Canonical text (the exact bytes fingerprinted)

TRUVACE RECORD VERSION
record: TRV-2026-0917
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-08-28T06:06:59.189792Z
status: published
lens: g_space
sector: health
headline: Noninvasive Profiling of the Glioma Vascular Microenvironment via 7T MRI: Decoding Angiogenic Signatures for Isocitrate Dehydrogenase and World Health Organization Grade Differentiation
dek: Accurate preoperative glioma grading and molecular subtyping are important for treatment. The vascular microenvironment promotes tumor progression. A noninvasive screening tool capable of mapping tumor vascularity may assist in preoperative grading and subtyping of gliomas. To evaluate 7T susceptibility-weighted imaging (SWI) for differentiating glioma isocitrate dehydrogenase (IDH) status and World Health Organization (WHO) grade based on vascular microenvironment features. Retrospective. Among 218 patients wit…
gain_title: Machine learning models using 7T SWI vascular topology, density and intensity features enabled noninvasive preoperative differentiation of glioma IDH status and WHO grade, with improved performance when integrated with clinical factors.
problem_title: (none)
trace_subject: (none)
gain_reading: Machine learning models using 7T SWI vascular topology, density and intensity features enabled noninvasive preoperative differentiation of glioma IDH status and WHO grade, with improved performance when integrated with clinical factors.
gain_evidence: Integrating 7T SWI vascular features with machine learning may aid noninvasive differentiation of glioma molecular status and grade. | Imaging models achieved AUCs of 0.816-0.843 for IDH status and 0.834-0.874 for WHO grade in the internal validation set.
problem_reading: (none)
problem_evidence: (none)
quick_read: Researchers evaluated 7T susceptibility-weighted imaging to map glioma vascular microenvironment features for preoperative differentiation. In 218 histologically confirmed cases, they extracted vascular topology, density and intensity after Frangi filtering and 3D skeletonization and trained machine learning models to distinguish IDH-mutant versus wildtype and WHO Grade 1-2 versus 3-4.

Accurate noninvasive grading and molecular subtyping could inform treatment planning without biopsy, but the reported AUCs come from a retrospective internal validation set as of August 2026. Whether the nine retained vascular features generalize across scanners, institutions, and prospective cohorts remains untested in this source.
limitation: Retrospective single-cohort design with only internal validation and no external or prospective testing reported, limiting generalizability of performance estimates.
tag: Evidence-backed gain
key_points: Retrospective study of 218 patients with histologically confirmed gliomas, split 152 training and 66 validation by stratified random sampling. | Vascular features extracted after Frangi filtering and 3D skeletonization from tumors delineated by two independent neuroradiologists on 7T SWI. | Four classifiers tested: logistic regression, support vector machine, naive Bayes, and random forest, with nine robust vascular features retained after selection.
rundown: Among 218 patients, training/validation sets included 61/26 IDH-mutant, 91/40 IDH-wildtype, 57/25 Grade 1-2, and 95/41 Grade 3-4 gliomas. Imaging used 7T T1-weighted MP2RAGE and SWI, with interobserver agreement assessed by Cohen's kappa and ICC.

Feature selection used t-test or Mann-Whitney U test and Spearman's rank correlation, and performance was reported with sensitivity, specificity, accuracy, F1-score, and AUC. Integrating clinical factors improved optimal AUCs to 0.888 for IDH status and 0.889 for WHO grade.
sources:
- peer_reviewed | Journal of Magnetic Resonance Imaging | https://doi.org/10.1002/jmri.70523 | 2026-08-26
prev: 0000000000000000000000000000000000000000000000000000000000000000

sha256

3d6c7e4e6cbaf2ad21b759abf8ceb7ccd14859f30b6765757555fa74fc21a0b6

previous

0000000000000000000000000000000000000000000000000000000000000000