VIEWS

• Research type

  Research Study

• Full title

  Visualising Intelligent Explanations for Weaning Support

• IRAS ID

  362795

• Contact name

  Brijesh Patel

• Contact email

  brijesh.patel@imperial.ac.uk

• Sponsor organisation

  Imperial College London

• Duration of Study in the UK

  1 years, 0 months, 1 days

• Research summary

  Aim
  Improve the care of patients who need help with their breathing.

  Background
  Every year, 70,000 patients in intensive care need help with breathing. This is delivered through a machine called a ventilator. To use a ventilator properly, clinicians must understand the patient and their disease.

  Deciding when to reduce ventilator support is important. Clinicians must balance different treatment goals to achieve the best outcome. Leaving patients on a ventilator too long can cause lung injury. However, patients must be strong enough to breathe on their own. Previous research has shown differences in clinical practice when reducing ventilator support. There is mixed success and no agreement on best practice.

  Artificial intelligence (AI) is when computers are used for tasks that usually need human intelligence. It is thought AI can help solve difficult clinical problems. Explainable AI (XAI) is the development of AI that can provide explanations for its findings. It is thought that visual tools like graphs, maps and dashboards, designed with clinicians and patients, can deliver on the ambition of XAI.

  Methods:
  1. Define key information needed for deciding when to reduce ventilator support using focus groups and expert opinion.
  2. Co-design visual tools with clinicians and patients that show information from AI models and include variables needed for reducing ventilator support. This will be done by using a questionnaire and two co-design workshops. The visual tools will be shown on an interactive computer system.

  Potential benefits
  This research can be used to support decision-making on intensive care, benefitting patients and clinicians. Consequently, treatment can be delivered more consistently, meaning improved safety, fewer complications and less time on the ventilator. The NHS will benefit through shorter hospital stays and lower healthcare costs.

• REC name

  London - Westminster Research Ethics Committee

• REC reference

  26/PR/0646

• Date of REC Opinion

  12 Jun 2026

• REC opinion

  Further Information Favourable Opinion