HIC- AI road map

Radiology analysisAutomatic audiometryPrediction of hearing health condition

Sara Valencia Cadavid PhDBRC HH Translational research manager

Road map HIC - AI

Front End Developer(s)

Staff required

Back End Developer(s)

User Story and Acceptance Criteria Researcher

Scientific /Clinical Analyst

Coordinator

ACTIONS

1. Characterisation of the user population
2. Define users needs and expectations
1. Workshops
2. Focus groups
3. Interviews.
3. Creation of User Stories
4. Graphic prototypes and visualisation (Mockups)

ACTIONS

1. Define Database infrastructure
2. Request acces to additional DB
3. Technological requirements (storage. processing capability, graphic interfase etc.)

Understanding of user experience and navigation requirements (Front-End)

Stage 4

Datasets and technological requirements

ACTIONS

1. Expert consensus
1. Advisor Scientific Commitee
2. Literature Review.
3. Defining input values
4. Defining outputs
5. Outpust presentation (predictive value or recomendation )

Identifying appropriate predictors and outpus

Stage 2

Defining the clinical question(s) or outcome(s) of interest (output).

Road Map

Stage 3

Stage 1

ACTIONS

1. Medicines and Healthcare products Regulatory Agency (MHRA) approval.
2. Meet NICE stardars

ACTIONS

1. Model training.
2. Connecting the back end to the front end.
3. Optimisation and validation of the model.
4. Traffic and stability testing of the platform.
5. User story validation tests.

DOCUMENTS

1. System architecture document.
2. User manuals, technical manual, installation manual
3. Data dictionary

Licensing the AI predictive model.

Stage 8

Integracion and Validation of the Model

Stage 7

Documentation

Stage 6

Developing the AI predictive model

Stage 5

ACTIONS

1. Bussines case - Bussines Model
2. Training
3. Adoption

ACTIONS

1. Updates
2. Complaince with NICE rules
3. UK Medical Research Council’s guidance for developing and evaluating complex interventions.

Commercialization or OpenLicencing of the product

Stage 9

Maintaining and evaluating the AI predictive model.

Stage 10

Key decisions

Experts to keep in the loop and invite to the initial stages and validation process

Identify additional sources of funding to develop the steps.

4. Funding

Define times to start and complete each stage

3. Time table

2. Population

Define the person or team that could support the activiy

1. Who?

Thank you