Ultra-fast deep-learned CNS tumor classification during surgery

Primary treatment of central nervous system (CNS) tumors includes neurosurgical resection, in which a delicate balance must be struck between maximizing the extent of resection and minimizing risk of neurological damage. However, prior to surgery, surgeons have limited knowledge of the precise tumor type. Current standard practice relies on preoperative imaging and intraoperative histological analysis, but these are not always conclusive and occasionally wrong. Using rapid nanopore sequencing, a sparse methylation profile can be obtained during surgery. Researchers in The Netherlands developed Sturgeon, a patient-agnostic transfer-learned neural network, to enable molecular subclassification of CNS tumors based on such sparse profiles. Sturgeon delivered an accurate diagnosis within 40 minutes of starting sequencing in 45 out of 50 retrospectively sequenced samples. The applicability was demonstrated in real time during 25 surgeries, achieving a diagnostic turnaround time of less than 90 minutes. Of these, 18 (72%) diagnoses were correct. The authors conclude that machine-learned diagnosis based on low-const intraoperative sequencing can assist neurosurgical decision-making, potentially preventing neurological comorbidity and avoiding additional surgeries.