Metabolic Insight into Glioma Heterogeneity: Mapping Whole Exome Sequencing to In Vivo Imaging with Stereotactic Localization and Deep Learning.

Intratumoral heterogeneity (ITH) complicates the diagnosis and treatment of glioma, partly due to the diverse metabolic profiles driven by underlying genomic alterations. While multiparametric imaging enhances the characterization of ITH by capturing both spatial and functional variations, it falls short in directly assessing the metabolic activities that underpin these phenotypic differences. This gap stems from the challenge of integrating easily accessible, colocated pathology and detailed genomic data with metabolic insights. This study presents a multifaceted approach combining stereotactic biopsy with standard clinical open-craniotomy for sample collection, voxel-wise analysis of MR images, regression-based GAM, and whole-exome sequencing. This work aims to demonstrate the potential of machine learning algorithms to predict variations in cellular and molecular tumor characteristics. This retrospective study enrolled ten treatment-naïve patients with radiologically confirmed glioma. Each patient underwent a multiparametric MR scan (T1W, T1W-CE, T2W, T2W-FLAIR, DWI) prior to surgery. During standard craniotomy, at least 1 stereotactic biopsy was collected from each patient, with screenshots of the sample locations saved for spatial registration to pre-surgical MR data. Whole-exome sequencing was performed on flash-frozen tumor samples, prioritizing the signatures of five glioma-related genes: IDH1, TP53, EGFR, PIK3CA, and NF1. Regression was implemented with a GAM using a univariate shape function for each predictor. Standard receiver operating characteristic (ROC) analyses were used to evaluate detection, with AUC (area under curve) calculated for each gene target and MR contrast combination. Mean AUC for five gene targets and 31 MR contrast combinations was 0.75 ± 0.11; individual AUCs were as high as 0.96 for both IDH1 and TP53 with T2W-FLAIR and ADC, and 0.99 for EGFR with T2W and ADC. These results suggest the possibility of predicting exome-wide mutation events from noninvasive, in vivo imaging by combining stereotactic localization of glioma samples and a semi-parametric deep learning method. The genomic alterations identified, particularly in IDH1, TP53, EGFR, PIK3CA, and NF1, are known to play pivotal roles in metabolic pathways driving glioma heterogeneity. Our methodology, therefore, indirectly sheds light on the metabolic landscape of glioma through the lens of these critical genomic markers, suggesting a complex interplay between tumor genomics and metabolism. This approach holds potential for refining targeted therapy by better addressing the genomic heterogeneity of glioma tumors.

Targeting intra-tumoral heterogeneity of human brain tumors with in vivo imaging: A roadmap for imaging genomics from multiparametric MR signals.

Resistance of high grade tumors to treatment involves cancer stem cell features, deregulated cell division, acceleration of genomic errors, and emergence of cellular variants that rely upon diverse signaling pathways. This heterogeneous tumor landscape limits the utility of the focal sampling provided by invasive biopsy when designing strategies for targeted therapies. In this roadmap review paper, we propose and develop methods for enabling mapping of cellular and molecular features in vivo to inform and optimize cancer treatment strategies in the brain. This approach leverages (1) the spatial and temporal advantages of in vivo imaging compared with surgical biopsy, (2) the rapid expansion of meaningful anatomical and functional magnetic resonance signals, (3) widespread access to cellular and molecular information enabled by next-generation sequencing, and (4) the enhanced accuracy and computational efficiency of deep learning techniques. As multiple cellular variants may be present within volumes below the resolution of imaging, we describe a mapping process to decode micro- and even nano-scale properties from the macro-scale data by simultaneously utilizing complimentary multiparametric image signals acquired in routine clinical practice. We outline design protocols for future research efforts that marry revolutionary bioinformation technologies, growing access to increased computational capability, and powerful statistical classification techniques to guide rational treatment selection.

Statistical multiscale mapping of IDH1, MGMT, and microvascular proliferation in human brain tumors from multiparametric MR and spatially-registered core biopsy.

We propose a statistical multiscale mapping approach to identify microscopic and molecular heterogeneity across a tumor microenvironment using multiparametric MR (mp-MR). Twenty-nine patients underwent pre-surgical mp-MR followed by MR-guided stereotactic core biopsy. The locations of the biopsy cores were identified in the pre-surgical images using stereotactic bitmaps acquired during surgery. Feature matrices mapped the multiparametric voxel values in the vicinity of the biopsy cores to the pathologic outcome variables for each patient and logistic regression tested the individual and collective predictive power of the MR contrasts. A non-parametric weighted k-nearest neighbor classifier evaluated the feature matrices in a leave-one-out cross validation design across patients. Resulting class membership probabilities were converted to chi-square statistics to develop full-brain parametric maps, implementing Gaussian random field theory to estimate inter-voxel dependencies. Corrections for family-wise error rates were performed using Benjamini-Hochberg and random field theory, and the resulting accuracies were compared. The combination of all five image contrasts correlated with outcome (P < 10-4) for all four microscopic variables. The probabilistic mapping method using Benjamini-Hochberg generated statistically significant results (α ≤ 0.05) for three of the four dependent variables: (1) IDH1, (2) MGMT, and (3) microvascular proliferation, with an average classification accuracy of 0.984 ± 0.02 and an average classification sensitivity of 1.567% ± 0.967. The images corrected by random field theory demonstrated improved classification accuracy (0.989 ± 0.008) and classification sensitivity (5.967% ± 2.857) compared with Benjamini-Hochberg. Microscopic and molecular tumor properties can be assessed with statistical confidence across the brain from minimally-invasive, mp-MR.

Self-directed down-regulation of auditory cortex activity mediated by real-time fMRI neurofeedback augments attentional processes, resting cerebral perfusion, and auditory activation.

In this work, we investigated the use of real-time functional magnetic resonance imaging (fMRI) with neurofeedback training (NFT) to teach volitional down-regulation of the auditory cortex (AC) using directed attention strategies as there is a growing interest in the application of fMRI-NFT to treat neurologic disorders. Healthy participants were separated into two groups: the experimental group received real feedback regarding activity in the AC; the control group was supplied sham feedback yoked from a random participant in the experimental group and matched for fMRI-NFT experience. Each participant underwent five fMRI-NFT sessions. Each session contained 2 neurofeedback runs where participants completed alternating blocks of "rest" and "lower" conditions while viewing a continuously-updated bar representing AC activation and listening to continuous noise. Average AC deactivation was extracted from each closed-loop neuromodulation run and used to quantify the control over AC (AC control), which was found to significantly increase across training in the experimental group. Additionally, behavioral testing was completed outside of the MRI on sessions 1 and 5 consisting of a subjective questionnaire to assess attentional control and two quantitative tests of attention. No significant changes in behavior were observed; however, there was a significant correlation between changes in AC control and attentional control. Also, in a neural assessment before and after fMRI-NFT, AC activity in response to continuous noise stimulation was found to significantly decrease across training while changes in AC resting perfusion were found to be significantly greater in the experimental group. These results may be useful in formulating effective therapies outside of the MRI, specifically for chronic tinnitus which is often characterized by hyperactivity of the primary auditory cortex and altered attentional processes. Furthermore, the modulation of attention may be useful in developing therapies for other disorders such as chronic pain.

Volitional down-regulation of the primary auditory cortex via directed attention mediated by real-time fMRI neurofeedback.

The present work assessed the efficacy of training volitional down-regulation of the primary auditory cortex (A1) based on real-time functional magnetic resonance imaging neurofeedback (fMRI-NFT). A1 has been shown to be hyperactive in chronic tinnitus patients, and has been implicated as a potential source for the tinnitus percept. 27 healthy volunteers with normal hearing underwent 5 fMRI-NFT sessions: 18 received real neurofeedback and 9 sham neurofeedback. Each session was composed of a simple auditory fMRI followed by 2 runs of A1 fMRI-NFT. The auditory fMRI alternated periods of no auditory with periods of white noise stimulation at 90 dB. A1 activity, defined from a region using the activity during the preceding auditory run, was continuously updated during fMRI-NFT using a simple bar plot, and was accompanied by white noise (90 dB) stimulation for the duration of the scan. Each fMRI-NFT run alternated "relax" periods with "lower" periods. Subjects were instructed to watch the bar during the relax condition and actively reduce the bar by decreasing A1 activation during the lower condition. Average A1 de-activation, representative of the ability to volitionally down-regulate A1, was extracted from each fMRI-NFT run. A1 de-activation was found to increase significantly across training and to be higher in those receiving real neurofeedback. A1 de-activation in sessions 2 and 5 were found to be significantly greater than session 1 in only the group receiving real neurofeedback. The most successful subjects reportedly adopted mindfulness tasks associated with directed attention. For the first time, fMRI-NFT has been applied to teach volitional control of A1 de-activation magnitude over more than 1 session. These are important findings for therapeutic development as the magnitude of A1 activity is altered in tinnitus populations and it is unlikely a single fMRI-NFT session will reverse the effects of tinnitus.

A Protocol for the Administration of Real-Time fMRI Neurofeedback Training.

Neurologic disorders are characterized by abnormal cellular-, molecular-, and circuit-level functions in the brain. New methods to induce and control neuroplastic processes and correct abnormal function, or even shift functions from damaged tissue to physiologically healthy brain regions, hold the potential to dramatically improve overall health. Of the current neuroplastic interventions in development, neurofeedback training (NFT) from functional Magnetic Resonance Imaging (fMRI) has the advantages of being completely non-invasive, non-pharmacologic, and spatially localized to target brain regions, as well as having no known side effects. Furthermore, NFT techniques, initially developed using fMRI, can often be translated to exercises that can be performed outside of the scanner without the aid of medical professionals or sophisticated medical equipment. In fMRI NFT, the fMRI signal is measured from specific regions of the brain, processed, and presented to the participant in real-time. Through training, self-directed mental processing techniques, that regulate this signal and its underlying neurophysiologic correlates, are developed. FMRI NFT has been used to train volitional control over a wide range of brain regions with implications for several different cognitive, behavioral, and motor systems. Additionally, fMRI NFT has shown promise in a broad range of applications such as the treatment of neurologic disorders and the augmentation of baseline human performance. In this article, we present an fMRI NFT protocol developed at our institution for modulation of both healthy and abnormal brain function, as well as examples of using the method to target both cognitive and auditory regions of the brain.