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.

Fully Automatic Adaptive Meshing Based Segmentation of the Ventricular System for Augmented Reality Visualization and Navigation.

OBJECTIVE: Effective image segmentation of cerebral structures is fundamental to 3-dimensional techniques such as augmented reality. To be clinically viable, segmentation algorithms should be fully automatic and easily integrated in existing digital infrastructure. We created a fully automatic adaptive-meshing-based segmentation system for T1-weighted magnetic resonance images (MRI) to automatically segment the complete ventricular system, running in a cloud-based environment that can be accessed on an augmented reality device. This study aims to assess the accuracy and segmentation time of the system by comparing it to a manually segmented ground truth dataset. METHODS: A ground truth (GT) dataset of 46 contrast-enhanced and non-contrast-enhanced T1-weighted MRI scans was manually segmented. These scans also were uploaded to our system to create a machine-segmented (MS) dataset. The GT data were compared with the MS data using the Sørensen-Dice similarity coefficient and 95% Hausdorff distance to determine segmentation accuracy. Furthermore, segmentation times for all GT and MS segmentations were measured. RESULTS: Automatic segmentation was successful for 45 (98%) of 46 cases. Mean Sørensen-Dice similarity coefficient score was 0.83 (standard deviation [SD] = 0.08) and mean 95% Hausdorff distance was 19.06 mm (SD = 11.20). Segmentation time was significantly longer for the GT group (mean = 14405 seconds, SD = 7089) when compared with the MS group (mean = 1275 seconds, SD = 714) with a mean difference of 13,130 seconds (95% confidence interval 10,130-16,130). CONCLUSIONS: The described adaptive meshing-based segmentation algorithm provides accurate and time-efficient automatic segmentation of the ventricular system from T1 MRI scans and direct visualization of the rendered surface models in augmented reality.

Fully automatic brain tumor segmentation for 3D evaluation in augmented reality.

OBJECTIVE: For currently available augmented reality workflows, 3D models need to be created with manual or semiautomatic segmentation, which is a time-consuming process. The authors created an automatic segmentation algorithm that generates 3D models of skin, brain, ventricles, and contrast-enhancing tumor from a single T1-weighted MR sequence and embedded this model into an automatic workflow for 3D evaluation of anatomical structures with augmented reality in a cloud environment. In this study, the authors validate the accuracy and efficiency of this automatic segmentation algorithm for brain tumors and compared it with a manually segmented ground truth set. METHODS: Fifty contrast-enhanced T1-weighted sequences of patients with contrast-enhancing lesions measuring at least 5 cm3 were included. All slices of the ground truth set were manually segmented. The same scans were subsequently run in the cloud environment for automatic segmentation. Segmentation times were recorded. The accuracy of the algorithm was compared with that of manual segmentation and evaluated in terms of Sørensen-Dice similarity coefficient (DSC), average symmetric surface distance (ASSD), and 95th percentile of Hausdorff distance (HD95). RESULTS: The mean ± SD computation time of the automatic segmentation algorithm was 753 ± 128 seconds. The mean ± SD DSC was 0.868 ± 0.07, ASSD was 1.31 ± 0.63 mm, and HD95 was 4.80 ± 3.18 mm. Meningioma (mean 0.89 and median 0.92) showed greater DSC than metastasis (mean 0.84 and median 0.85). Automatic segmentation had greater accuracy for measuring DSC (mean 0.86 and median 0.87) and HD95 (mean 3.62 mm and median 3.11 mm) of supratentorial metastasis than those of infratentorial metastasis (mean 0.82 and median 0.81 for DSC; mean 5.26 mm and median 4.72 mm for HD95). CONCLUSIONS: The automatic cloud-based segmentation algorithm is reliable, accurate, and fast enough to aid neurosurgeons in everyday clinical practice by providing 3D augmented reality visualization of contrast-enhancing intracranial lesions measuring at least 5 cm3. The next steps involve incorporation of other sequences and improving accuracy with 3D fine-tuning in order to expand the scope of augmented reality workflow.

Current Accuracy of Augmented Reality Neuronavigation Systems: Systematic Review and Meta-Analysis.

BACKGROUND: Augmented reality neuronavigation (ARN) systems can overlay three-dimensional anatomy and disease without the need for a two-dimensional external monitor. Accuracy is crucial for their clinical applicability. We performed a systematic review regarding the reported accuracy of ARN systems and compared them with the accuracy of conventional infrared neuronavigation (CIN). METHODS: PubMed and Embase were searched for ARN and CIN systems. For ARN, type of system, method of patient-to-image registration, accuracy method, and accuracy of the system were noted. For CIN, navigation accuracy, expressed as target registration error (TRE), was noted. A meta-analysis was performed comparing the TRE of ARN and CIN systems. RESULTS: Thirty-five studies were included, 12 for ARN and 23 for CIN. ARN systems could be divided into head-mounted display and heads-up display. In ARN, 4 methods were encountered for patient-to-image registration, of which point-pair matching was the one most frequently used. Five methods for assessing accuracy were described. Ninety-four TRE measurements of ARN systems were compared with 9058 TRE measurements of CIN systems. Mean TRE was 2.5 mm (95% confidence interval, 0.7-4.4) for ARN systems and 2.6 mm (95% confidence interval, 2.1-3.1) for CIN systems. CONCLUSIONS: In ARN, there seems to be lack of agreement regarding the best method to assess accuracy. Nevertheless, ARN systems seem able to achieve an accuracy comparable to CIN systems. Future studies should be prospective and compare TREs, which should be measured in a standardized fashion.

Clinical Accuracy of Holographic Navigation Using Point-Based Registration on Augmented-Reality Glasses.

BACKGROUND: As current augmented-reality (AR) smart glasses are self-contained, powerful computers that project 3-dimensional holograms that can maintain their position in physical space, they could theoretically be used as a low-cost, stand-alone neuronavigation system. OBJECTIVE: To determine feasibility and accuracy of holographic neuronavigation (HN) using AR smart glasses. METHODS: We programmed a fully functioning neuronavigation system on commercially available smart glasses (HoloLens®, Microsoft, Redmond, Washington) and tested its accuracy and feasibility in the operating room. The fiducial registration error (FRE) was measured for both HN and conventional neuronavigation (CN) (Brainlab, Munich, Germany) by using point-based registration on a plastic head model. Subsequently, we measured HN and CN FRE on 3 patients. RESULTS: A stereoscopic view of the holograms was successfully achieved in all experiments. In plastic head measurements, the mean HN FRE was 7.2 ± 1.8 mm compared to the mean CN FRE of 1.9 ± 0.45 (mean difference: -5.3 mm; 95% confidence interval [CI]: -6.7 to -3.9). In the 3 patients, the mean HN FRE was 4.4 ± 2.5 mm compared to the mean CN FRE of 3.6 ± 0.5 (mean difference: -0.8 mm; 95% CI: -3.0 to 4.6). CONCLUSION: Owing to the potential benefits and promising results, we believe that HN could eventually find application in operating rooms. However, several improvements will have to be made before the device can be used in clinical practice.

[Augmented reality glasses in the operating room]. / Augmented-realitybril in de operatiekamer.

Augmented reality is a technology that makes use of special glasses to combine various virtual images, such as holograms and scans, with reality. This technology offers important advantages for surgery in particular, because stereoscopic three-dimensional images of anatomical structures can be projected almost perfectly on the immobilised patient before and during surgery. This technology also has a lot of potential when it comes to education and providing information to patients. There are still some major obstacles, but it is expected that this technology will eventually find its way to virtually all operating rooms.