Spectroscopie par résonance magnétique des tumeurs cérébrales

Precision Neurosurgery: A Path Forward

Since the inception of their profession, neurosurgeons have defined themselves as physicians with a surgical practice. Throughout time, neurosurgery has always taken advantage of technological advances to provide better and safer care for patients. In the ongoing precision medicine surge that drives patient-centric healthcare, neurosurgery strives to effectively embrace the era of data-driven medicine. Neuro-oncology best illustrates this convergence between surgery and precision medicine with the advent of molecular profiling, imaging and data analytics. This convenient convergence paves the way for new preventive, diagnostic, prognostic and targeted therapeutic perspectives. The prominent advances in healthcare and big data forcefully challenge the medical community to deeply rethink current and future medical practice. This work provides a historical perspective on neurosurgery. It also discusses the impact of the conceptual shift of precision medicine on neurosurgery through the lens of neuro-oncology.

Development of a Machine Learning Classifier Based on Radiomic Features Extracted From Post-Contrast 3D T1-Weighted MR Images to Distinguish Glioblastoma From Solitary Brain Metastasis

Objectives

To differentiate Glioblastomas (GBM) and Brain Metastases (BM) using a radiomic features-based Machine Learning (ML) classifier trained from post-contrast three-dimensional T1-weighted (post-contrast 3DT1) MR imaging, and compare its performance in medical diagnosis versus human experts, on a testing cohort.

Methods

We enrolled 143 patients (71 GBM and 72 BM) in a retrospective bicentric study from January 2010 to May 2019 to train the classifier. Post-contrast 3DT1 MR images were performed on a 3-Tesla MR unit and 100 radiomic features were extracted. Selection and optimization of the Machine Learning (ML) classifier was performed using a nested cross-validation. Sensitivity, specificity, balanced accuracy, and area under the receiver operating characteristic curve (AUC) were calculated as performance metrics. The model final performance was cross-validated, then evaluated on a test set of 37 patients, and compared to human blind reading using a McNemar’s test.

Results

The ML classifier had a mean [95% confidence interval] sensitivity of 85% [77; 94], a specificity of 87% [78; 97], a balanced accuracy of 86% [80; 92], and an AUC of 92% [87; 97] with cross-validation. Sensitivity, specificity, balanced accuracy and AUC were equal to 75, 86, 80 and 85% on the test set. Sphericity 3D radiomic index highlighted the highest coefficient in the logistic regression model. There were no statistical significant differences observed between the performance of the classifier and the experts’ blinded examination.

Conclusions

The proposed diagnostic support system based on radiomic features extracted from post-contrast 3DT1 MR images helps in differentiating solitary BM from GBM with high diagnosis performance and generalizability.

[Imaging methods used in the differential diagnosis between brain tumour relapse and radiation necrosis after stereotactic radiosurgery of brain metastases: Literature review]

After stereotactic radiosurgery for a cerebral metastasis, one of the dreaded toxicities is radionecrosis. In the follow-up of these patients, it is impossible to distinguish radiation necrosis from tumour relapse either clinically or with MRI. In current practice, many imaging methods are designed such as special sequences of MRI (dynamic susceptibility contrast perfusion and susceptibility-weighted imaging, diffusion), proton magnetic resonance spectroscopy, positron emission tomography, or more seldom 201-thallium single-photon emission computerized tomography. This article is a required literature analysis in order to establish a decision tree with the analysis of retrospective and prospective data.

[Imaging of response to treatment in oncology]

Imaging plays a crucial role in oncology to assist in the management of patients and selection of drug regimen. Recent advances in imaging techniques allowing to predict and evaluate response to treatments in oncology will be reviewed. The standard in the evaluation of response to treatment is based on the measurement of lesion size. Functional imaging assesses physiological or molecular processes that may be earlier indicators of early response to treatment. Dynamic imaging of tumor vascularization assesses the biodistribution of a contrast agent within tumoral tissues. Diffusion-weighted MR imaging can differentiate free water from water restricted by tissues, providing an assessment of tumor cellularity. MR spectroscopy assesses the relative quantity of specific chemical components within normal and tumoral tissues. 18 FDG PET imaging provides an assessment of the metabolic activity of tissues. FDG uptake is proportional to cellular proliferation and number of viable cells within a tumor. Results from studies assessing the role of these emerging imaging techniques remain preliminary and the medical community must determine their respective role in the routine evaluation of response to treatment in oncological patients.