WIP & Methods

Hao-Lun Fu, Jae-Chang Kim, Amy Wong

Wednesday, 28 January 2026, 12pm to 1pm

Hybrid via Teams or in-person in the Cowey rooms, FMRIB Annexe

Join via Teams

WIN Wednesday Works In Progress Identifying and Validating Biomarkers of Mild Traumatic Brain Injury (mTBI) Using Low-Density EEG

Presented by Hao-Lun Fu

Abstract: My DPhil project aims to assess the signal fidelity and translational potential of a portable EEG device for capturing brain activity relevant to mild traumatic brain injury (mTBI). As a critical first step, we aim to validate low-density EEG measures by comparing them to gold-standard modalities (high-density EEG and MEG) in healthy individuals.

MEG will be used as a high-resolution benchmark, as it is widely considered the gold standard for detecting neural signals disrupted in mTBI (e.g., frontal theta/delta activity and cognitive task-related responses). It will also allow evaluation of signal quality across systems. High-density EEG serves as an intermediate comparison between MEG and portable EEG. Demonstrating strong correspondence across devices in healthy individuals is essential to justify future deployment of portable EEG in post-injury or pitch-side assessments.

WIN Wednesday Works In Progress

How humans solve computationally complex resource procurement problems: the example of food choice

Presented by Jae-Chang Kim

Abstract: Choosing the best foods is not always trivial and can easily go wrong. While poor food choice is often attributed to carelessness, poor planning, medical reasons, or the attractiveness of unhealthy foods, the underlying mechanisms may be much deeper. We propose that economic food choice constitutes a complex combinatorial problem: individual decision-makers need to choose the best combination of foods that maximises benefits (utility) by optimising energy intake (calories) for fuelling physiological processes, nutrient balance (fat, carbohydrate, protein) for assuring physical health, and staying within a limited budget. Solving this optimisation problem requires experience and cognitive ability, both of which can be limited and fail.

We approach the confusing complexity of food choice mathematically, by conceptualising utility maximization as a ‘knapsack’ (limited-capacity food bag) problem. The two crucial parameters, energy and nutrient balance, have a cost that can quickly exceed ones’ income; this constraint is represented by knapsack’s limited capacity. To solve the problem, even the most efficient algorithms require decision-makers to consider exponentially increasing numbers of item (food) combinations. This computational complexity is formalised by the Sahni-k index—the minimal number of item combinations that need to be chosen before the greedy algorithm (which adds individual items in decreasing order of value) can fill the knapsack optimally. The knapsack problem represents tasks consumers face in the supermarket (monthly allowance being the limited capacity). Suboptimal problem-solving would cause malnourishment or obesity.

WIN Wednesday Methods Series Detecting Human Neural Replay with Temporally Delayed Linear Modelling (TDLM)

Presented by Amy Wong

Abstract: Neural replay, the spontaneous and rapid reactivation of past neural sequences, is thought to play a key role in memory and planning. Detecting neural replay noninvasively in humans remains challenging. We present a Python-based implementation of Temporally Delayed Linear Modelling (TDLM), a method for detecting neural replay in human MEG data. We outline key analysis steps, replication efforts of Liu et al. (Cell, 2019), and discuss challenges such as parameter sensitivity and oscillatory noise, to promote transparent and reproducible methods for studying human neural replay. imaging method to date.

How do I...?

Watch past Seminars?

Our past seminars, together with talk recordings (available to those with an Oxford University SSO) are listed here.

Watch other recorded seminars and courses?

Other WIN Wednesdays, courses and talks may also be recorded and available on the University Replay Pages (SSO required).

Find out more?

Find out how to book and prepare a WIP, advertise your OxCIN event on these pages, recommend content, subscribe to the events calendar, and more in our main 'How do I...?' pages.