Cynthia M. Stonnington

Automatic classification of MR scans in Alzheimer's disease

To be diagnostically useful, structural MRI must reliably distinguish Alzheimers disease (AD) from normal aging in individual scans. Recent advances in statistical learning theory have led to the application of support vector machines to MRI for detection of a variety of disease states. The aims of this study were to assess how successfully support vector machines assigned individual diagnoses and to determine whether data-sets combined from multiple scanners and different centres could be used to obtain effective classification of scans. We used linear support vector machines to classify the grey matter segment of T1-weighted MR scans from pathologically proven AD patients and cognitively normal elderly individuals obtained from two centres with different scanning equipment. Because the clinical diagnosis of mild AD is difficult we also tested the ability of support vector machines to differentiate control scans from patients without post-mortem confirmation. Finally we sought to use these methods to differentiate scans between patients suffering from AD from those with frontotemporal lobar degeneration. Up to 96% of pathologically verified AD patients were correctly classified using whole brain images. Data from different centres were successfully combined achieving comparable results from the separate analyses. Importantly, data from one centre could be used to train a support vector machine to accurately differentiate AD and normal ageing scans obtained from another centre with different subjects and different scanner equipment. Patients with mild, clinically probable AD and age/sex matched controls were correctly separated in 89% of cases which is compatible with published diagnosis rates in the best clinical centres. This method correctly assigned 89% of patients with post-mortem confirmed diagnosis of either AD or frontotemporal lobar degeneration to their respective group. Our study leads to three conclusions: Firstly, support vector machines successfully separate patients with AD from healthy aging subjects. Secondly, they perform well in the differential diagnosis of two different forms of dementia. Thirdly, the method is robust and can be generalized across different centres. This suggests an important role for computer based diagnostic image analysis for clinical practice.

Accuracy of dementia diagnosis—a direct comparison between radiologists and a computerized method

There has been recent interest in the application of machine learning techniques to neuroimaging-based diagnosis. These methods promise fully automated, standard PC-based clinical decisions, unbiased by variable radiological expertise. We recently used support vector machines (SVMs) to separate sporadic Alzheimers disease from normal ageing and from fronto-temporal lobar degeneration (FTLD). In this study, we compare the results to those obtained by radiologists. A binary diagnostic classification was made by six radiologists with different levels of experience on the same scans and information that had been previously analysed with SVM. SVMs correctly classified 95% (sensitivity/specificity: 95/95) of sporadic Alzheimers disease and controls into their respective groups. Radiologists correctly classified 65–95% (median 89%; sensitivity/specificity: 88/90) of scans. SVM correctly classified another set of sporadic Alzheimers disease in 93% (sensitivity/specificity: 100/86) of cases, whereas radiologists ranged between 80% and 90% (median 83%; sensitivity/specificity: 80/85). SVMs were better at separating patients with sporadic Alzheimers disease from those with FTLD (SVM 89%; sensitivity/specificity: 83/95; compared to radiological range from 63% to 83%; median 71%; sensitivity/specificity: 64/76). Radiologists were always accurate when they reported a high degree of diagnostic confidence. The results show that well-trained neuroradiologists classify typical Alzheimers disease-associated scans comparable to SVMs. However, SVMs require no expert knowledge and trained SVMs can readily be exchanged between centres for use in diagnostic classification. These results are encouraging and indicate a role for computerized diagnostic methods in clinical practice.

Interpreting scan data acquired from multiple scanners: A study with Alzheimer's disease

Large, multi-site studies utilizing MRI-derived measures from multiple scanners present an opportunity to advance research by pooling data. On the other hand, it remains unclear whether or not the potential confound introduced by different scanners and upgrades will devalue the integrity of any results. Although there are studies of scanner differences for the purpose of calibration and quality control, the current literature is devoid of studies that describe the analysis of multi-scanner data with regard to the interaction of scanner(s) with effects of interest. We investigated a data-set of 136 subjects, 62 patients with mild to moderate Alzheimers disease and 74 cognitively normal elderly controls, with MRI scans from one center that were acquired over 10 years with 6 different scanners and multiple upgrades over time. We used a whole-brain voxel-wise analysis to evaluate the effect of scanner, effect of disease, and the interaction of scanner and disease for the 6 different scanners. The effect of disease in patients showed the expected significant reduction of grey matter in the medial temporal lobe. Scanner differences were substantially less than the group differences and only significant in the thalamus. There was no significant interaction of scanner with disease group. We describe the rationale for concluding that our results were not confounded by scanner differences. Similar analyses in other multi-scanner data-sets could be used to justify the pooling of data when needed, such as in studies of rare disorders or in multi-center designs.

Irritability in pre-clinical Huntington's disease

Irritability, together with depression and anxiety, form three salient clinical features of pre-symptomatic Huntingtons disease (HD). To date, the understanding of irritability in HD suffers from a paucity of experimental data and is largely based on questionnaires or clinical anecdotes. Factor analysis suggests that irritability is related to impulsivity and aggression and is likely to engage the same neuronal circuits as these behaviours, including areas such as medial orbitofrontal cortex (OFC) and amygdala. 16 pre-symptomatic gene carriers (PSCs) and 15 of their companions were asked to indicate the larger of two squares consecutively shown on a screen while undergoing functional magnetic resonance imaging (fMRI). Despite correct identification of the larger square, participants were often told that they or their partner had given the wrong answer. Size differences were subtle to make negative feedback credible but detectable. Although task performance, baseline irritability, and reported task-induced irritation were the same for both groups, fMRI revealed distinct neuronal processing in those who will later develop HD. In controls but not PSCs, task-induced irritation correlated positively with amygdala activation and negatively with OFC activation. Repetitive negative feedback induced greater amygdala activations in controls than PSCs. In addition, the inverse functional coupling between amygdala and OFC was significantly weaker in PSCs compared to controls. Our results argue that normal emotion processing circuits are disrupted in PSCs via attenuated modulation of emotional status by external or internal indicators. At later stages, this dysfunction may increase the risk for developing recognised, HD-associated, psychiatric symptoms such as irritability.

Comparison of Psychogenic Movement Disorders and Psychogenic Nonepileptic Seizures: Is Phenotype Clinically Important?

BACKGROUND Psychogenic non-epileptic seizures (PNES) and psychogenic movement disorders (PMDs) are common in neurology practice, yet it is not established whether clinically relevant differences between these two groups exist. METHODS In this retrospective chart review 172 patients were identified (PNES n = 116, PMD n = 56). RESULTS The whole group was characterized by female gender (82%), abuse history (45%), chronic pain (70%), depression (42%), subjective fatigue (47%), subjective cognitive complaints (55%), and referral for psychiatric evaluation (54%). Statistically significant differences (P <. 01) were found for age, education, frequency of symptoms, altered consciousness, developmental abuse, and coexisting anxiety. Clinical practice also differed for the two groups in history-taking and referrals for neuropsychological testing and/or psychiatric evaluation. CONCLUSIONS This retrospective study revealed more similarities than differences suggesting these are manifestations of the same psychopathology, with age and co-morbid anxiety potentially being important factors in predicting the symptomatic presentation. Prospective studies are needed to confirm our results. Future studies focusing more globally on somatoform disorders, rather than each phenotypic presentation, are likely needed to improve clinical care and outcomes.

Confirming psychogenic nonepileptic seizures with video-EEG: Sex matters

The influence of gender on psychogenic nonepileptic seizures (PNES) diagnosis was examined retrospectively in 439 subjects undergoing video-EEG (vEEG) for spell classification, of whom 142 women and 42 men had confirmed PNES. The epileptologists predicted diagnosis was correct in 72% overall. Confirmed epilepsy was correctly predicted in 94% men and 88% women. In contrast, confirmed PNES was accurately predicted in 86% women versus 61% men (p=0.003). Sex-based differences in likelihood of an indeterminate admission were not observed for predicted epilepsy or physiologic events, but were for predicted PNES (39% men, 12% women, p=0.0002). More frequent failure to record spells in men than women with predicted PNES was not explained by spell frequency, duration of monitoring, age, medication use, or personality profile. PNES are not only less common in men, but also more challenging to recognize in the clinic, and even when suspected more difficult to confirm with vEEG.

Localizing sadness activation within the subgenual cingulate in individuals: a novel functional MRI paradigm for detecting individual differences in the neural circuitry underlying depression

Variations in frontal lobe (FL) functional anatomy, especially the subgenual cingulate gyrus (SGC) suggest that mapping on an individual rather than group level may give greater insight regarding dysregulation of the neural circuitry involved in depression, as well as potentially provide more specific or individualized treatment plans for depressed patients. We designed a functional MRI task capable of imaging FL activity in individuals, including the SGC region, using a transient sadness paradigm. We sought to develop a method that may better detect individual differences of FL subregions related to sadness, since this region has been implicated to show dysregulation in depression. The task was based on a block design that also accommodates individual differences in responsivity to a sadness induction paradigm. Individual differences from nine non-depressed healthy volunteers were analyzed. We also performed functional connectivity analyses to further characterize our findings to the networks associated with the SGC in each individual. The study was designed to account for individual variation rather than using a true experimental design; therefore, no control group was necessary. As expected, due to inter-individual variability, the specific site of SGC activation during sadness varied across individuals. Activation was also observed in other brain regions consistent with other studies of induced sadness and depression. Patterns of functional connectivity to the SGC also highlighted neural circuits known to subserve sadness and depression. This task promises to more precisely localize a given individual’s functional organization of the brain circuitry underlying sadness, and potentially depression, in an efficient, standardized way. This task could potentially aid in providing individualized targets in the treatment of depression.

Looking Toward DSM-V : Should Factitious Disorder Become a Subtype of Somatoform Disorder?

BACKGROUND Factitious and somatoform-disorder patients are alike in that they both organize their lives around seeking medical services in spite of having primarily a psychiatric condition. In DSM-IV, the key difference is that factitious-disorder patients feign illness, and somatoform-disorder patients actually believe they are ill. Although patients may not be conscious of their motivation or even their behaviors, deliberately embellishing history or inducing symptoms exemplifies behaviors designed to enhance a self-concept of being ill. CONCLUSION For DSM-V, we propose reclassifying factitious disorder as a subtype within the somatoform-spectrum disorders or the proposed physical-symptom disorder, premised on our belief that deliberate deceptions serve primarily to portray to treaters the sense of being ill.

Double-Blind Crossover Study of the Cognitive Effects of Lorazepam in Healthy Apolipoprotein E ( APOE )-ε4 Carriers

OBJECTIVE To examine cognitive effects of pharmacologically induced somnolence in cognitively normal carriers and noncarriers of the apolipoprotein E (APOE)-epsilon4 allele, a common Alzheimers disease susceptibility gene. METHOD Between December 2005 and July 2007, healthy and cognitively normal carriers of the APOE-epsilon4 allele (heterozygotes; n = 18) and noncarriers (n = 18), 50 to 65 years old, participated in a double-blind crossover study of cognitive function before, 2.5 hours after, and 5 hours after administration of 2 mg oral lorazepam or placebo. Main outcome measures included the Groton Maze Learning Test (GMLT) for executive functioning and visuospatial working memory, the Rey Auditory-Verbal Learning Test (AVLT) for verbal memory, and the one-back test for attention and simple working memory. RESULTS At 2.5 hours after lorazepam administration, GMLT total errors score (P = .04), AVLT long-term memory (P = .01), and AVLT percent recall (P = .005) reflected worse performance in heterozygotes. By multivariate analysis, the combined set of all 6 measures for heterozygotes versus noncarriers yielded P = .003 for 2.5 hours and P = .58 for 5 hours. No differences were observed for somnolence, speed, attention, or simple working memory at any time points. CONCLUSIONS Despite comparable levels of associated somnolence, lorazepam appears to diminish verbal and visuospatial memory more in healthy late-middle-aged heterozygotes than in noncarriers, whereas attention and reaction time are similarly affected in both. Additional studies are needed to determine whether substantial lorazepam-induced memory detriments predict subsequent onset of cognitive decline and conversion to mild cognitive impairment or Alzheimers disease. Clinicians should be aware of the potential for cognitive decline with lorazepam in healthy late-middle-aged individuals, especially those at a higher risk for Alzheimers disease. TRIAL REGISTRATION clinicaltrials.gov Identifier: NCT00586430.

Studying ventricular abnormalities in mild cognitive impairment with hyperbolic Ricci flow and tensor-based morphometry

Mild Cognitive Impairment (MCI) is a transitional stage between normal aging and dementia and people with MCI are at high risk of progression to dementia. MCI is attracting increasing attention, as it offers an opportunity to target the disease process during an early symptomatic stage. Structural magnetic resonance imaging (MRI) measures have been the mainstay of Alzheimers disease (AD) imaging research, however, ventricular morphometry analysis remains challenging because of its complicated topological structure. Here we describe a novel ventricular morphometry system based on the hyperbolic Ricci flow method and tensor-based morphometry (TBM) statistics. Unlike prior ventricular surface parameterization methods, hyperbolic conformal parameterization is angle-preserving and does not have any singularities. Our system generates a one-to-one diffeomorphic mapping between ventricular surfaces with consistent boundary matching conditions. The TBM statistics encode a great deal of surface deformation information that could be inaccessible or overlooked by other methods. We applied our system to the baseline MRI scans of a set of MCI subjects from the Alzheimers Disease Neuroimaging Initiative (ADNI: 71 MCI converters vs. 62 MCI stable). Although the combined ventricular area and volume features did not differ between the two groups, our fine-grained surface analysis revealed significant differences in the ventricular regions close to the temporal lobe and posterior cingulate, structures that are affected early in AD. Significant correlations were also detected between ventricular morphometry, neuropsychological measures, and a previously described imaging index based on fluorodeoxyglucose positron emission tomography (FDG-PET) scans. This novel ventricular morphometry method may offer a new and more sensitive approach to study preclinical and early symptomatic stage AD.