Genetic Algorithm with Logistic Regression for Diagnosis and Prognosis of Alzheimer’s Disease — ASN Events
  1. Schedule
  2. Disease Informatics
  3. Genetic Algorithm with Logistic Regression for Diagnosis and Prognosis of Alzheimer’s Disease

Genetic Algorithm with Logistic Regression for Diagnosis and Prognosis of Alzheimer’s Disease (#61)

Ping Zhang 1 , Piers Johnson 1 , Luke Vandewater 1 , William Wilson 2 , Paul Maruff 3 , Greg Savage 4 , Petra Graham 4 , Lance Macaulay 5 , Kathryn A Ellis 6 , Cassandra Szoeke 6 , Ralph Martins 7 , Christopher Rowe 6 , Colin Masters 6 , David Ames 6
  1. CCI, CSIRO, Marsfield, NSW, Australia
  2. CCI, CSIRO, North Ryde, NSW, Australia
  3. Cogstate Ltd, Melbourne, VIC, Australia
  4. Macquarie University, North Ryde, NSW, Australia
  5. CMSE, CSIRO, Parkville, VIC, Australia
  6. The University of Melbourne, Parkville, VIC, Australia
  7. Edith Cowan University, Perth, WA, Australia

Background

Assessment of risk and early diagnosis of Alzheimer’s disease (AD) is a key to its prevention or slowing the progression of the disease.  Previous research on risk factors for AD typically utilizes statistical comparison tests or stepwise selection with regression models. Outcomes of these methods tend to emphasize single risk factors rather than a combination of risk factors. However, a combination of factors, rather than any one alone, is likely to affect disease development. Genetic algorithms (GA) can be useful and efficient for searching a combination of variables for the best achievement (eg. accuracy of diagnosis), especially when the search space is large, complex or poorly understood, as in the case in prediction of AD development.

Results

GA in combination with logistic regression (LR) was used for finding one or more sets of neuropsychological tests which can best predict the progression to AD. Data from the Australian Imaging, Biomarkers & Lifestyle (AIBL) Study of Ageing with 36 months follow up were examined. A set of 37 neuropsychological variables including depression and anxiety measures was used for identifying the best subsets for prediction of conversion from healthy to mild cognitive impairment (MCI) or AD and for conversion from MCI to AD. Multiple sets of neuropsychological variables were identified by GA to best predict conversions between clinical categories, with a cross validated AUC of 0.90 for prediction of HC conversion to MCI/AD and 0.86 for MCI conversion to AD within 36 months.

Conclusions

This study showed the potential of GA application in the neural science area. It demonstrated that the combination of the variables is superior in performance than the use of single significant variables for prediction of progression of disease. Variables more frequently selected by GA might be more important as part of the algorithm for prediction of disease development.