Purpose: To develop a method for statistically significant tumor delineation. Glioblastoma (GBM) remains resistant to current therapeutic strategies and is unequivocally associated with a dismal prognosis. Hyperpolarized 13C MRI (hpMRI) provides unique insights into tissue metabolism, enabling outlining of tumor boundaries. Statistical confidence of this outline may help guide surgical resection, inform therapeutic decisions, and evaluate treatment efficacy. Materials and Methods: The new method was applied to time-resolved hyperpolarized [1-13C]pyruvate data from a previous study in a rat glioma model as well as from a new clinical acquisition from a patient with brain cancer. MATLAB R2021a was used for the comprehensive extraction and analysis of metabolite profiles. Time-points corresponding to high pyruvate and lactate signal intensities were combined, smoothed using a two-dimensional (2D) moving average, and analyzed with a sliding window to identify regions statistically distinct from manually segmented normal tissue. The negative predictive value (NPV) and the positive predictive value (PPV) of the detection method were evaluated by comparison with cancer regions delineated using standard radiological criteria. Results: The new approach enabled glioblastoma delineation with the NPV of 99% and the PPV of 59% in small animals. While in the larger brain of the clinical patient, with bicarbonate taken into account, NPV and PPV obtained 95% and 92% respectively. Conclusions: Combining time-points of high pyruvate and lactate signal intensities increases the statistical power of two-dimensional testing. This promising technique requires further evaluation in a larger patient cohort.
Abdallah Salemdawood, Dirk Mayer, A. Eldirdiri et al.· Magnetic Resonance in Medici...· 0 citations
Background / Introduction: Biologically inspired spiking neural networks can model adaptive behavior, but learning multiple goals is difficult because synaptic updates for different targets can interfere. We tested whether multi-timescale plasticity and context-specific credit assignment could improve continual multi-goal learning in a spiking navigation system inspired by entorhinal-hippocampal circuitry. Methods: We developed a closed-loop spiking model containing grid-like, place-like, target-related, association, and motor-output populations. An agent navigated in a two-dimensional environment with randomized starting locations and learned through reward-modulated spike-timing dependent plasticity (STDP/RL) and a novel evidence-gated plasticity (EGP) framework. EGP accumulates candidate synaptic modifications, evaluates them using reward evidence, and consolidates only changes that improve performance. A target-context variant maintained separate proposal stores and reward evaluation for each target. Results: STDP/RL learned and retained a single-target navigation policy, but multi-target training produced substantial interference, including attraction to incorrect targets after learning. Across 10 connectivity seeds, target-context EGP achieved higher late-stage reward than global EGP, improved weakest-target performance, and increased the fraction of targets achieving positive reward. In a longer continual-learning simulation, reward increased for all targets, TEST-phase performance increasingly exceeded TRAIN-phase performance, and proposal magnitudes grew over learning. Dwell-time confusion analyses showed that target-context EGP reduced wrong-target attraction and improved target selectivity relative to multi-target STDP/RL. Conclusions: These results demonstrate that spiking navigation circuits can learn goal-directed behavior using local plasticity, but robust multi-goal learning benefits from context-specific evidence-based consolidation. Target-context EGP provides a biologically motivated mechanism for reducing interference during continual reinforcement learning in spiking neural networks.
Samuel A Neymotin, Hananel Hazan, Gozde Unal et al.· Research Square· 0 citations
