Search

Extent of resection after surgery is one of the main prognostic factors for patients diagnosed with glioblastoma. To achieve this, accurate segmentation and classification of residual tumor from post-operative MR images is essential. The current standard method for estimating it is subject to high inter- and intra-rater variability, and an automated method for segmentation of residual tumor in early post-operative MRI could lead to a more accurate estimation of extent of resection. In this study, two state-of-the-art neural network architectures for pre-operative segmentation were trained for the task. The models were extensively validated on a multicenter dataset with nearly 1000 patients, from 12 hospitals in Europe and the United States. The best performance achieved was a 61% Dice score, and the best classification performance was about 80% balanced accuracy, with a demonstrated ability to generalize across hospitals. In addition, the segmentation performance of the best models was on par with human expert raters. The predicted segmentations can be used to accurately classify the patients into those with residual tumor, and those with gross total resection., (© 2023. The Author(s).)

Background: Neurosurgical resection represents an important treatment option in the modern, multimodal therapy approach of brain metastases (BM). Guidelines for perioperative imaging exist for primary brain tumors to guide postsurgical treatment. Optimal perioperative imaging of BM patients is so far a matter of debate as no structured guidelines exist., Methods: A comprehensive questionnaire about perioperative imaging was designed by the European Association of Neuro-Oncology (EANO) Youngsters Committee. The survey was distributed to physicians via the EANO network to perform a descriptive overview on the current habits and their variability on perioperative imaging. Chi square test was used for dichotomous variables., Results: One hundred twenty physicians worldwide responded to the survey. MRI was the preferred preoperative imaging method (93.3%). Overall 106/120 (88.3%) physicians performed postsurgical imaging routinely including MRI alone (62/120 [51.7%]), postoperative CT (29/120 [24.2%]) and MRI + CT (15/120 [12.5%]). No correlation of postsurgical MRI utilization in academic vs. non-academic hospitals (58/89 [65.2%] vs. 19/31 [61.3%], p = 0.698) was found. Early postoperative MRI within ≤72 h after resection is obtained by 60.8% of the participants. The most frequent reason for postsurgical imaging was to evaluate the extent of tumor resection (73/120 [60.8%]). In case of residual tumor, 32/120 (26.7%) participants indicated to adjust radiotherapy, 34/120 (28.3%) to consider re-surgery to achieve complete resection and 8/120 (6.7%) to evaluate both., Conclusions: MRI was the preferred imaging method in the preoperative setting. In the postoperative course, imaging modalities and timing showed high variability. International guidelines for perioperative imaging with special focus on postoperative MRI to assess residual tumor are warranted to optimize standardized management and adjuvant treatment decisions for BM patients.