Integrative Neuroscience – The Role of a Standardized Database
Evian Gordon*
*The Brain Dynamics Centre, Westmead Hospital, NSW 2145, Australia * Brain Resource International Database (The Brain Resource Company), NSW 2007, Australia * Department of Psychological Medicine, University of Sydney, NSW 2006, Australia
Brain ResourceTM
Summary Neuroimaging databases usually consist of large subject numbers, with flexibility and diversity of measures, and analyses distributed across laboratories.1,2 This poster outlines preliminary outcomes from the first entirely standardized and centralized International Human Brain Database. Rationale • There is a paucity of consistent and specific findings in neuroimaging, psychometrics and genetics across neuropsychiatric conditions. • This lack of consistency may be significantly confounded by the methodological variability across research facilities. • Strict standardization allows an “evidence based” evaluation of the sensitivity and specificity of findings across laboratories. Our Integrative Approach 1. Spatio-temporal Integration. 2. Integration across modalities. 3. Central-Autonomic Integration. 4. Individual Differences. Methods • Database currently comprises 1,000 normal subjects (age range 6 – 70 yrs). Also growing numbers of neurological and psychiatric disorders acquired from nine laboratories (New York, Rhode Island, London, Holland, Adelaide, Melbourne and Sydney). • All data is acquired in precisely the same manner (H/W, S/W, task instructions) with identical paradigms that are designed to reflect a profile of brain functions and structure. • Psychometrics include memory, attention and executive function. Psychophysiology (EEG and ERP) include oddball, go/no-go, working memory, PPI and affect faces. sMRI via MPRAGE, Dual Echo and DTI. fMRI uses 4 of the Psychophysiology paradigms. Genetics is via cheek swabs (or bloods).
Results
Selected examples of preliminary “Integrative” outcomes from the first 550 normative subjects
(1) Spatio-temporal Integration
sMRi: Gray matter
mean time to complete (ms)
Discussion Selected outcomes from the standardized international database show its capacity for: •Sensitivity: The consistency of the effects of individual differences, including age, sex, history of early life stress and personality dimensions, can be seen across modalities in a large standardized database. This allows us to control for these effects in the comparison of clinical and non-clinical subjects. •Specificity: Elucidating the specificity of brain function findings in a range of psychiatric disorders is increased with the use of completely standardized protocols across multiple laboratories worldwide. •Treatment Efficacy: Testing the same subjects before and after medication allows for an “evidence based” assessment of brainbody functional changes with treatment. •Conclusion: The focus of this consortium is to bring together explicitly complementary “Integrative” multimodal parameters, including the use of a numerical simulation of fundamental physiological mechanisms across scale.4
Time to Completion (Switching of Attention)
]
54000 51000 48000
]
45000 42000
] ]
39000
11-20
21-30
31-50
51-70
Each 1mL increase in gray matter is associated with a 36 ms decrease in time to completion of the switching of attention task (ß= -.359, p = .018).
age
(2) Integration Across Modalities
Psychophysiology (Alpha EEG peak frequency) 21-30 yrs 51-70 yrs
Reverse digitspan
A
6
Psychometrics (Reverse digitspan)
‘Without -arousal’ D
‘With -arousal’ B
5
] ]
C
]
4
]
E
11-20
21-30
F
31-50
51-70
The slowing of EEG alpha peak frequency with age is associated with the decline in working memory performance seen in subjects over 50 years of age. Each 1Hz increase in frequency is associated with a 0.21 increase in reverse digit span score and this was independent of age (p = .15).
Age
(3) Central-Autonomic Integration
A C
5.6
E
P has ic arous al (S CRs for a s tim ulus bloc k )
s kin c o nd uc t a nc e le ve l
M icros iemens
5.4 5.2 5 4.8 4.6
-arousal’
Start
3
6
9
12
15
18
B
D
5.6
Tim e (s ec onds )
‘Without 21 24 27 30
s kin c o nd uc t a nc e le ve l
F
Tonic arous al (no S CRs for a s ubs equent bloc k )
M ic ros iem ens
5.4 5.2 5 4.8 4.6 Star t 123 126 129 132 135 138 141 144 147 150
Simultaneously recorded fMRI and skin conductance data shows that the amygdala (A) and MFC (B) is activated during increased skin conductance responses (E), whilst the hippocampus (C) and LFC (D) is activated during low levels of arousal (F). 3 #
References
1. Chicurel M. Databasing the brain. Nature 2000, 822-825. 2. OHBM-The Governing Council of the Organization for Human Brain Mapping. Neuroimaging Databases.Science 2001; 292: 1-4. 3. Williams LM, et al. Arousal dissociates amygdala and hippocampal fear responses: Evidence from simultaneous fMRI and skin conductance recording. NeuroImage 2001; 14: 1070 – 1079. 4. Robinson PA, et al. Prediction of electroencephalographic spectra from neurophysiology. Physics Review 2000; 63: 021903. # fMRI data is not acquired in all subjects.
Tim e (s ec onds )
(4) Individual Differences (Aging)
ERP latency to targets
300000
Time to completion (executive maze)
]
250000
]
200000
150000
]
]
11-20
21-30
31-50
51-70
Delayed latency of ERPs to target stimuli with age is associated with a significantly longer time to completion on the executive maze. Age accounts for 35% of the variance associated with time to complete the executive maze (R2= .349).
time to complete (ms)
age
For further details contact: Evian Gordon:eviang@brainresource.com