An artificial intelligence framework for automatic segmentation and volumetry of vestibular schwannomas from contrast-enhanced T1-weighted and high-resolution T2-weighted MRI

Abstract

<jats:sec><jats:title>OBJECTIVE</jats:title><jats:p>Automatic segmentation of vestibular schwannomas (VSs) from MRI could significantly improve clinical workflow and assist in patient management. Accurate tumor segmentation and volumetric measurements provide the best indicators to detect subtle VS growth, but current techniques are labor intensive and dedicated software is not readily available within the clinical setting. The authors aim to develop a novel artificial intelligence (AI) framework to be embedded in the clinical routine for automatic delineation and volumetry of VS.</jats:p></jats:sec><jats:sec><jats:title>METHODS</jats:title><jats:p>Imaging data (contrast-enhanced T1-weighted [ceT1] and high-resolution T2-weighted [hrT2] MR images) from all patients meeting the study’s inclusion/exclusion criteria who had a single sporadic VS treated with Gamma Knife stereotactic radiosurgery were used to create a model. The authors developed a novel AI framework based on a 2.5D convolutional neural network (CNN) to exploit the different in-plane and through-plane resolutions encountered in standard clinical imaging protocols. They used a computational attention module to enable the CNN to focus on the small VS target and propose a supervision on the attention map for more accurate segmentation. The manually segmented target tumor volume (also tested for interobserver variability) was used as the ground truth for training and evaluation of the CNN. We quantitatively measured the Dice score, average symmetric surface distance (ASSD), and relative volume error (RVE) of the automatic segmentation results in comparison to manual segmentations to assess the model’s accuracy.</jats:p></jats:sec><jats:sec><jats:title>RESULTS</jats:title><jats:p>Imaging data from all eligible patients (n = 243) were randomly split into 3 nonoverlapping groups for training (n = 177), hyperparameter tuning (n = 20), and testing (n = 46). Dice, ASSD, and RVE scores were measured on the testing set for the respective input data types as follows: ceT1 93.43%, 0.203 mm, 6.96%; hrT2 88.25%, 0.416 mm, 9.77%; combined ceT1/hrT2 93.68%, 0.199 mm, 7.03%. Given a margin of 5% for the Dice score, the automated method was shown to achieve statistically equivalent performance in comparison to an annotator using ceT1 images alone (p = 4e−13) and combined ceT1/hrT2 images (p = 7e−18) as inputs.</jats:p></jats:sec><jats:sec><jats:title>CONCLUSIONS</jats:title><jats:p>The authors developed a robust AI framework for automatically delineating and calculating VS tumor volume and have achieved excellent results, equivalent to those achieved by an independent human annotator. This promising AI technology has the potential to improve the management of patients with VS and potentially other brain tumors.</jats:p></jats:sec>

Manual segmentation versus semi-automated segmentation for quantifying vestibular schwannoma volume on MRI

“…Deep learning is a sub-type of artificial intelligence that utilises multiple layers of analysis to process an image. A variety of applications of deep learning are postulated [ 7 , 20 , 42 ], and one study has shown this to be a useful approach in automated VS segmentation [ 32 ] in terms of both time and accuracy. Despite the accuracy of automated approaches, interactive corrections may continue to play a role even with deep learning due to the lack of adaptability of automated methods to the specific imaging sequences and protocols used clinically [ 39 ].…”

“…Automated segmentation may be accurate in the assessment of tumour progression and in overall survival prediction in glioma [1,26] as well as for the clinical assessment of biomarkers in glioma [4]. For VS imaging, automated segmentation has been applied with positive results [32,40] and there is growing interest in the field [10]. An automated segmentation tool could also improve clinical workflow and operational efficiency during the planning of stereotactic radiosurgery (SRS) by using the tool as an initialisation step in the process.…”

Deep learning for glioblastoma segmentation using preoperative magnetic resonance imaging identifies volumetric features associated with survival

“…A major advantage of our method is that it is applicable to data from different scanners and institutions. Furthermore, curating a large sample of uniform images for deep learning training is time-and labour-intensive [28]. Even when training time is considered, use of our computer-assisted segmentation method can reduce annotation time by 20 min per patient compared to fully manual segmentation.…”

“…It is one of the largest studies investigating volumetric features and prognosis incorporating quantitative measurement of postoperative imaging and clinical variables. Future work to evaluate our approach should quantitatively compare segmentation measurements with manual segmentations from multiple observers to assess inter-rater variability [28]. In addition, an independent dataset is necessary for us to compare the relative prognostic performance of automated segmentations against manual segmentations and determine the reproducibility of our segmentations.…”

Editorial for: “Primary Central Nervous System Lymphoma: Clinical Evaluation of Automated Segmentation on Multiparametric MRI Using Deep Learning”

“…Recently, artificial intelligence has shown great potential in the application of medical imaging, especially the deep-learning model based on the neural network in the detection of diseases, segmentation, and quantitative evaluation of the lesions. [4][5][6][7][8] In this issue of the Journal of Magnetic Resonance Imaging, Pennig et al explored the performance of a deep-learning model (DLM) initially trained on gliomas to detect and segment PCNSL using multiparametric MRI at different field strengths from several imaging manufacturers, models, and study centers. 9 They used a DLM-based 3D convolutional neural network to achieve fully automated voxelwise segmentation of tumor components in a cohort of 43 patients with histologically proven PCNSL, which was compared to manual ground truth segmentations by two observers in contrastenhanced T 1 -weighted and fluid-attenuated inversion recovery images.…”