Ai Eeg Analytics

Electrode Technology Development

Doctoral work develops sensing interfaces converting brain electrical activity into recordable signals. Electrode performance sets the quality ceiling for every subsequent analytical step.

Dry Electrode Systems

Research develops electrodes requiring no conductive gel for reliable contact. Gel free operation is essential for practical everyday and home recording.

Wearable Headset Design

Doctoral study designs comfortable head worn systems for extended recording. Comfort and stability determine whether long recordings are actually tolerated.

Ear Based Recording Systems

Research examines recording from within and around the ear canal. Ear placement offers discreet recording acceptable for continuous daily wear.

Textile And Flexible Electrodes

Doctoral work develops conformable electrodes integrated into fabric and flexible substrates. Conformable materials improve contact stability during natural movement.

High Density Array Systems

Research examines recording from very large numbers of closely spaced contacts. Dense sampling resolves spatial detail that sparse arrays cannot capture.

Intracranial Recording Methods

Doctoral study examines signals recorded directly from within the skull. Intracranial recording offers resolution that scalp measurement cannot approach.

Stereotactic Depth Recording Analysis

Research analyses signals from electrodes placed along trajectories into deep structures. Depth recording accesses regions entirely invisible from the scalp.

Electrocorticography Signal Analysis

Doctoral work analyses activity recorded directly from the cortical surface. Surface recording combines high resolution with broad spatial coverage.

Subscalp Implant Systems

Research examines minimally invasive devices placed beneath the scalp. These devices enable continuous monitoring across months without external hardware.

Amplifier And Front End Design

Doctoral study designs the electronics amplifying extremely small neural voltages. Front end noise performance limits what activity can be detected at all.

Sampling And Digitisation Methods

Research examines conversion of continuous signals into digital records. Sampling choices determine which frequency content is preserved for analysis.

Reference Montage Selection

Doctoral work examines how reference choice shapes the recorded signal. Reference selection materially changes apparent activity and connectivity results.

Impedance Monitoring Methods

Research tracks electrode contact quality continuously during recording. Contact quality degrades over time and silently corrupts recorded data.

Motion Tolerant Recording

Doctoral study develops recording that remains usable during subject movement. Movement tolerance is essential for recording outside laboratory conditions.

Ambulatory Recording Systems

Research examines portable systems recording during ordinary daily activity. Ambulatory recording captures events that clinic sessions never observe.

Home Based Monitoring Technology

Doctoral work develops systems patients can apply and operate themselves. Home monitoring extends assessment beyond scarce specialist facilities.

Long Term Continuous Monitoring

Research examines recording sustained across days, weeks or longer periods. Extended recording captures rare events that brief sessions inevitably miss.

Low Power Recording Electronics

Doctoral study minimises energy consumption in continuously recording devices. Power consumption determines how long a wearable device can operate.

Wireless Transmission Of Neural Data

Research examines reliable wireless communication of recorded neural signals. Wireless operation removes cables that restrict movement and comfort.

On Device Signal Processing

Doctoral work performs analysis within the recording device itself. Local processing reduces transmission volume and protects sensitive information.

Edge Computing In Neural Devices

Research embeds learned inference within constrained neural recording hardware. Embedded inference enables immediate response without any network connection at all.

Simultaneous Multimodal Acquisition

Doctoral study records neural activity alongside other physiological signals. Combined recording resolves ambiguity that single modality data leaves.

Concurrent Imaging And Recording

Research combines electrical recording with simultaneous brain imaging. Combined methods link fast electrical activity to spatial anatomical detail.

Recording During Stimulation

Doctoral work records neural activity while stimulation is being applied. Stimulation artefacts overwhelm signals and require dedicated handling.

Neonatal Recording Systems

Research develops recording suited to newborn infants and their fragile skin. Newborn recording demands entirely distinct hardware and interpretation.

Paediatric Recording Adaptation

Doctoral study adapts recording methods to children and their tolerance limits. Developmental change means adult reference values do not apply.

Recording In Intensive Care

Research examines continuous monitoring within critical care environments. Intensive care units are electrically noisy and operationally demanding settings.

Field And Remote Recording

Doctoral work examines recording outside clinical and laboratory facilities. Field capability extends assessment to populations with no specialist access.

Consumer Device Validation

Research evaluates whether inexpensive consumer systems produce usable signals. Validation is essential given widespread claims made for these devices.

Automated Artefact Detection

Doctoral study automates recognition of non neural contamination within recordings. Artefact handling consumes the majority of expert review time.

Ocular Artefact Removal

Research removes contamination arising from eye movement and blinking. Eye related contamination is pervasive and overlaps neural frequency content.

Muscle Artefact Suppression

Doctoral work suppresses contamination from scalp and facial muscle activity. Muscle contamination is broadband and particularly difficult to separate.

Cardiac Interference Removal

Research removes heartbeat related contamination from neural recordings. Cardiac contamination is rhythmic and can be mistaken for neural oscillation.

Line Noise Suppression Methods

Doctoral study removes interference from electrical power infrastructure. Power line interference is ubiquitous in clinical and home environments.

Movement Artefact Correction

Research corrects contamination produced by head and body movement. Movement contamination is the principal obstacle to recording during activity.

Electrode Contact Failure Detection

Doctoral work detects when an electrode has lost adequate contact. Undetected contact failure produces data that appears valid but is meaningless.

Bad Channel Identification

Research identifies channels carrying unusable signal within a recording. Automated identification removes a slow and inconsistent manual task.

Signal Quality Assessment Metrics

Doctoral study develops objective measures of recording usability. Quality measures determine which segments should enter downstream analysis.

Independent Component Analysis Methods

Research separates recorded mixtures into statistically independent underlying sources. Component separation remains the established route to removing contamination.

Blind Source Separation Techniques

Doctoral work recovers underlying sources without prior knowledge of mixing. Source separation addresses the fundamental mixing that scalp recording imposes.

Adaptive Filtering Approaches

Research develops filters adjusting continuously to changing signal conditions. Adaptive filters handle interference whose characteristics shift over time.

Wavelet Based Denoising

Doctoral study applies multiresolution decomposition to separate signal from noise. Wavelet methods suit signals whose properties change rapidly over time.

Empirical Mode Decomposition

Research decomposes signals into intrinsic oscillatory components adaptively. Data driven decomposition avoids assuming fixed frequency bands.

Baseline Correction Methods

Doctoral work removes slow signal drift and establishes stable reference levels. Baseline handling substantially influences measured response amplitudes.

Interpolation Of Missing Channels

Research reconstructs plausible signal for channels that failed during a recording. Reconstruction preserves spatial analysis when individual contacts are unusable.

Data Segmentation Strategies

Doctoral study divides long continuous recordings into analysable segments. Segmentation choices silently influence every statistic computed from the data afterwards.

Epoch Selection Automation

Research automates the choice of which recording periods should be analysed. Automated selection removes subjective and inconsistent manual decisions from the process.

Standardised Preprocessing Pipelines

Doctoral work develops consistent processing sequences applied across studies. Standardisation is required for results to be comparable between groups.

Pipeline Choice Sensitivity Analysis

Research examines how processing decisions influence eventual conclusions. Reasonable processing choices can produce substantially different findings.

Reproducible Preprocessing Frameworks

Doctoral study builds systems recording every processing step that was applied. Complete processing records are a prerequisite for independent replication.

Automated Quality Control Systems

Research automates checking of recordings before analysis proceeds. Automated checking catches problems that routine review reliably misses.

Deep Learning For Denoising

Doctoral work applies learned models to separating signal from contamination. Learned denoising must never invent structure that was not present.

Learned Artefact Rejection

Research trains models to recognise contamination as expert reviewers would. Learned rejection scales expert judgement across very large datasets.

Real Time Preprocessing Methods

Doctoral study develops processing operating within strict latency limits. Real time capability is mandatory for interface and monitoring applications.

Preprocessing For Constrained Devices

Research adapts processing methods to limited embedded computing resources. Wearable devices cannot support conventional processing pipelines.

Harmonisation Across Recording Sites

Doctoral work reconciles systematic differences between recording centres. Site differences frequently exceed the biological effects being studied.

Cross Device Signal Alignment

Research aligns recordings made with differing hardware and montages. Alignment permits combining data across equipment generations and vendors.

Data Standardisation Formats

Doctoral study develops common formats for storing and exchanging recordings. Format fragmentation obstructs data sharing across the whole field.

Metadata Standards For Recordings

Research develops structured description of recording conditions and context. Incomplete metadata renders otherwise valuable recordings unusable.

Spectral Power Analysis Methods

Doctoral work quantifies energy distribution across signal frequencies. Spectral measures remain the most widely used descriptors in the field.

Time Frequency Decomposition

Research characterises how signal frequency content changes across time. Neural activity is fundamentally nonstationary and demands time resolved description.

Nonstationary Signal Analysis

Doctoral study develops methods for signals whose statistics change continuously. Standard methods assume stationarity that neural signals never satisfy.

Oscillatory Rhythm Characterisation

Research characterises rhythmic activity and the functions associated with it. Rhythms remain central to how the field interprets and reports brain state.

Cross Frequency Coupling Analysis

Doctoral work examines interactions between activity at different frequencies. Coupling is proposed as a mechanism coordinating distributed processing.

Phase Amplitude Coupling Methods

Research quantifies how slow rhythm phase modulates faster activity. Measurement of this relationship is technically contested and easily biased.

Aperiodic Component Separation

Doctoral study separates rhythmic activity from broadband background structure. Confusing the two has produced substantial misinterpretation in the literature.

Microstate Analysis Methods

Research examines brief quasi stable spatial patterns of scalp activity. Microstates offer a discrete description of continuously varying activity.

Complexity And Entropy Measures

Doctoral work quantifies signal irregularity and its information content. Complexity measures track consciousness level and general neurological state closely.

Fractal And Scaling Analysis

Research examines self similar structure across temporal scales. Scaling properties characterise the organisation of ongoing brain activity.

Nonlinear Dynamics Approaches

Doctoral study applies dynamical systems theory to recorded brain activity. Nonlinear methods capture structure that linear analysis entirely misses.

Criticality Analysis In Neural Signals

Research examines whether brain activity operates near a critical regime. Criticality is proposed to explain the balance of stability and flexibility.

Connectivity Estimation Methods

Doctoral work estimates statistical relationships between recording locations. Connectivity estimates are highly sensitive to methodological choices.

Functional Connectivity Analysis

Research examines statistical dependence between activity at separate sites. Functional relationships describe coordination without implying causal direction.

Effective Connectivity Modelling

Doctoral study infers directed influence between brain regions. Directional inference from scalp signals remains methodologically demanding.

Graph Theoretic Network Analysis

Research characterises brain networks using formal graph measures. Network measures summarise complex connectivity into interpretable quantities.

Dynamic Network Analysis

Doctoral work examines how network organisation changes moment to moment. Static network descriptions conceal rapid reconfiguration during behaviour.

Source Localisation Methods

Research estimates where within the brain recorded activity originated. Localisation converts surface measurement into anatomically meaningful findings.

Inverse Problem Regularisation

Doctoral study addresses the fundamentally underdetermined localisation problem. Regularisation assumptions strongly determine the solutions obtained.

Head Model Construction

Research builds electrical models of the head for localisation computation. Head model accuracy directly limits achievable localisation precision.

Individual Anatomy Integration

Doctoral work incorporates each subject own anatomy into analysis. Individual anatomy substantially improves localisation over template approaches.

Beamforming Techniques

Research develops spatial filters isolating activity from chosen locations. Beamforming suppresses interference arising from other brain regions.

Spatial Filtering Methods

Doctoral study combines channels to emphasise particular activity patterns. Spatial filtering substantially improves signal to noise for weak responses.

Common Spatial Pattern Approaches

Research derives spatial filters maximising discrimination between mental states. These methods underpin a great deal of practical interface decoding work.

Riemannian Geometry Methods

Doctoral work analyses signals within the geometry of covariance matrices. Geometric methods have proved unusually robust across subjects and sessions.

Covariance Based Representation

Research represents recordings through relationships between channels. Covariance representations discard timing but capture spatial structure robustly.

Self Supervised Representation Learning

Doctoral study learns useful representations from entirely unlabelled recordings. Labelled neural data is extremely scarce while raw recordings are abundant.

Contrastive Learning For Neural Signals

Research learns representations by distinguishing related from unrelated segments. Contrastive objectives exploit structure without requiring expert labels.

Foundation Models For Neural Data

Doctoral work develops large pretrained models spanning many recording datasets. Pretrained models bring capability to tasks with very few labels.

Transformer Architectures For Signals

Research adapts attention based architectures to long neural recordings. Attention mechanisms capture dependencies across extended time spans.

Graph Neural Networks For Electrodes

Doctoral study represents electrode arrangements as graphs for learning. Graph structure encodes the spatial relationships between recording sites.

Recurrent Models For Temporal Structure

Research applies sequential models to the temporal evolution of recorded signals. Temporal modelling captures dynamics that summary static features entirely discard.

Spiking Network Models For Signals

Doctoral work applies event driven computational models to neural signal processing. Spiking models suit very low power hardware performing continuous monitoring.

Interpretable Feature Design

Research develops measures clinicians and scientists can meaningfully interpret. Interpretable features are essential where findings must be explained.

Dimensionality Reduction Methods

Doctoral study compresses high dimensional recordings into compact descriptions. Reduction reveals structure obscured by very large channel counts.

Seizure Detection Algorithms

Research develops automated recognition of seizure activity within recordings. Automated detection is essential given the volume of continuous monitoring.

Seizure Prediction Methods

Doctoral work examines whether seizures can be anticipated before onset. Reliable warning would transform daily life for many affected people.

Interictal Discharge Detection

Research detects brief abnormal events occurring between seizures. These events support diagnosis and indicate where abnormality originates.

Epileptogenic Zone Localisation

Doctoral study identifies the brain region from which seizure activity originates. Accurate localisation determines whether surgical treatment can succeed at all.

Surgical Planning In Epilepsy

Research supports planning of resective and disconnection procedures. Planning quality determines both seizure freedom and preserved function.

Treatment Response Prediction In Epilepsy

Doctoral work predicts which patients will respond to particular medications. Prediction avoids prolonged sequential trials of ineffective treatment.

Status Epilepticus Monitoring

Research monitors prolonged seizure activity requiring emergency management. Continuous monitoring guides treatment in a time critical condition.

Neonatal Seizure Detection

Doctoral study detects seizures in newborn infants where signs are subtle. Newborn seizures are frequently invisible clinically and go untreated.

Absence Seizure Classification

Research classifies brief generalised events that are common in childhood epilepsy. These events occur frequently and are easily mistaken for ordinary inattention.

Epilepsy Syndrome Classification

Doctoral work classifies epilepsy types from recording characteristics. Syndrome classification determines treatment choice and expected course.

Responsive Neurostimulation Analytics

Research analyses data from implanted devices delivering stimulation on detection. These devices generate continuous long term recordings from within the brain.

Nerve Stimulation Response Monitoring

Doctoral study examines neural changes accompanying peripheral nerve stimulation therapy. Monitoring supports individualised setting of stimulation parameters.

Sleep Stage Classification

Research automates scoring of sleep stages across whole night recordings. Manual scoring is slow, costly and shows substantial observer variation.

Sleep Microstructure Analysis

Doctoral work examines brief events and transitions within sleep stages. Microstructure carries information that conventional stage scoring discards.

Sleep Disorder Diagnosis Support

Research supports recognition of disorders affecting sleep quality and structure. Sleep disorders are common, underdiagnosed and highly treatable.

Breathing Disturbance Detection In Sleep

Doctoral study detects disturbed breathing events from neural and related signals. Simplified detection could extend assessment beyond specialist sleep laboratories.

Circadian Rhythm Analysis

Research examines daily rhythms expressed within brain electrical activity. Rhythm disturbance accompanies many neurological and psychiatric conditions.

Insomnia Assessment Methods

Doctoral work develops objective assessment of disturbed sleep initiation and maintenance. Objective measures frequently diverge from what patients report experiencing.

Anaesthesia Depth Monitoring

Research monitors consciousness level during surgical anaesthesia. Monitoring aims to prevent both awareness during surgery and excessive dosing.

Intraoperative Neuromonitoring

Doctoral study monitors neural function continuously during surgical procedures. Monitoring warns of developing injury while it remains reversible.

Coma And Consciousness Assessment

Research assesses consciousness level in unresponsive patients. Assessment informs decisions of the greatest possible consequence for families.

Disorders Of Consciousness Research

Doctoral work examines prolonged states of severely reduced responsiveness. Behavioural assessment misclassifies a substantial proportion of these patients.

Covert Awareness Detection

Research detects preserved awareness in patients unable to respond behaviourally. Covert awareness has profound implications for care and communication.

Prognostication After Brain Injury

Doctoral study predicts neurological recovery following severe brain injury. These predictions influence continuation of life sustaining treatment.

Cardiac Arrest Outcome Prediction

Research predicts neurological recovery after resuscitation from arrest. Prediction must be exceptionally cautious given the decisions it informs.

Traumatic Brain Injury Monitoring

Doctoral work monitors evolving brain function after traumatic injury. Continuous monitoring detects secondary injury while intervention remains possible.

Stroke Detection And Monitoring

Research examines neural signatures accompanying cerebrovascular events. Early detection determines eligibility for time critical treatment.

Cerebral Ischaemia Detection

Doctoral study detects inadequate blood supply from changes in brain activity. Neural change precedes irreversible injury by a usable interval.

Delirium Detection Methods

Research develops objective recognition of acute confusional states in patients. Delirium is common in hospital settings and very frequently goes entirely unrecognised.

Encephalopathy Assessment

Doctoral work examines diffuse brain dysfunction arising from systemic illness. Recording patterns indicate both severity and likely underlying cause.

Dementia Biomarker Research

Research examines electrophysiological markers of neurodegenerative disease. Inexpensive markers could support screening at population scale.

Mild Cognitive Impairment Detection

Doctoral study detects early cognitive change preceding established dementia. Early detection matters as disease modifying treatments become available.

Parkinsonian Disorder Analysis

Research examines neural signatures associated with parkinsonian conditions. Signatures support both diagnosis and adjustment of stimulation therapy.

Movement Disorder Assessment

Doctoral work examines neural correlates of tremor and involuntary movement. Objective measurement supports assessment of treatment effectiveness.

Multiple Sclerosis Monitoring

Research examines electrophysiological measures of nerve conduction change. These measures detect functional change before structural imaging does.

Migraine And Headache Research

Doctoral study examines neural excitability changes associated with headache disorders. Excitability markers may permit anticipation of impending attacks.

Psychiatric Disorder Biomarkers

Research examines electrophysiological markers across psychiatric conditions. Objective markers would substantially strengthen diagnosis currently based on interview.

Depression Biomarker Research

Doctoral work examines neural markers of depressive illness and treatment response. Response prediction could shorten the lengthy search for effective treatment.

Schizophrenia Neural Signatures

Research examines electrophysiological abnormalities in psychotic disorders. Signatures inform both mechanism research and early identification efforts.

Attention Disorder Assessment

Doctoral study examines neural measures relating to attention regulation. Objective measures could supplement diagnosis currently based on report.

Autism Neural Marker Research

Research examines neural response differences associated with autistic development. Findings must be interpreted with care regarding difference and disorder.

Developmental Disorder Assessment

Doctoral work examines neural markers across childhood developmental conditions. Early identification enables support during the most responsive period.

Substance Effect Monitoring

Research examines how psychoactive substances affect recorded brain activity. Neural measures support both research and clinical assessment of effects.

Medication Effect Assessment

Doctoral study examines neural changes produced by prescribed medications. Many medicines affect recordings and complicate clinical interpretation.

Treatment Response Monitoring

Research tracks neural change accompanying therapeutic intervention. Objective tracking supports adjustment before clinical change becomes apparent.

Event Related Potential Analysis

Doctoral work examines averaged neural responses to specific stimuli or actions. These responses are the established route to studying cognitive processing.

Single Trial Response Estimation

Research recovers responses from individual events without averaging. Single trial analysis reveals variability that averaging deliberately removes.

Attention State Decoding

Doctoral study infers what a person is attending to from neural activity. Attention decoding supports both interfaces and cognitive research.

Working Memory Load Estimation

Research estimates how much information a person is actively maintaining. Load estimation supports adaptive interfaces and educational applications.

Cognitive Workload Monitoring

Doctoral work measures mental effort during demanding real world tasks. Workload measurement supports safety in operationally critical roles.

Fatigue And Drowsiness Detection

Research detects declining alertness from continuously recorded activity. Drowsiness detection has direct safety applications in transport settings.

Vigilance State Assessment

Doctoral study tracks sustained alertness across extended monitoring tasks. Vigilance declines predictably and measurably during prolonged watch duties.

Emotion Recognition From Signals

Research examines whether emotional state can be inferred from neural activity. Claims in this area frequently exceed the demonstrated evidence.

Affective State Modelling

Doctoral work models continuous dimensions of emotional experience. Dimensional models fit neural evidence better than discrete emotion categories.

Language Processing Studies

Research examines neural activity accompanying comprehension and production. Language studies connect neural measurement to a defining human capability.

Speech Decoding From Neural Signals

Doctoral study reconstructs attempted speech from recorded brain activity. Speech decoding could restore communication to people who cannot speak.

Imagined Speech Research

Research examines neural activity accompanying internally generated speech. Imagined speech offers a natural control modality for communication interfaces.

Motor Imagery Decoding

Doctoral work decodes imagined movement from recorded brain activity. Motor imagery remains the most widely used interface control strategy.

Movement Intention Detection

Research detects preparation for movement before it is executed. Early detection reduces the delay users experience when controlling devices.

Brain Computer Interface Design

Doctoral study designs systems translating neural activity into device control. Interface design determines usability far more than raw decoding accuracy.

Non Invasive Interface Systems

Research develops interfaces requiring no surgical implantation. Non invasive systems are accessible but face substantially lower signal quality.

Invasive Interface Systems

Doctoral work examines implanted interfaces offering high signal quality. Implantation carries surgical risk justified only by substantial benefit.

Assistive Communication Systems

Research develops communication for people with severe motor impairment. Restored communication transforms quality of life and personal autonomy.

Neuroprosthetic Control Methods

Doctoral study develops neural control of prosthetic and robotic limbs. Control fidelity determines whether prosthetics restore useful function.

Rehabilitation Interface Systems

Research uses neural interfaces to support recovery after neurological injury. Interfaces may promote recovery by reinforcing intended movement attempts.

Neurofeedback System Design

Doctoral work develops systems presenting neural activity back to users. Rigorous evaluation is needed given widespread and weakly supported claims.

Closed Loop Stimulation Control

Research develops stimulation responding automatically to recorded activity. Responsive stimulation delivers therapy only when it is actually needed.

Transcranial Stimulation Analytics

Doctoral study examines neural changes produced by non invasive stimulation. Response varies widely between individuals and requires personalisation.

Deep Brain Stimulation Signal Analysis

Research analyses signals recorded from implanted stimulation electrodes. Recorded activity guides individualised adjustment of stimulation settings.

Adaptive Stimulation Parameter Tuning

Doctoral work automates adjustment of stimulation according to measured state. Automated tuning removes lengthy manual programming appointments.

Calibration Reduction Methods

Research reduces the setup time required before an interface becomes usable. Lengthy calibration is a leading reason users abandon these systems.

Cross Session Transfer Learning

Doctoral study transfers decoding models between separate recording sessions. Session differences otherwise force retraining before every use.

Cross Subject Generalisation

Research develops models working across individuals without any personal training. Generalisation would allow a new user to begin immediately without calibration.

User Adaptation And Co Learning

Doctoral work examines user and system adapting to one another over time. Mutual adaptation frequently outperforms optimising either side alone.

Interface Usability Research

Research examines whether interfaces are practically usable in daily life. Laboratory performance rarely survives contact with everyday conditions.

Long Term Interface Reliability

Doctoral study examines whether interface performance persists across months. Signal quality and decoding stability both degrade over extended use.

Driver State Monitoring

Research examines neural monitoring of alertness during vehicle operation. Detection of impaired alertness has clear road safety applications.

Operator Monitoring In Safety Roles

Doctoral work examines monitoring of personnel in safety critical positions. Monitoring raises significant consent and workplace surveillance concerns.

Educational Neurotechnology Research

Research examines neural monitoring applied within learning settings. Evidence for educational benefit remains weak and requires careful scrutiny.

Sport And Performance Monitoring

Doctoral study examines neural measurement in athletic and performance contexts. Applications include concussion assessment and readiness evaluation.

Benchmark Datasets For Neural Signals

Research constructs shared datasets enabling fair comparison of methods. Common benchmarks let claimed improvements be independently verified.

Data Sharing Infrastructure

Doctoral work develops systems for responsibly sharing recordings between groups. Sharing multiplies the value of expensive and difficult data collection.

Synthetic Signal Generation

Research generates realistic artificial recordings for method development. Synthetic data permits testing where real recordings cannot be shared.

Simulation Of Neural Recordings

Doctoral study simulates recordings with fully known underlying sources. Known ground truth permits evaluation impossible with real recordings.

Data Augmentation Techniques

Research expands limited training data through principled transformation. Augmentation addresses the persistent scarcity of labelled neural data.

Class Imbalance Handling

Doctoral work addresses datasets where events of interest are extremely rare. Rare event detection dominates clinical neural signal applications.

Label Quality And Expert Disagreement

Research examines disagreement between experts labelling the same recordings. Label disagreement places a hard ceiling on achievable model accuracy.

Uncertainty Estimation In Predictions

Doctoral study attaches calibrated confidence to automated neural assessments. Knowing when a system is unreliable is clinically as valuable as accuracy.

Explainability Of Neural Signal Models

Research develops explanation of automated conclusions from recordings. Clinicians will not act on findings whose basis they cannot examine.

Model Validation In Clinical Settings

Doctoral work evaluates automated analysis under real clinical conditions. Reported accuracy frequently fails to survive genuine deployment.

Prospective Clinical Evaluation

Research designs studies testing whether automated analysis improves outcomes. Predictive accuracy alone says nothing about clinical benefit.

Regulatory Approval Of Neural Software

Doctoral study examines evidence requirements for diagnostic neural software. Approval pathways determine which developments reach clinical practice.

Clinical Workflow Integration

Research examines how automated analysis fits existing neurophysiology practice. Tools disrupting established workflow are abandoned regardless of accuracy.

Clinician Trust In Automated Analysis

Doctoral work examines how appropriate reliance on automated review develops. Both uncritical acceptance and blanket rejection produce patient harm.

Reproducibility In Neural Signal Research

Research establishes practices allowing analyses to be independently repeated. Processing flexibility makes reproducibility especially difficult in this field.

Statistical Methods For Neural Data

Doctoral study develops inference suited to high dimensional dependent signals. Standard statistical assumptions fail badly for these data structures.

Multiple Comparison Correction

Research addresses statistical testing across many channels, frequencies and time points. Uncorrected testing across these dimensions generates a great many spurious findings.

Federated Learning Across Centres

Doctoral work trains shared models without centres exchanging patient recordings. Federation addresses both privacy rules and institutional sensitivity.

Privacy Of Neural Data

Research examines protection of recordings that may reveal sensitive information. Neural recordings carry health and potentially cognitive information.

Neural Data Governance And Rights

Doctoral study examines rights and control over recorded brain activity. Governance frameworks for this data remain substantially undeveloped.

Ethics Of Consumer Neurotechnology

Research examines consumer devices recording brain activity outside clinical oversight. These devices operate with minimal regulation and expansive marketing claims.

Algorithmic Fairness In Neural Analysis

Doctoral work measures whether methods perform equally across population groups. Training data in this field represents populations very unevenly.

Access And Equity In Neurotechnology

Research examines who can obtain neurological monitoring and who cannot. Specialist neurophysiology is unavailable across much of the world.

Health Economics Of Neural Monitoring

Doctoral study evaluates whether monitoring delivers value proportionate to cost. Economic evidence determines which services health systems can sustain.

Workforce And Training In Neurophysiology

Research examines skills required as neurophysiology becomes more computational. Specialist shortages constrain access more than technology maturity does.