Please enable JavaScript.

Coggle requires JavaScript to display documents.

AI Award writing [2021] - Coggle Diagram

- - - - Predict optimal amlodipine dose for the individual to secure BP control with minimal side-effects
        
        Using a mechanistic approaches, augmented by AI techniques to refine dose predictions
    - - Proof of Concept
      - Pilot data
      - Efficacy
      - Usability data
      - Validation data
    - - Intellectual property position
      - Review of existing evidence
      - Identify any ongoing research in area of focus
      - Proposed barriers to clinical / social adoption
  - - - Reduced mortality...?
    - - Health data quality / availability?
      - Improved efficacy
    - - More efficient care pathway management
- - - - Create production AI/ML training infrastructure / pipeline
      - Train production model(s) for deployment
      - Develop production Hypertension treatment algorithm, with AI component
      - Integrate algorithm with AI into product code, in "silent running" mode
      - Update QMS to align with AI regs / policies
      - Implement CLM Treatment Engine for hosting HTN algorithm
      - Verification and Validation of updated product / algorithm
      - Build infrastructure & processes for model updates / increments based on collected data
    - - TLR4 feasibility output model (Bayesian network + PK/PD + DenseNet for dose prediction)
      - Comparative evaluations...
      - Consensus models?
      - Potential competitive or complementary models
      - Create experimental AI/ML training pipeline / infra
    - - Iterate deployed models based on data
      - Attempt to adopt rolling clinical development model (SaMD compatible?)
      - Potentially deploy non-SaMD initially, then transition?
      - Study objective (more data + SaMD evidence)
    - - CPRD
      - UKHTD
      - Provision data storage and integration infrastructure
      - Ascot