Linda Bryant
What drew me to psychosis research
Psychosis is not limited to one diagnosis. It can occur in schizophrenia, but also in neurological conditions such as Parkinson's disease and behavioural variant frontotemporal dementia (bvFTD). I became interested in psychosis research through my clinical work in an NHS Early Intervention in Psychosis service. This experience highlighted the clinical complexity of psychosis and the substantial variation in its presentation between individuals. It also underscored how much remains to be understood about why psychosis emerges across different brain disorders and what underlies this individual variation.
This led me to ask whether psychosis is associated with shared brain features across diagnoses or whether it emerges through different pathways in different conditions. Answering this is difficult because brain imaging produces large and complex datasets, and the most informative patterns may be subtle and spread across many regions. Artificial intelligence (AI) offers a way to bring these patterns together and examine how they differ between individuals. This interest now forms the basis of my PhD at King's College London, where I use AI and brain imaging to study psychosis across psychiatric and neurodegenerative conditions, with the aim of supporting more personalised care.
My research at and since the SIRS Congress
At SIRS 2026, I presented research examining the brain features associated with psychosis in Parkinson's disease. Although Parkinson's is usually known for its movement symptoms, more than half of people may experience psychosis over the course of the illness, often as visual hallucinations or delusions. This is associated with faster cognitive decline, a greater risk of dementia, increased pressure on caregivers and earlier admission to residential care. Clinicians, however, still have limited tools for identifying who is most at risk.
We analysed structural MRI scans from 481 people with Parkinson's disease, including 135 who had experienced psychosis. Using machine learning, we found that patterns across visual, motor and deeper dopamine-related brain systems could distinguish those who had experienced psychosis better than chance. Visual-processing regions were particularly informative, which is especially relevant because visual hallucinations are a prominent feature of Parkinson's psychosis. This work is a first step towards moving from average group differences to information that may eventually be useful for an individual person.
My wider research also examines the boundary between schizophrenia and neurodegeneration. In a recent study, we found preliminary evidence that some people with schizophrenia express a brain pattern associated with lower dopamine production and higher iron levels in a deep brain region called the striatum. This pattern resembles changes seen in behavioural variant frontotemporal dementia (bvFTD), suggesting that a subgroup of people with schizophrenia may show neurodegeneration-like features. These findings may help us better understand the biological heterogeneity within schizophrenia.
Why SIRS matters
SIRS provides a valuable forum for exchanging ideas with researchers and clinicians who approach psychosis from different perspectives. Receiving an Early Career Award made the Congress especially rewarding, as it offered a fantastic opportunity to connect with other awardees, follow their research beyond the meeting and benefit from mentorship within the SIRS community.
Where I want to take the research next
In the future, I want to continue using AI to advance precision medicine across psychosis and neurodegenerative conditions. I plan to continue combining brain imaging with clinical information to better predict individual risk and treatment outcomes. The longer-term aim is to translate these tools into clinical workflows, so that information from each person can guide earlier and more precise decisions about monitoring and treatment. Ultimately, I hope this work helps move care away from one-size-fits-all approaches and towards support that is tailored to the individual.
