ARMOR Optimize Chad Byers - Project Manager Brendan Grebur - Domain - - PowerPoint PPT Presentation

armor optimize
SMART_READER_LITE
LIVE PREVIEW

ARMOR Optimize Chad Byers - Project Manager Brendan Grebur - Domain - - PowerPoint PPT Presentation

ARMOR Optimize Chad Byers - Project Manager Brendan Grebur - Domain Expert / Customer Liaison Rory Hool - Project Facilitator Caitlin Russ - Artifacts Manager What Optimization Means Finding solution(s) to maximize / minimize a set of defined


slide-1
SLIDE 1

ARMOR Optimize

Chad Byers - Project Manager Brendan Grebur - Domain Expert / Customer Liaison Rory Hool - Project Facilitator Caitlin Russ - Artifacts Manager

slide-2
SLIDE 2

What Optimization Means

  • Finding solution(s) to maximize / minimize a set of defined criteria, subject

to a set of constraints, for a specific problem

  • Problem
  • Improving a patient's quality of life by reducing polypharmacy

interactions

  • Criteria
  • Increasing or Reducing
  • Bodily aspects (appetite, balance, vision, etc.)
  • Constraints
  • Drug constraints

§ Increase / Reduce / Remove / No Alteration

  • Condition constraints

§ "To treat or not to treat"

slide-3
SLIDE 3

The Basic Terminology

  • Continuity of Care Record (CCR)
  • Body Measurements
  • Afflicting Conditions
  • Drugs (Past / Present)
  • Drugs
  • Prescribed and Over-the-counter (OTC)
  • Frequency and Dosage
  • Set of affected body aspects and a quantitative effect
  • Increase or Decrease
  • Condition
  • Treated by a group of drugs
  • Body Aspect
  • Similar to side effects
  • Appetite, vision, balance, depression, diarrhea
  • Treatment Plan
  • Current: Existing prescribed treatment plan
  • Recommended: Optimized treatment plans for review
slide-4
SLIDE 4

The Extended Terminology

  • Goal Profile
  • A collection of:
  • Body aspects each denoting a desired change
  • Other Goal Profiles
  • Example:
  • Better Balance { vision+ , dizziness- , lightheaded- }
  • Goal Set
  • An accumulation of weighted:
  • Goal profiles
  • Individual body aspects each with denoted change
  • Example:
  • (5 * Better Balance ) + (3 * depression-)
slide-5
SLIDE 5

Use Case Diagram

slide-6
SLIDE 6

Web Interface Front-end Class Diagram

slide-7
SLIDE 7

System Back (Server) - Evolutionary Computation

  • Modification Encoding
  • Representation of adjustments to current treatment plan
  • Treatment Adjustment ("Gene")
  • Per drug containing:
  • Increase / Decrease / Reduce / No Alteration
  • Population
  • Encodings competing to provide the best solution
  • Selection Protocol
  • Roulette
  • Tournament*
  • Rank
slide-8
SLIDE 8

Evolutionary Computation Back-end Class Diagram

slide-9
SLIDE 9

Evolutionary Optimization Algorithm

slide-10
SLIDE 10

System Front (Client) Statechart

slide-11
SLIDE 11

System Back (Server) Statechart

slide-12
SLIDE 12

Sequence Diagram: Creating And Using New Goal Profile

slide-13
SLIDE 13

Sequence Diagram: Evolutionary Optimization

slide-14
SLIDE 14

Sequence Diagram:

Display and Review of Treatment Plans

slide-15
SLIDE 15

Sequence Diagram: Goals Not Specified for Optimization

slide-16
SLIDE 16

Prototype Demo

  • High Level Features of User Interface
  • Upload CCR with Analysis Results
  • Select drug / condition constraints
  • Create custom Goal Profile
  • Select Goal Profiles and Body Aspects for Goal Set
  • Select Weighting Criteria for members of Goal Set
  • View Optimization summary
  • View Optimized treatment plans
  • Save / Delete treatment plans

Web Prototype