Lab News
Eight abstracts from members of the Medical Imaging Lab have been accepted for presentation at fNIRS 2026 in Macau, highlighting the breadth of research taking place across the lab, from diffuse optical imaging and brain connectivity to hyperscanning, computational modelling and multi frequency fNIRS.
The Medical Imaging Lab at the University of Birmingham is proud to announce that eight abstracts involving members of the lab have been accepted for presentation at fNIRS 2026 in Macau, China.
The Society for functional near infrared spectroscopy describes fNIRS 2026 as the IX Biennial Meeting of the Society and one of the major international meetings dedicated to the field. The conference will bring together researchers working across neuroscience, biomedical optics, instrumentation, modelling and data analysis. Abstract submissions were subject to peer review as part of the conference selection process.
The range of accepted work from MI Lab reflects the breadth of research being carried out across the group and provides an exciting opportunity for our researchers to share their work with the wider international fNIRS community.
Improving diffuse correlation spectroscopy data
Ben Fry will present work on improving the quality of diffuse correlation spectroscopy data.
His study focuses on the problem of noise in Diffuse Correlation Spectroscopy, or DCS, which can be used to continuously and non invasively estimate a blood flow index. One of the challenges facing DCS is that its signal to noise ratio can become poorer as source detector separation increases and photon counts decrease.
Ben has developed a denoising algorithm designed to work across a range of continuous wave DCS device configurations. Their results demonstrate improved signal quality across several datasets, including a 34 percent increase in signal to noise ratio in resting state in vivo data.
The work has the potential to provide a practical off the shelf approach for improving the quality of DCS measurements and highlights the important role of signal processing in making diffuse optical measurements more robust.
Reproducibility of fNIRS during longitudinal attention tasks
Maisie Sanderson will present research investigating the reproducibility of fNIRS measurements during a longitudinal attention task.
The study followed five participants across five milestone sessions over a total period of 21 sessions. K means clustering was used to group fNIRS channels into regions of interest, with reproducibility assessed using measures of spatial overlap and size similarity.
The results showed moderate to high intra subject stability, with mean spatial overlap of 0.42 and mean size similarity of 0.47. The strongest consistency was observed in the left and right medial prefrontal cortex.
These findings suggest that the 0 back task can provide a reliable functional baseline for longitudinal studies investigating neuroplasticity using fNIRS. The work is particularly relevant to understanding how stable fNIRS measurements remain when participants are assessed repeatedly over an extended period.
Understanding the neurodivergent brain
Jennifer Button will present “Shining a light on the Neurodivergent Brain: Fronto parietal functional connectivity during Attentional Control: An fNIRS Study”, as an oral talk, representing The Centre for Human Brain Health and the Schools of Psychology and Computer Science.
The research investigates functional connectivity within the fronto parietal control network during attentional control using high density fNIRS. Functional connectivity provides a way of examining how activity in different brain regions changes in relation to one another, allowing researchers to investigate communication across distributed brain networks.
In the study, 23 adults completed a frequency manipulated Go No Go task designed to investigate response inhibition and sustained attention. Following signal quality assessment, 13 datasets were retained for analysis. The results showed increased inter regional connectivity between frontal and parietal regions during the task compared with resting conditions, suggesting that high density fNIRS can capture distributed fronto parietal functional connectivity during attentional control.
The work represents an important step towards understanding whether fNIRS can provide information about distributed brain networks traditionally studied using functional MRI.
Comparing hyperscanning approaches during mother infant interactions
M. Juliana Gutierrez Camperos will present research examining different approaches to analysing hyperscanning fNIRS data during mother infant interactions, in collaboration with researchers from Mackenzie Presbyterian University and Dr Rickson Mesquita.
Hyperscanning allows brain activity to be measured from two people at the same time, making it possible to investigate shared or coordinated neural responses during social interaction. In this study, Juliana compared intersubject correlation and general linear model approaches for analysing neural synchrony between mothers and their six month old infants.
The project aims to address an important methodological challenge in hyperscanning research. Different processing pipelines and analytical approaches can produce different results, making reproducibility difficult between studies.
By comparing complementary approaches and investigating their sensitivity to shared neural dynamics, the work aims to contribute towards more consistent and reproducible analysis of mother infant hyperscanning data. Preliminary results from the ongoing project will be presented at the conference.
Improving fNIRS image reconstruction with an end to end framework
Biao Zheng will present research investigating an end to end framework for improving fNIRS haemodynamic response recovery.
Conventional fNIRS general linear model analysis typically estimates haemodynamic response coefficients independently at individual channels. The work presented at fNIRS 2026 instead investigates a spatially informed approach that considers the spatial, temporal and spectral relationships within the optimisation process.
Using a three dimensional layered model representing the adult head, Biao simulated a localised cortical activation and generated continuous wave measurements using NIRFASTerFF. The resulting haemodynamic response images were reconstructed using both the end to end approach and conventional methods.
The results showed improvements in contrast and localisation, with the greatest improvement observed for HbO compared with HbR. The findings suggest that incorporating spatial information directly into the optimisation process can improve the accuracy and depth sensitivity of haemodynamic response recovery.
Exploring causal discovery for effective brain connectivity
Yaru Zhou will present research on causal discovery methods for estimating effective brain connectivity using fNIRS.
While functional connectivity examines statistical relationships between brain signals, effective connectivity goes a step further by investigating causal influences between neural systems. This can provide insight into the direction of information flow within the brain.
The study evaluates four causal discovery methods, including Granger causality, vector autoregressive based Granger causality, Greedy Equivalence Search and the Peter Clark algorithm. These methods were tested using simulated neural time series across different network sizes, transmission delays and levels of noise.
The results showed that the time lag based approaches generally outperformed the distribution based approaches across the simulated conditions. The findings also highlighted the importance of temporal modelling when estimating effective connectivity from time series data.
Understanding sleep haemodynamics across timescales
Robert Ward will present research investigating how brain blood flow and oxygenation change throughout sleep using functional near infrared spectroscopy.
Rather than treating sleep as a sequence of separate stages, the study explores whether haemodynamic activity can be understood as a continuous process evolving through a low dimensional state space. The analysis examines haemodynamic features, the underlying manifold structure and the rate at which activity changes through feature space, while also considering how the choice of temporal window affects these representations.
Across 17 participants, the haemodynamic features themselves remained relatively stable across different window lengths, while the geometry and dynamics of the data changed substantially. Longer windows revealed more coherent and slowly evolving patterns, while deeper NREM sleep was associated with slower haemodynamic dynamics than lighter sleep, REM sleep and wakefulness.
These findings highlight the importance of temporal scale when analysing sleep haemodynamics and suggest that fNIRS can provide a useful perspective on sleep as a continuous physiological process operating across multiple timescales.
Depth resolved quantitative multi frequency fNIRS
Manraaj Singh will present work on depth resolved quantitative multi frequency fNIRS.
Conventional continuous wave fNIRS cannot independently recover tissue absorption and reduced scattering, creating ambiguity when estimating haemoglobin changes. Frequency domain fNIRS addresses this limitation by measuring both the amplitude and phase of the optical signal, allowing absolute quantification of these optical properties.
Manraaj’s work investigates the potential of using multiple modulation frequencies to provide depth resolved sensitivity. Higher modulation frequencies attenuate more rapidly and therefore provide greater sensitivity to shallower tissue, while lower frequencies can penetrate more deeply.
Through simulation, phantom and in vivo studies, the research aims to establish how modulation frequency relates to tissue depth sensitivity and whether appropriate frequency combinations can improve the recovery of cortical haemodynamic responses. The longer term goal is to move fNIRS towards more quantitative and subject specific measurements of cortical haemodynamics.
Recognition for Jennifer Button
Alongside having her research accepted for presentation at fNIRS 2026, Jennifer Button has also been awarded the SfNIRS Diversity, Equity and Inclusion Travel Award for the conference.
The DEI Travel Awards are designed to support and recognise researchers from groups that have historically been under represented in scientific work related to fNIRS, while also supporting research with relevance to culturally diverse and under represented populations. The award provides support for researchers to participate in the international fNIRS community.
Jennifer has also been awarded the LES College Travel Award for her attendance at the conference. Together, these awards are a fantastic recognition of her work and provide valuable support as she travels to Macau to present her research.
A proud moment for MI Lab
Having eight accepted abstracts across such a diverse range of topics is a fantastic achievement for MI Lab and reflects the breadth of research being undertaken across the laboratory.
From improving the quality of diffuse optical measurements and developing new approaches to fNIRS reconstruction, to studying functional and effective connectivity, longitudinal reproducibility and brain to brain interactions, the accepted work represents a broad cross section of the questions being addressed by researchers within the lab.
We are extremely proud of everyone who has contributed to these studies and look forward to seeing Ben, Maisie, Jennifer, Juliana, Biao, Yaru, Manraaj and Robert representing the Medical Imaging Lab at fNIRS 2026 in Macau.
The conference will take place in October 2026, with the main conference running from 16 to 19 October and the Educational Workshop taking place on 15 and 16 October.
