Brain sulcal folding and mental illness: investigating causal associations and stratification approaches for psychosis
Lead Research Organisation:
University of Cambridge
Department Name: Psychiatry
Abstract
The folding of the outer surface of the brain is highly variable across individuals. Variations in folding are associated with brain function, potentially contributing to mental illness. Some folds of the brain shows a particularly high level of variability between individuals, including the cingulate fold (or known to scientists as the cingulate gyrus and its associated furrow, the cingulate sulcus), which presents as a single fold in some people and a double fold in others. Variability in cingulate folding is an almost uniquely human feature. Whilst the cingulate fold is present in other mammals, humans and chimpanzees are the only animals that can have a double cingulate fold. Initial evidence suggests that the folding pattern may also be important for vulnerability to mental health problems, but providing definitive answers has been limited by small samples sizes and technical challenges in measuring the degree of folding on brain scans.
To drive this work forward, we have improved the way we can measure variable folding on brain scans. Currently the gold standard way to measure it is for a trained expert to spend about 30 minutes per scan per variable sulcus, looking at a brain scan from various points of view on a screen and measuring it using a computer mouse to draw lines on the scan. This method is prone to human error and is too slow to do at scale. We have developed a new computer program to measure cingulate folding rapidly, reliably and accurately, and we will refine this program, extending it to measure folding in other parts of the brain, and make it freely available for other scientists and clinicians to use in future. We will run the program on the brain scans in over 50,000 people (from the large UK Biobank and USA ABCD studies), giving each participant summary measures of several variable folds. We will deposit the results in the relevant database so that researchers studying this acclaimed resource can easily access the results and relate cingulate folding to other measures of interest.
It is not at all known what causes folding variability. We will conduct the first study to examine the molecular genetic basis of sulcal folding variation, and we will use a genetic technique called mendelian randomisation that uses genetic information to check whether particular brain folding variants are contributory causes to mental illness.
We will illustrate the clinical importance of folding by testing whether it can predict an important clinical outcome in schizophrenia, namely response to treatment. If we can identify, at first presentation, those patients who will need to treatments that are usually only introduced as last resort, we could potentially improve early outcomes.
Our project will clarify the importance of folding for psychological function and mental illness in much greater detail and in much larger numbers than has been done before. To do this we will take advantage of several existing high quality studies that have already collected data that we will use to clarify the role of the PCS in health and illness. In particular, we will look at cingulate folding in relation to hallucinations using scans from over 1100 patients with schizophrenia in studies drawn from Europe and the USA - a sample five time larger than the largest previous study. We will examine the effects of cingulate folding in the general population in large studies of over 50,000 people. We will use data from these studies to see if the cingulate folding pattern is linked to vulnerability to certain psychiatric symptoms, psychological abilities, and the wiring patterns of the brain.
This line of work is exciting: our computer program could in future be used on brain scans to calculate "biomarkers" that help doctors and patients make better treatment decisions, and will provide new knowledge on how the brain folds and what difference this makes to our thinking and vulnerability to mental illness.
To drive this work forward, we have improved the way we can measure variable folding on brain scans. Currently the gold standard way to measure it is for a trained expert to spend about 30 minutes per scan per variable sulcus, looking at a brain scan from various points of view on a screen and measuring it using a computer mouse to draw lines on the scan. This method is prone to human error and is too slow to do at scale. We have developed a new computer program to measure cingulate folding rapidly, reliably and accurately, and we will refine this program, extending it to measure folding in other parts of the brain, and make it freely available for other scientists and clinicians to use in future. We will run the program on the brain scans in over 50,000 people (from the large UK Biobank and USA ABCD studies), giving each participant summary measures of several variable folds. We will deposit the results in the relevant database so that researchers studying this acclaimed resource can easily access the results and relate cingulate folding to other measures of interest.
It is not at all known what causes folding variability. We will conduct the first study to examine the molecular genetic basis of sulcal folding variation, and we will use a genetic technique called mendelian randomisation that uses genetic information to check whether particular brain folding variants are contributory causes to mental illness.
We will illustrate the clinical importance of folding by testing whether it can predict an important clinical outcome in schizophrenia, namely response to treatment. If we can identify, at first presentation, those patients who will need to treatments that are usually only introduced as last resort, we could potentially improve early outcomes.
Our project will clarify the importance of folding for psychological function and mental illness in much greater detail and in much larger numbers than has been done before. To do this we will take advantage of several existing high quality studies that have already collected data that we will use to clarify the role of the PCS in health and illness. In particular, we will look at cingulate folding in relation to hallucinations using scans from over 1100 patients with schizophrenia in studies drawn from Europe and the USA - a sample five time larger than the largest previous study. We will examine the effects of cingulate folding in the general population in large studies of over 50,000 people. We will use data from these studies to see if the cingulate folding pattern is linked to vulnerability to certain psychiatric symptoms, psychological abilities, and the wiring patterns of the brain.
This line of work is exciting: our computer program could in future be used on brain scans to calculate "biomarkers" that help doctors and patients make better treatment decisions, and will provide new knowledge on how the brain folds and what difference this makes to our thinking and vulnerability to mental illness.
Technical Summary
The cingulate gyrus presents as a single fold in some people (about 10%) and a double fold in others. About 50% of people have a prominent double fold (a neurodevelopmental feature, seen only in humans and chimpanzees). Individual variation in the degree of folding is linked to liability to specific forms of psychopathology, indicating a potential use of quantitative measures of cingulate folding to be used as clinical biomarkers. We will probe the role of brain folding variation (focussed on the cingulate but also considering other regions) in health and illness. Folding variation is poorly handled by existing automated neuroimaging workflows. The gold standard of measurement is currently laborious manual tracing, greatly restricting the pace of discovery in this field. We have developed a deep learning method to measure the degree of cingulate folding accurately in an automated manner using standard MRI structural data. We will refine and extend this and produce a freely available software tool to easily implement this.
We will examine the genetic causes of cingulate folding in much greater detail than has been done before, and assess the consequences of variation in sulcal folding for treatment response, psychopathology, brain connectivity and cognition. In particular, we will examine if brain folding, especially in the cingulate, associates with liability to hallucinations in >1000 patients with schizophrenia, and/or predicts treatment resistance. We will use large scale, high quality datasets where MRI data, psychopathology and cognitive function are already available. This research will clarify the neural and psychological underpinnings of psychopathology (potentially helping treatment development), and lead to sulcal folding biomarkers that (1) can be used in large scale future basic science studies (eg to investigate the genetic and environmental basis of sulcal folding), and (2) can be further developed for clinical utility.
We will examine the genetic causes of cingulate folding in much greater detail than has been done before, and assess the consequences of variation in sulcal folding for treatment response, psychopathology, brain connectivity and cognition. In particular, we will examine if brain folding, especially in the cingulate, associates with liability to hallucinations in >1000 patients with schizophrenia, and/or predicts treatment resistance. We will use large scale, high quality datasets where MRI data, psychopathology and cognitive function are already available. This research will clarify the neural and psychological underpinnings of psychopathology (potentially helping treatment development), and lead to sulcal folding biomarkers that (1) can be used in large scale future basic science studies (eg to investigate the genetic and environmental basis of sulcal folding), and (2) can be further developed for clinical utility.
