From a single MRI to a real-time EEG read, Tale Research builds the methods, the normative references, and the platforms that turn raw neural data into decisions clinicians can act on. Two research streams, both grounded in open-source components and traceable numbers.
Tale Research works across two complementary modalities: structural neuroimaging, where a single 3D T1 MRI becomes a quantitative volumetric read, and real-time EEG, where continuous neural signals become clinically meaningful markers session by session.
Both streams share one philosophy. Open methods over proprietary classifiers. Documented normative references over hand-tuned thresholds. Numbers that trace back to the voxels or the band power they came from. The platforms — NeuroSentinel and Vexis — are how the research reaches clinicians.
Two clinical streams and one cross-domain track. Each area is grounded in published methods, traceable numbers, and reference data we keep open whenever we can.
Volumetric biomarkers, normative references, and longitudinal change tracking — the research stack underneath NeuroSentinel. Built on FreeSurfer SynthSeg parcellation and an open 205-subject reference cohort.
Building on contrast-agnostic SynthSeg parcellation, we develop volumetric biomarkers that detect structural change earlier than visual radiological review. The current focus is a TBI structural signature — a single composite score that captures the regional volume-loss patterns characteristic of traumatic injury, scored against a 205-subject normative cohort.
Quantitative neuroimaging only works if the reference matters. We curate and publish open normative cohorts with documented inclusion criteria, demographic balance, and acquisition diversity — and design percentile inference that holds up under real-world clinical scan quality.
A single-timepoint volume is useful; the same brain measured twice is more useful. We develop methods for hemispheric asymmetry indices (left vs right structural divergence) and within-subject longitudinal volumetric tracking — both critical for follow-up imaging in neurosurgery, neurology, and TBI rehabilitation.
Real-time spectral methods, ICA-based artifact handling, and the validation work that turns raw EEG into markers a clinician can read second by second — the research stack underneath Vexis and several upstream applications.
Vexis turns continuous EEG into a single, clinically meaningful marker of parasympathetic dominance — the Vexis Relaxation Index — designed to make the effect of a manual-therapy session visible second by second. The research underneath develops band-decomposition methods, clinic-grade artifact handling, and validation against autonomic-tone measures.
A multi-modal EEG analysis pipeline that combines spectral methods with non-linear dynamics to surface pre-ictal biomarkers before clinical manifestation. The early-warning window gives patients and caregivers preparation time and reduces injury risk in epilepsy management.
EEG-based methods to objectively measure cognitive effort in real time, with applications in safety-critical environments where subjective self-report falls short — aviation, healthcare, training simulators, and human-machine teaming.
Independent Component Analysis is fundamental to every downstream EEG application. Our methodology isolates neural signals from artifacts (eye movements, muscle activity, line noise) so the markers we surface — relaxation indices, workload metrics, pre-ictal biomarkers — sit on clean component decompositions.
Applied work that takes the same EEG and signal-processing methods into non-clinical settings. Independent of the clinical platforms, but built on the same toolchain.
Bridging neuroscience and behavioral economics to decode the neural mechanisms underlying financial decision-making. By monitoring brain activity during simulated trading scenarios, we identify neural markers that predict risk assessment, loss aversion, and decision confidence.