Introducing Bayesian Networks in Neuropsychological Measures

1
Introducing Bayesian Networks in
Neuropsychological Measures

• Presented at 2006 APS Annual Conference
• Toshi Yumoto
• University of Maryland, Abt Associate
• Gregory Anderson
• Xtria, Adler School of Professional Psychology
• Daisy Wise
• University of Maryland

2
Bayesian Networks

• It is very difficult for practitioners to combine
  this information objectively
• Paul Meehl, Clinical versus Statistical
  Prediction (1954)
• Bayesian Networks provide an effective way to
  combine different sources of information such as
• Information from distinct domains
• Demographic (e.g. gender, race/ethnicity) and
  Clinical Information (e.g. previous diagnosis,
  different type of tests)
• It is easy to add new information or modify
  existing information

3
Bayesian Network (2)

• A Bayes Net is suitable for discrete data
• Continuous data needs to be converted into
  categorical data
• Complex models can be divided into several
  conditionally independent parts
• Model estimation is easier
• Straightforward to add conditional independent
  data
• Bayesian Network is visual and intuitive
• Class assignments are expressed as a proportion
  (e.g. percent)
• Provides a prediction for all parameters based on
  limited (or no) information
• Allows both subjective and objective evaluation
  of a network
• Microsoft Belief Networks (MSBNx, Kadie, Hovel,
  Horvitz, 2001) was used to build a Bayesian
  Network

4
Steps in Network Development

• Part I
• Conversion of continuous scores to discrete
  categories by Finite Mixture Approach
• Part II
• Create Bayesian Network based on hypothesized
  model
• Estimate conditional probabilities
• This research used a Latent Class Model
• Part III
• Examine and Modify model

5
Measures and Sample

• The ATLAS is a comprehensive assessment for ADHD
  containing 7 different sections (Anderson Post,
  2006). One of the sections is a series of
  neuropsychological measures and observations of
  performance during testing.
• This paper examines three of the major traits of
  the neuropsychological measures for ADHD.
• Diagnoses of ADHD LD were gathered from a
  parent report.
• The sample is around 220 subjects, 8-18 years of
  age, from across the nation, gathered in the test
  field trials.

6
Creation of Discrete scores

• The assumption was made that test scores are from
  more than one distribution
• A Finite Mixture Model was therefore utilized
• Cut scores were established using the
  intersection of the distributions
• of mixture distribution of cut scores 1
• A discrete score was assigned based on a persons
  mixture distribution characteristic
• For more information contact authors.

7
Finite Mixture Analysis for Trail A
Cut Score
8
Hypothesized Model
9
Model Specification

• Three Distinct Domains
• Visual Trace/Sequence
• Three Latent Classes
• Memory
• Three Latent Classes
• Impulsive/Error
• Two Latent Classes
• Second Order Latent Class
• Four Latent Classes
• Diagnosis (LD/ADHD)
• Four manifest categories
• Typical, LD, ADHD, and LD/ADHD

10
Specification of the Bayesian Network

• For each domain conditional probability of
  indicator variables are specified given latent
  class membership (for that domain).
• These probabilities are first specified based on
  the assumption that we have no information.
• These are then updated, given information such as
  test scores or clinical observations.

11
Partial Model Visual Trace NetworkNo
information known
12
Partial Model Visual Trace NetworkAll item
scores known
13

• Probability of LD/ADHD state given
• Middle Visual Trace level
• Middle Memory level
• High Impulsive level
• SOClass and LD/ADHD states are unobserved
• Proportions are expected class states

14

• Probability of LD/ADHD state with Six test
  results
• Trail A time (middle) and error (middle)
• Trail B time (middle-low) and error (middle)
• Word Memory 1 (low) and Word Memory 3 (low)
• Other nodes have expected category distribution
• Highest probability indicates most likely category

15

• It is easy to combine additional information such
  as clinical observations and gender to improve
  model prediction.
• Clinical observation is conditionally independent
  from other nodes given LD/ADHD states (i.e. only
  affect the probability of LD/ADHD node)
• Gender has direct effect on LD/ADHD states and
  Impulsive/Error level, which indirectly affect
  second order class states.
• This type of information is harder to add later
  and should be included from the beginning, if
  appropriate.

16
Summary

• Bayesian Networks provide an effective way to
  express and examine hypothesized models.
• The model performance can be compared with
  precision of prediction (e.g. LD/ADHD diagnoses
  in this research).
• Any statistical procedures estimating expected
  scores (i.e. probability of responses) may be
  used to build a network.
• Bayes Net uses discrete data, therefore latent
  class model and latent trait model (with discrete
  proficiency levels) nicely fit model development.
• Combining additional information is straight
  forward and relatively easy
• Understanding of conditional independence is the
  key
• Bayes Net estimates expected probability from
  available information
• Makes best possible diagnosis without complete
  data

17
Contacts

• Toshi Yumoto
• fyumoto_at_umd.edu
• Gregory Anderson
• ganderson_at_xtria.com
• Daisy Wise
• dawise_at_umd.edu