Skip to content Skip to menu
This website uses cookies to help us understand the way visitors use our website. We can't identify you with them and we don't share the data with anyone else. If you click Reject we will set a single cookie to remember your preference. Find out more in UCL's privacy notices.

Understanding Brain Disorders

Clinical impact

What goes wrong in neurological and psychiatric disorders? How do we classify each disorder?

We investigate what defines neurological and psychiatric disorders at the level of brain structure and activity, with the goal of understanding the cause and consequence of recognisable symptoms.

Differences in brain function between patients, associated with the same symptoms, may be used to reclassify disorders in terms of pathophysiological mechanisms that could help inform the best treatment for an individual.


Understanding brain disorders involves three steps:

  • First, we need to understand behaviour, computations and brain structure/function in healthy individuals who differ from one another in a variety of typical ways.
  • Then, we investigate how patients differ from the normal range so that we can understand what has gone wrong.
  • Finally, we can examine differences between patients, to check whether they would be better grouped according to behavioural and brain measurements. We do this by comparing brain structure and activity in groups of patients who differ in the combination of their symptoms, their genetics, and other factors.


Uncovering abnormalities that are hidden from the human eye

For many conditions, it is not possible to see what is wrong using the naked eye. Structural abnormalities are often subtle and located in small parts of the brain. These can be hard to detect unless researchers already know precisely where to look. We can use our non-invasive methods to find differences between groups across the whole brain. We can then use this information to examine the regional abnormalities in more detail using a variety of specialised techniques, and the results can help inform other scientists using different experimental methods to study these conditions.

Some abnormalities are not always accompanied by detectable changes in brain structure. This is particularly the case for psychiatric disorders. Here we use brain activity measures to understand the brain circuitry and computations associated with typical mental states and tasks.

Understanding some mental states (e.g. fluctuating levels of regret) can also be challenging as they cannot be directly assessed from behavioural scores. Here we use computational approaches that track and integrate the multiple factors that influence behaviour and how these change over time.

Early diagnosis

We are working to develop ways of identifying brain disorders before the major symptoms of disease emerge. These will allow targeted interventions to be commenced earlier, before potentially irreversible damage to brain tissue has occurred.

Finding better treatments

By distinguishing between different causes of disorders, we open up new avenues of research to develop innovative treatments and interventions that can be tailored to individual patients.

Safe and accessible techniques

By demonstrating that our non-invasive neuroimaging techniques can be used to accurately diagnose, monitor and understand disorders, we can help replace some of the more invasive tests that are in use today.

Tailored help for patients

Across a wide range of brain conditions, developing ways to more accurately identify sub-types of disease will provide more relevant and specific diagnoses, prognoses and treatments targeted to individual patients.

Iteratively deepening our understanding

Accurate and useful classifications of patient groups can motivate further research to better understand the mechanisms of specific disorders and guide future treatments.




Psychiatric traits tend to arise in late adolescence, hinting at a feature in this stage of development that makes the brain uniquely vulnerable to psychiatric disorders. We found that lower myelin growth in specific brain regions during adolescence is linked to psychiatric traits such as compulsivity and impulsivity (Ziegler, Hauser et al., 2019). This finding brings us closer to a developmental understanding of psychiatric disorders, establishing a new direction in computational psychiatry (Hauser et al., 2019).


A characteristic of many psychiatric disorders is distorted self-awareness, more specifically known as metacognition. However, it is unclear whether self-awareness processes are affected independently or as a consequence of distorted decision-making processes. Stephen Fleming’s team has used computational modelling to disentangle the roles of these two processes in psychiatric symptoms. Self-awareness varied across different psychiatric disorders, but the changes in speed and accuracy of decision making did not – suggesting these two processes are affected separately (Rouault et al., 2018; Hauser et al., 2017). Understanding the processes behind the symptoms in detail could help identify which cognitive processes to target for treatment.


Decision-making processes are known to be impaired in psychiatric disorders. With computational modelling, we are determining the neural mechanisms underlying normal decision-making.   For example, Ray Dolan’s team found that the quality and quantity of an option are represented in two separate brain areas. Both are then integrated in a third region of the brain, providing a single value signal that the brain uses to quickly compare different options (De Berker et al., 2019).

Another decision-making process is finding when to stop gathering information and commit to a chosen option. In this process, Ray Dolan’s team showed that the neuromodulator noradrenaline modulates decision urgency to determine the cut-off point for information gathering (Hauser et al., 2018).  These advances in our understanding of neural mechanisms pave the way for understanding how impairments in decision-making processes contribute to the symptoms of psychiatric disorders.


One of our goals is to be able to predict how damage to different brain regions will impact upon behaviour. This involves studying anatomical and functional brain imaging data and cognitive abilities in large populations of healthy individuals and those with brain damage caused by stroke, tumours and neurosurgery. The anatomical imaging indicates which parts of the brain have been damaged. The cognitive assessments tell us which cognitive functions have been affected following damage to different regions. By monitoring how cognitive functions change over time we can determine how patients typically recover.  Functional imaging data are used to understand how parts of the brain change their function when other parts of the brain are damaged. This provides clues as to how the brain is able to recover from brain damage and might give clues as to why others don’t recover and how they might be helped to recover.

Cathy Price’s team are investigating the brain regions involved in a wide variety of functions that support the ability to perform language tasks (e.g. reading aloud and comprehending speech). These functions include speech perception, object recognition, semantic knowledge, word retrieval, sentence construction and speech production. Studying how brain activity associated with these functions differs across individuals is providing clues to why patients differ in the degree to which they recover after stroke.


Damage to part of the medial temporal lobes known as the hippocampus can arise in patients with LGI1-limbic encephalitis and cause problems with a number of cognitive functions including memory, mind-wandering, and spatial navigation.  Eleanor Maguire’s team has shown that these patients are unable to construct scene imagery, and this underpins their other cognitive problems.

To understand why this happens, the normal mechanisms underlying these functions need to be investigated. Eleanor Maguire’s team have conducted a large-scale cognitive and neuroimaging study of these functions in healthy volunteers. It revealed that constructing scene imagery is the process that mediates the relationship between autobiographical memory, future thinking and spatial navigation (Clark et al., 2019).  A breakdown in scene imagery, therefore, offers an explanation for the difficulties that patients with hippocampal damage have on a range of cognitive tests. Understanding this can focus efforts to help patients with hippocampal damage.


Autism is a disorder that affects how people interact with others. Geraint Rees’s team have shown that differences in the structure and function of local brain areas are linked to higher-order cognitive symptoms in autism (Watanabe et al., 2019). This could lead to earlier diagnosis if these differences also exist in young children and better diagnosis for autistic people who do not present typical behavioural symptoms.


N-Methyl-D-aspartate receptor (NMDAR) antibody encephalitis is an autoimmune disorder. It affects a range of the brain’s functions, with symptoms including psychosis, disorientation, amnesia, and seizures. Early treatment is critical to improving outcomes for patients, but it is difficult to diagnose because symptoms are so varied.  Karl Friston’s team have shown that, in patients with NMDA antibody encephalitis, non-invasive EEG data (with measures electrical activity from the brain) can be used to pick out the same recognisable pattern of abnormal receptor signalling as invasive techniques. This research demonstrates that readily available EEG can be used to safely test for the disorder.  The same study also furthers our understanding of the physiological basis of the disorder, finding that the excitatory NMDA neurons showed a greater loss of signalling than inhibitory interneurons (Symmonds et al., 2018).


Deep Brain Stimulation (DBS) is an effective treatment for several neurological and psychiatric disorders.   In order to gain insights into the therapeutic mechanisms of DBS and to advance future therapies, Vladimir Litvak’s team (Oswal et al., 2016) introduced an experimental protocol and analysis pipeline for recording how stimulation to small deep brain regions affects widespread brain activity measured by magnetoencephalography (MEG) (which measures magnetic fields produced by the electrical activity of neurons).  The optimal deep brain stimulation site can also be determined by investigating brain structure in patients who have the best response to DBS (Akram et al., 2017; 2018).


Dynamic causal modelling of MEG data can be used to look at the balance between inhibitory (GABA) or excitatory (glutamate) neurotransmitter function (Bhatt et al., 2016). Modelling this can help decide whether drugs that prolong hyper-excitability or reverse inhibition are best for recovery (eg. after stroke).