Billings Clinic

Billings, Montana  |  January 20, 2019

2006

2010-2013

2014

2015

2016-2020

Modeling the Reading the Brain

Using "Bench" Science to Address Clinical Problems

My talk

  • The Observation
  • The Plan
  • The Study
  • Moving Forward

The Observation

130+ years later

  • Diagnosis is only based on history and experience
  • No agreed upon tests
  • No quantitative measures/biomarkers
  • It's more complicated than just reading
  • Serious methodology problems

Needs

  • Defined approach with objective measures
  • Enable detection, dx, triage, and grading
  • Widely deploy these measures
  • Personalized therapy

To support that we need:

  • Establish behavioral model and neuroetiology
  • Quantitative characterization for predictive framework

The Plan

The Plan: Defining Neuroetiology

Cross-sectional Design

  • Healthy (n = 75) vs. Diagnosed (n = 75)
  • Controlled for development, education, language
  • Ideally, reading ability

The Plan: Behavioral Model

f_x(Fixation\,Duration) =Text_{features} + Dx + Dx_{sub}

The Plan: Behavioral Model

f_x(Fixation\,Map) = Picture\,predictability + Dx + Dx_{sub}

The Plan: Behavioral Model

f_x(Correct\, Answers) = Dx + Dx_{sub}

small

smoll

LexTale

The Plan: Quantitative Characterization

Structural MRI

f_{x}(Structure) = Behavior_{read} + Dx

Eye tracking

Part. 1

Part. 2

Part. n

...

3D group T-test

The Plan: Quantitative Characterization

FIRE-fMRI

The Plan: Quantitative
                  Characterization

f_x(IQ) \neq Dx + D_{subgroup}
f_x(Fluency) = Dx + D_{subgroup}
f_x(Phonology) = Dx + D_{subgroup}

The Study

The Study: Current State

Data to date:

  • 25 Dx participants
  • 6 healthy participants (14 more to write dissertation)
  • More weekly thanks to UAC

Moving Forward

Billings Clinic

Billings, Montana  |  January 20, 2019

Made with Slides.com