NnU-Net versus mesh growing algorithm as a tool for the robust and timely segmentation of neurosurgical 3D images in contrast-enhanced T1 MRI scans.

PURPOSE: This study evaluates the nnU-Net for segmenting brain, skin, tumors, and ventricles in contrast-enhanced T1 (T1CE) images, benchmarking it against an established mesh growing algorithm (MGA). METHODS: We used 67 retrospectively collected annotated single-center T1CE brain scans for training models for brain, skin, tumor, and ventricle segmentation. An additional 32 scans from two centers were used test performance compared to that of the MGA. The performance was measured using the Dice-Sørensen coefficient (DSC), intersection over union (IoU), 95th percentile Hausdorff distance (HD95), and average symmetric surface distance (ASSD) metrics, with time to segment also compared. RESULTS: The nnU-Net models significantly outperformed the MGA (p < 0.0125) with a median brain segmentation DSC of 0.971 [95CI: 0.945-0.979], skin: 0.997 [95CI: 0.984-0.999], tumor: 0.926 [95CI: 0.508-0.968], and ventricles: 0.910 [95CI: 0.812-0.968]. Compared to the MGA's median DSC for brain: 0.936 [95CI: 0.890, 0.958], skin: 0.991 [95CI: 0.964, 0.996], tumor: 0.723 [95CI: 0.000-0.926], and ventricles: 0.856 [95CI: 0.216-0.916]. NnU-Net performance between centers did not significantly differ except for the skin segmentations Additionally, the nnU-Net models were faster (mean: 1139 s [95CI: 685.0-1616]) than the MGA (mean: 2851 s [95CI: 1482-6246]). CONCLUSIONS: The nnU-Net is a fast, reliable tool for creating automatic deep learning-based segmentation pipelines, reducing the need for extensive manual tuning and iteration. The models are able to achieve this performance despite a modestly sized training set. The ability to create high-quality segmentations in a short timespan can prove invaluable in neurosurgical settings.

Longitudinal diffusion-weighted imaging in infants with hydrocephalus: decrease in tissue water diffusion after cerebrospinal fluid diversion.

OBJECT: Progressive hydrocephalus may lead to edema of the periventricular white matter and to damage of the brain parenchyma because of compression, stretching, and ischemia. The aim of the present study was to investigate whether cerebral edema can be quantified using diffusion-weighted imaging in infants with hydrocephalus and whether CSF diversion could decrease cerebral edema. METHODS: Diffusion-weighted MR imaging was performed in 24 infants with progressive hydrocephalus before and after CSF diversion. Parametric images of the trace apparent diffusion coefficients (ADCs) were obtained. The ADCs of 5 different cortical and subcortical regions of interest were calculated pre- and postoperatively in each patient. The ADC values were compared with age-related normal values. Mean arterial blood pressure and anterior fontanel pressure were measured immediately after each MR imaging study. RESULTS: After CSF diversion, the mean ADC decreased from a preoperative value of 1209 +/- 116 x 10(-6) mm(2)/second to a postoperative value of 928 +/- 64 x 10(-6) mm(2)/second (p < 0.005). Differences between pre- and postoperative ADC values were most prominent in the periventricular white matter, supporting the existence of preoperative periventricular edema. Compared with age-related normal values, the preoperative ADC values were higher and the postoperative ADC values were lower, although within normal range. The decrease in ADC after CSF drainage was more rapid than the more gradual physiological decrease that is related to age. The preoperative ICP was elevated in all patients. After CSF diversion the ICP decreased significantly to within the normal range. A linear correlation between ADC values and ICP was found (correlation coefficient 0.496, p < 0.001). In all patients the mean arterial blood pressure was within physiological limits both pre- and postoperatively. CONCLUSIONS: This study shows a rapid and more extensive decrease in ADC values after CSF diversion than is to be expected from physiological ADC decrease solely due to increasing patient age. The preoperative ADC increase can be explained by interstitial edema caused by transependymal CSF leakage or by vasogenic edema caused by capillary compression and stretching of the brain parenchyma. This study population of infants with (early recognized) hydrocephalus did not suffer from cytotoxic edema. These findings may help to detect patients at risk for cerebral damage by differentiating between progressive and compensated hydrocephalus.