. And existing studies using chest radiographs for multiple rib fracture detection have used more complex and slower detection algorithms, so we aimed to create a multiple rib fracture detection model by using a convolutional neural network (CNN), based on multi-center and quality-normalised chest radiographs. Methods: A total of 1080 radiographs with rib fractures were obtained and randomly divided into the training set (918 radiographs, 85%) and the testing set (162 radiographs, 15%). An object detection CNN, You Only Look Once v3 (YOLOv3), was adopted to build the detection model. Receiver operating characteristic (ROC) and free-response ROC (FROC) were used to evaluate the model’s performance. A joint testing group of 162 radiographs with rib fractures and 233 radiographs without rib fractures was used as the internal testing set. Furthermore, an additional 201 radiographs, 121 with rib fractures and 80 without rib fractures, were independently validated to compare the CNN model performance with the diagnostic efficiency of radiologists. Results: The sensitivity of the model in the training and testing sets was 92.0% and 91.1%, respectively, and the precision was 68.0% and 81.6%, respectively. FROC in the testing set showed that the sensitivity for whole-lesion detection reached 91.3% when the false-positive of each case was 0.56. In the joint testing group, the case-level accuracy, sensitivity, specificity, and area under the curve were 85.1%, 93.2%, 79.4%, and 0.92, respectively. At the fracture level and the case level in the independent validation set, the accuracy and sensitivity of the CNN model were always higher or close to radiologists’ readings. Conclusions: The CNN model, based on YOLOv3, was sensitive for detecting rib fractures on chest radiographs and showed great potential in the preliminary screening of rib fractures, which indicated that CNN can help reduce missed diagnoses and relieve radiologists’ workload. In this study, we developed and verified the performance of a novel CNN model for rib fracture detection by using radiography.

BACKGROUND AND PURPOSE: Punctate white matter lesions are common in preterm neonates. Neurodevelopmental outcomes of the neonates are related to the degree of extension. This study aimed to characterize the extent of microstructural variations for different punctate white matter lesion grades. MATERIALS AND METHODS: Preterm neonates with punctate white matter lesions were divided into 3 grades (from mild to severe: grades I–III). DTI-derived fractional anisotropy, axial diffusivity, and radial diffusivity between patients with punctate white matter lesions and controls were compared with Tract-Based Spatial Statistics and tract-quantification methods. RESULTS: Thirty-three preterm neonates with punctate white matter lesions and 33 matched controls were enrolled. There were 15, 9, and 9 patients, respectively, in grades I, II, and III. Punctate white matter lesions were mainly located in white matter adjacent to the lateral ventricles, especially regions lateral to the trigone, posterior horns, and centrum semiovale and/or corona radiata. Extensive microstructural changes were observed in neonates with grade III punctate white matter lesions, while no significant changes in DTI metrics were found for grades I and II. A pattern of increased axial diffusivity, increased radial diffusivity, and reduced/unchanged fractional anisotropy was found in regions adjacent to punctate white matter lesion sites seen on T1WI and T2WI. Unchanged axial diffusivity, increased radial diffusivity, and reduced/unchanged fractional anisotropy were observed in regions distant from punctate white matter lesion sites. CONCLUSIONS: White matter microstructural variations were different across punctate white matter lesion grades. Extensive change patterns varied according to the distance to the lesion sites in neonates with severe punctate white matter lesions. These findings may help in determining the outcomes of punctate white matter lesions and selecting treatment strategies.