Skip to main content
Department of Psychiatry and Behavioral Sciences

Department of Psychiatry and Behavioral Sciences

Research projects

Concordance of onset types in siblings with autism spectrum disorders

Mentor:  Sally Ozonoff, Ph.D.
Duration:  Long-term
Description:  Autism begins in the first three years of life, with two distinct onset patterns evident. The early onset phenotype presents early in life, with parents reporting developmental abnormalities before the first birthday. The regression phenotype includes a period of typical or mostly typical development (18-24 months), followed by a loss of previously acquired skills and onset of symptoms of autism. The etiologies of these phenotypes, and whether they differ, are unknown. In multiplex families (those with more than one child with autism), it is unknown if both children demonstrate the same onset type or if this can differ within a family and similar genetic background. Trainees will conduct brief phone interviews with mothers of multiple children with autism regarding symptom onset. They will then analyze data and assist in publication write-up.

Background reading

  • Lainhart, J.E., Ozonoff, S., Coon, H., Krasny, L., & McMahon, W. (2002). Autism, regression and genetics. American Journal of Medical Genetics, 113, 231-237.
  • Werner, E., & Dawson, G. (2005). Validation of the phenomenon of autistic regression using home videotapes. Archives of General Psychiatry, 62, 889-895.

Neurocognitive foundations of intellectual impairments and psychiatric disorders in children with chromosome 22q11.2 deletion syndrome

Mentor:  Tony J. Simon, Ph.D.
Duration:  Long-term
Description :  Chromosome 22q11.2 deletion syndrome (22q) is a very common yet ill-understood neurogenetic disorder that produces characteristic cognitive impairments and elevated risk for psychological/psychiatric disorders (including ADHD, OCD & Schizophrenia). Our lab carries our translational research using cognitive processing, neuroimaging and genetics tools as well as psychological, behavioral and psychiatric assessments. Trainees will get direct experience of patient interaction along with the opportunity to learn about and get involved with data collection and analysis using the methods of cognitive neuroscience as well as preparing results for publication.

Background reading

  • Simon, T.J., Ding, L., Bish, J.P., McDonald-McGinn, D., Zackai, E.H., & Gee, J. (2005) Volumetric, connective and morphologic changes in the brains of children with chromosome 22q11.2 deletion syndrome: An integrative study. NeuroImage , 25: 169-180.
  • Simon, T.J., Bish, J.P., Bearden, C.E., Ferrante, S., Ding, L., Nguyen, V., Gee, J., McDonald-McGinn, D., Zackai, E.H., & Emanuel, B. (2005) A multi-level analysis of cognitive dysfunction and psychopathology associated with chromosome 22q11.2 deletion syndrome in children. Development and Psychopathology , 17, 753-784.

Developing screening and diagnostic tools using novel data mining techniques

Mentor:  Peter Yellowlees, M.D.
Duration:  Long-term
Description :  To further develop, and validate, a novel machine learning computerized screening tool analyzing visual or auditory material for depression, anxiety and substance abuse, and potentially other medical conditions, which is ideally suited for use in primary care., especially in rural regions. Ex-Ray is a new application of machine learning technology developed by PY that uses samples of a few minutes of a patient's objective behavior, including recorded voice, transcribed language, and visual images, to screen for and monitor common mental health or other conditions. It can be operated by a medical assistant or layperson with minimal training, and can maintain the confidentiality of the results and the nature of the screening. It can be configured to screen for or monitor a wide range of disorders using a single session's sample, and can be used repeatedly with long or short inter-test intervals to monitor change in a chronic condition. It has been piloted on a small sample of clinically confirmed patients and has shown better than 85% accuracy for distinguishing patients from controls. Trainees will develop hypotheses about specific groups of patients, collect data using audio or video interviews, and analyse the data using the Ex-Ray data mining technology before writing up their results.

Background reading

  • Diederich, J., & Yellowlees, P. (2002). Ex-ray: Text classification and the assessment of mental health. Proceedings of the Australasian Document Computing Symposium , Sydney, Australia.

Assessment of the BOLD signal in schizophrenia in three cortical regions

Mentor:  Jong H. Yoon, M.D.
Duration:  Short-term
Description :  fMRI presents an unprecedented opportunity to examine the neural underpinnings of schizophrenia. However, a fundamental methodological issue in the application of fMRI analysis to the study of this illness remain unresolved-- the verification that no difference in the fMRI marker of neural activity, the blood oxygen dependent level (BOLD) signal, is unaltered in schizophrenia. The presence of a difference between subjects with schizophrenia and controls would have significant impact on how we should conduct fMRI analysis of any study comparing activations between groups. With this study, trainees will be exposed to general experimental methods, statistics, and fMRI analysis pertinent to conducting studies comparing neural function in healthy controls and subjects with schizophrenia. The trainee will then apply this knowledge to analyze fMRI data, which have already been gathered. The trainees will also assist in generating a manuscript for publication.

The effect of neuroleptic medication on the BOLD signal and type II error in detecting activations using the general linear model

Mentor:  Jong H. Yoon, M.D.
Duration:  Short-term
Description :  This study is related to the above experiment in its examination of the BOLD signal in schizophrenia. This study's main aim is, however, on determining the effect of psychiatric medication on the BOLD signal. The trainee will analyze fMRI data that has already been generated, which are derived from patients with schizophrenia who have undergone fMRI scanning pre- and post-medication treatment. This unique data set provides a within subject design capable of evaluating whether psychiatric medications alter the BOLD response.