Carolina Makowski

Evaluating accuracy of striatal, pallidal, and thalamic segmentation methods: Comparing automated approaches to manual delineation

ABSTRACT Accurate automated quantification of subcortical structures is a greatly pursued endeavour in neuroimaging. In an effort to establish the validity and reliability of these methods in defining the striatum, globus pallidus, and thalamus, we investigated differences in volumetry between manual delineation and automated segmentations derived by widely used FreeSurfer and FSL packages, and a more recent segmentation method, the MAGeT‐Brain algorithm. In a first set of experiments, the basal ganglia and thalamus of thirty subjects (15 first episode psychosis [FEP], 15 controls) were manually defined and compared to the labels generated by the three automated methods. Our results suggest that all methods overestimate volumes compared to the manually derived “gold standard”, with the least pronounced differences produced using MAGeT. The least between‐method variability was noted for the striatum, whereas marked differences between manual segmentation and MAGeT compared to FreeSurfer and FSL emerged for the globus pallidus and thalamus. Correlations between manual segmentation and automated methods were strongest for MAGeT (range: 0.51 to 0.92; p<0.01, corrected), whereas FreeSurfer and FSL showed moderate to strong Pearson correlations (range 0.44–0.86; p<0.05, corrected), with the exception of FreeSurfer pallidal (r=0.31, p=0.10) and FSL thalamic segmentations (r=0.37, p=0.051). Bland‐Altman plots highlighted a tendency for greater volumetric differences between manual labels and automated methods at the lower end of the distribution (i.e. smaller structures), which was most prominent for bilateral thalamus across automated pipelines, and left globus pallidus for FSL. We then went on to examine volume and shape of the basal ganglia structures using automated techniques in 135 FEP patients and 88 controls. The striatum and globus pallidus were significantly larger in FEP patients compared to controls bilaterally, irrespective of the method used. MAGeT‐Brain was more sensitive to shape‐based group differences, and uncovered widespread surface expansions in the striatum and globus pallidus bilaterally in FEP patients compared to controls, and surface contractions in bilateral thalamus (FDR‐corrected). By contrast, after using a recommended cluster‐wise thresholding method, FSL only detected differences in the right ventral striatum (FEP>Control) and one cluster of the left thalamus (Control>FEP). These results suggest that different automated pipelines segment subcortical structures with varying degrees of variability compared to manual methods, with particularly pronounced differences found with FreeSurfer and FSL for the globus pallidus and thalamus. HIGHLIGHTSManual segmentation of subcortical structure is evaluated against 3 automated tools.Correspondence with manual labels was greatest with MAGeT‐Brain.Greatest between‐method variability was noted for FreeSurfer and FSL‐FIRST.Larger striatum and pallidum were found in first episode psychosis across methods.Subcortical shape profiles in patients vs. controls differed between MAGeT and FSL.

Age-related cortical thickness trajectories in first episode psychosis patients presenting with early persistent negative symptoms.

Recent work has clearly established that early persistent negative symptoms (ePNS) can be observed following a first episode of psychosis (FEP), and can negatively affect functional outcome. There is also evidence for cortical changes associated with ePNS. Given that a FEP often occurs during a period of ongoing complex brain development and maturation, neuroanatomical changes may have a specific age-related component. The current study examines cortical thickness (CT) and trajectories with age using longitudinal structural imaging. Structural T1 volumes were acquired at three time points for ePNS (N=21), PNS due to secondary factors (N=31), non-PNS (N=45) patients, and controls (N=48). Images were processed using the CIVET pipeline. Linear mixed models were applied to test for the main effects of (a) group, (b) time, and interactions between (c) time and group membership, and (d) age and group membership. Compared with the non-PNS and secondary PNS patient groups, the ePNS group showed cortical thinning over time in temporal regions and a thickening with age primarily in prefrontal areas. Early PNS patients also had significantly different linear and quadratic age relationships with CT compared with other groups within cingulate, prefrontal, and temporal cortices. The current study demonstrates that FEP patients with ePNS show significantly different CT trajectories with age. Increased CT may be indicative of disruptions in cortical maturation processes within higher-order brain regions. Individuals with ePNS underline a unique subgroup of FEP patients that are differentiated at the clinical level and who exhibit distinct neurobiological patterns compared with their non-PNS peers.

The effect of second-generation antipsychotics on hippocampal volume in first episode of psychosis: longitudinal study

Background Current neuroscience literature has related treatment with aripiprazole to improved memory performance and subcellular changes in the hippocampus. Aims To explore the volumetric changes in hippocampal grey matter in people with a first episode of psychosis (FEP) treated with second-generation antipsychotics. Method Baseline and 1-year follow-up magnetic resonance images were obtained. Hippocampal volumes were estimated by using FreeSurfer and MAGeT-Brain. Subgroups included: aripiprazole (n=13), olanzapine (n=12), risperidone/paliperidone (n=24), refused-antipsychotics (n=13) and controls (n=44). Results Aripiprazole subgroup displayed significant increases in bilateral hippocampal volume compared with all other subgroups (FreeSurfer: all P’s<0.012; MAGeT-Brain: all P’s<0.040). Conclusions Aripiprazole is a first-line, second-generation treatment option that may provide an added benefit of pro-hippocampal growth. The biological underpinnings of these changes should be the focus of future investigations and may be key towards achieving a better clinical outcome for more individuals. Declaration of interest M.L. received financial assistance/compensation for research and educational events from Janssen-Ortho, Eli Lilly, Roche and Otsuka/Lundbeck Alliance. A.K.M. received financial assistance/compensation for research and educational activities from Pfizer, Janssen-Ortho, AstraZeneca and Bristol-Myers Squibb. R.J. received consultancy honorariums from Pfizer and Janssen-Ortho. Copyright and usage

A longitudinal study of cognitive insight and cortical thickness in first-episode psychosis

Among individuals with psychosis, those with poor cognitive insight (lower Self-Reflectiveness, higher Self-Certainty) show volumetric reductions in cortical structure. We evaluated whether changes in cognitive insight are associated with progressive changes in cortical structure in first-episode psychosis (FEP) and control subjects. Beck Cognitive Insight Scale ratings and magnetic resonance imaging scans were acquired at baseline for 130 FEP and 52 controls, 59 FEP and 28 controls at 1-year, and 53 FEP and 20 controls at 2-years. Cortical thickness was computed across scans and analyzed with linear mixed models. At baseline, groups did not differ on Self-Reflectiveness or Self-Certainty. At baseline, higher Self-Reflectiveness significantly correlated with thinner right occipital cortex in FEP, and higher Self-Certainty was significantly negatively correlated with cortical thickness in left posterior cingulate in controls. Longitudinal analysis showed that Self-Reflectiveness and Self-Certainty did not change over time in either group. Interestingly, the lack of change in cognitive insight aligned with longitudinal cortical thickness results, where no interaction effects were seen with cortical thickness between time and either Self-Reflectiveness or Self-Certainty. Exploratory analyses with a reduced threshold found that in FEP, across all time-points, higher Self-Certainty associated with thinner cortex in left posterior cingulate/precuneus. Results suggest that the posterior cingulate may be a common neural correlate for Self-Certainty in FEP and non-clinical subjects.

Functional Neural Correlates of Social Approval in Schizophrenia

Social approval is a reward that uses abstract social reinforcers to guide interpersonal interactions. Few studies have specifically explored social reward processing and its related neural substrates in schizophrenia. Fifteen patients with schizophrenia and fifteen healthy controls participated in a two-part study to explore the functional neural correlates of social approval. In the first session, participants were led to believe their personality would be assessed based on their results from various questionnaires and an interview. Participants were then presented with the results of their supposed evaluation in the scanner, while engaging in a relevant fMRI social approval task. Subjects provided subjective reports of pleasure associated with receiving self-directed positive or negative feedback. Higher activation of the right parietal lobe was found in controls compared with individuals with schizophrenia. Both groups rated traits from the high social reward condition as more pleasurable than the low social reward condition, while intergroup differences emerged in the low social reward condition. Positive correlations were found in patients only between subjective ratings of positive feedback and right insula activation, and a relevant behavioural measure. Evidence suggests potential neural substrates underlying the cognitive representation of social reputation in schizophrenia.

Longitudinal trajectory of clinical insight and covariation with cortical thickness in first-episode psychosis

Among people with a first-episode of psychosis, those with poorer clinical insight show neuroanatomical abnormalities in frontal, temporal and parietal cortices compared to those with better clinical insight. Whether changes in clinical insight are associated with progressive structural brain changes is unknown. We aimed to evaluate 1) associations between clinical insight and cortical thickness at a baseline assessment, 2) covariation between clinical insight and cortical thickness across baseline, one-year and two-year follow-up assessments, and 3) the predictive value of clinical insight for cortical thickness at one-year and two-year follow-ups. Scale for the assessment of Unawareness of Mental Disorder ratings and magnetic resonance imaging scans were acquired at baseline, one-year, and two-year follow-ups in 128, 74, and 44 individuals with a first-episode psychosis, respectively. Cortical thickness metrics were then computed at baseline, one-year and two-year follow-ups and analyzed with linear mixed models. At baseline, clinical insight was not significantly associated with cortical thickness in any region. Longitudinal mixed effects models showed that a worsening in clinical insight between the one-year and two-year assessments was significantly associated with cortical thinning in dorsal pre-central and post-central gyri. Cortical thinning in left fusiform gyrus at two-years was predicted by poorer clinical insight at baseline. Results suggest that poor clinical insight soon after the onset of a first-episode psychosis may lead to progressive cortical changes in temporal lobe, while changes in clinical insight during the second year covary with cortical thinning in circumscribed dorsal frontal and parietal cortices.

Clarifying associations between cortical thickness, subcortical structures, and a comprehensive assessment of clinical insight in enduring schizophrenia

BACKGROUND The relationship between poor insight and less favorable outcomes in schizophrenia has promoted research efforts to understand its neurobiological basis. Thus far, research on neural correlates of insight has been constrained by small samples, incomplete insight assessments, and a focus on frontal lobes. The purpose of this study was to examine associations of cortical thickness and subcortical volumes, with a comprehensive assessment of clinical insight, in a large sample of enduring schizophrenia patients. METHODS Two dimensions of clinical insight previously identified by a factor analysis of 4 insight assessments were used: Awareness of Illness and Need for Treatment (AINT) and Awareness of Symptoms and Consequences (ASC). T1-weighted structural images were acquired on a 3 T MRI scanner for 110 schizophrenia patients and 69 healthy controls. MR images were processed using CIVET (version 2.0) and MAGeT and quality controlled pre and post-processing. Whole-brain and region-of-interest, vertex-wise linear models were applied between cortical thickness, and levels of AINT and ASC. Partial correlations were conducted between volumes of the amygdala, thalamus, striatum, and hippocampus and insight levels. RESULTS No significant associations between both insight factors and cortical thickness were observed. Moreover, no significant associations emerged between subcortical volumes and both insight factors. CONCLUSIONS These results do not replicate previous findings obtained with smaller samples using single-item measures of insight into illness, suggesting a limited role of neurobiological factors and a greater role of psychological processes in explaining levels of clinical insight.

T172. MULTIMODAL QUANTIFICATION OF MEMORY CIRCUIT MICROSTRUCTURE IN FIRST EPISODE PSYCHOSIS

Abstract Background Integrity of hippocampal subfield structure and associated limbic circuitry subserves various memory processes, a domain that is impaired in psychosis and an important predictor of functional outcome. We use a novel atlas that encapsulates both hippocampal subfields and surrounding white matter (WM), forming the ‘memory circuit’, to assess volumes with high-resolution MRI, and microstructure with quantitative T1 (qT1). Our aims were to examine 1) group by time interactions on memory measures and the memory circuit, and 2) explore the relationships between the chosen memory measures and limbic structures, informed by results from 1), in a longitudinal sample of first episode of psychosis (FEP) patients. Methods Nineteen FEP and 20 controls with baseline and 3-month follow-up data were included. Logical Memory and Visual Reproduction Subscales of the Weschler Memory Scale, and MRI scans on a 3T scanner were collected. High-resolution T2-weighted images (0.64 mm3) were input to the MAGeT Brain algorithm to obtain volumes of hippocampal subfields and surrounding WM, defined by fimbria, alveus, fornix, and mammillary bodies. Mean qT1 values for each hippocampal subfield and WM structure were sampled from MP2RAGE (1 mm3) qT1 maps. Linear mixed models were used to assess group by time interactions on memory measures, volumes and qT1. To begin, total hippocampal volumes and WM structure for each hemisphere were examined using a Bonferroni correction for multiple comparisons, followed by post-hoc tests of individual subfields and WM structures. Linear models were then used to assess relationships between baseline memory and change in anatomical measures of interest in FEP. Models controlled for sex, education, age, and brain volume. Results Significant group by time interactions emerged on bilateral mean WM qT1 (left: F1,65=9.3, p=.003; right: F1,65=10.6, p=.002), where it was found that within the FEP group, qT1 (relaxation time in ms) increased over the 3-month follow-up period. Looking at WM structures separately, the interaction was driven by qT1 changes in fimbria, fornix, and mammillary bodies bilaterally (p’s<.05). No significant group by time interactions were found with respect to volumes or memory, although a trend-like group by time interaction on right fornix volume was found (F1,64=5.6, p-uncorrected=.02). Finally, brain-behaviour relationships were explored, restricting our anatomical measure of interest to mean qT1 values within bilateral WM. Although no tests passed correction for multiple comparisons, there was a trend association between better delayed recall of Visual Reproduction and decreases in qT1 of combined WM on the right hemisphere (F1,11=3.72, p=.08), driven by changes in qT1 of the right fornix (F1,11=4.4, p=.06). Discussion This study reveals significant microstructural changes in WM output circuitry of the hippocampus shortly after a FEP. Specifically, increases in qT1 were found within fimbria, fornix, and mammillary bodies bilaterally. Given that T1 relaxation times are typically shorter in WM, an increase in qT1 may reflect a combination of decreased myelin content and increased inflammation. Furthermore, preliminary data suggest better visual memory at baseline is associated with lower qT1 within WM microstructure over a 3-month period, suggesting that preserved non-verbal memory ability shortly after a FEP may manifest in a protective anatomical phenotype, particularly within the fornix. Given the importance of the hippocampal-fornix circuit in FEP, both with respect to memory and as a theorized hub of pathophysiology in psychosis, a better understanding of WM microstructure in relation to cognitive profiles in patients may offer a new perspective for treatment targets.

T178. PRIOR SUB-THRESHOLD PSYCHOTIC SYMPTOMS ASSOCIATED WITH THICKER RIGHT INFERIOR FRONTAL GYRUS AMONG PATIENTS IN A FIRST EPISODE OF PSYCHOSIS

Abstract Background Individuals with attenuated or sub-threshold psychotic symptoms (STPS) are considered at-risk for psychosis. The notion that STPS represent “early psychosis” holds promise as it suggests the possibility of charting the developmental course of psychotic illness with neuroimaging. However, recent evidence suggests that a significant minority of patients in a first episode of psychosis (FEP) do not recall pre-onset STPS, suggesting diversity in early positive symptom course. This diversity may be reflected at the level of neurodevelopment. While imaging studies of at-risk youth and FEP patients reveal progressive trends in cortical thinning across stages of illness, none have considered the STPS history of FEP patients. To better understand neurobiological trends across illness stages, we investigate the relationships between STPS history and cortical thickness in FEP patients using a whole-brain approach. Methods Patients (N=93) were recruited from a specialized early intervention clinic for FEP at the Douglas Mental Health University Institute. The Circumstances of Onset and Relapse Schedule was administered to identify youth who recalled at least one of nine expert-selected STPS prior to their FEP (STPS+, N=67) compared to those who did not (STPS-, N=26). These STPS include: Suspiciousness or odd ideas of reference, odd or bizarre ideas that are not delusional, unusual or eccentric behavior, unusual perceptual experiences that are not clearly psychotic, disorganized or odd speech, inappropriate affect, hallucinations or delusions (sub-threshold), and passivity experiences. Age and sex-matched healthy controls were recruited (N=83) for comparison. Participants were scanned on a 1.5T MRI scanner between 1 and 3 times at baseline, at 1-year follow-up, and at 2-year follow-up. Structural T1-weighted images were processed through the CIVET 2.1 pipeline. Cortical thickness values of 320 scans (143 HC, 123 STPS+, 54 STPS-) that passed quality control were extracted for group analysis. Linear mixed effects models controlling for effects of age, sex, education, and mean whole-brain thickness were applied to obtain vertex-wise F-test maps comparing groups. Results Post-hoc vertex-wise t-test maps were thresholded with Random Field Theory (p-cluster=0.001) and revealed that compared to controls, only STPS- patients exhibited significantly thinner cortical thickness in the right inferior frontal gyrus (peak t(162.3)=4.13, p<0.001). Examination of mean cortical thickness within this cluster, comparing patient groups only, revealed that compared to STPS+ patients, STPS- patients exhibited significantly thinner cortical thickness (t(172)=-2.55, p=0.01). This difference was most pronounced at baseline. Discussion These results indicate that within the right prefrontal cortex, STPS+/- patients may undergo different cortical maturation trajectories leading up to and through a first episode of psychosis. These differences may explain differential vulnerability to sub-threshold psychotic symptomology before a full-blown episode. In addition to suggesting differential underlying neurobiology related to STPS, these results suggest the importance of considering STPS history in mapping the trajectories of cortical thickness among FEP patients.