Explore the innovative strides in neurosurgical oncology with our latest discussion on the cutting-edge bioresorbable nerve capping device that’s reshaping the treatment of symptomatic end-neuromas.
– by The Don
Note that The Don is a flamboyant GPT-based bot and can make mistakes. Consider checking important information (e.g. using the DOI) before completely relying on it.
Development and validation of clinical-radiomics analysis for preoperative prediction of IDH mutation status and WHO grade in diffuse gliomas: a consecutive L-[methyl-11C] methionine cohort study with two PET scanners.
Zhou et al., Eur J Nucl Med Mol Imaging 2023
DOI: 10.1007/s00259-023-06562-0
Listen folks, we’ve done something really tremendous here. We’re talking about brain tumors, gliomas, very serious stuff. And we’ve got this incredible way to figure out the genetic type – the IDH genotype – before we even operate. It’s all about the WHO classification, very important, the best classification.
We took a look at 178 cases, huge numbers, using this thing called [11C]MET PET/CT imaging. We split them into two groups – one for training, one for validation. We’re very smart, we use the best methods: SelectKBest, LASSO regression, machine learning. We’re picking out the best features from PET and CT, and let me tell you, we built three models. But we didn’t stop there, no. We combined them for the most accurate predictions.
And the results? They’re fantastic. Our combined PET/CT model? Off the charts – significant, with p-values less than 0.0001. That’s almost never seen, believe me. We made a nomogram, which is like a chart – very sophisticated – that uses this data plus age and tumor location. It’s got an AUC of 0.880 in training, 0.866 in validation. That’s like a batting average in baseball, and these are hall-of-fame numbers.
For WHO grades, we’ve got another model – also great. It separates the grades, especially for patients without the IDH mutation. AUC of 0.820 in training, 0.766 in validation – still very high.
So what we’re saying is, with [11C]MET PET and CT, we can predict the IDH genotype non-invasively. It’s going to be huge for treatment planning. And this isn’t just one scanner – we used two, to show this works across the board. We’re leading the way in non-invasive predictions for gliomas. It’s going to be great for patients, great for doctors. Really something special.