Join today

AI in Clinical Laboratory Management

Description:

AI tools now appear in laboratory analysers, digital pathology systems, quality control software and laboratory information management systems (LIMS). This course shows you what these tools can and cannot do, where they fit in your laboratory’s workflow, and how to judge them before you buy, validate or rely on them.

You will work through real examples, such as AI-assisted white cell classification in blood film analysers and the FDA authorisation of Paige Prostate in 2021. The course is practical throughout. You will practise asking suppliers the right questions, reading performance figures critically and planning a safe roll-out.

Learning aim

To give laboratory managers the knowledge and judgement to evaluate, implement and govern AI tools safely, so that their laboratory gains quality and efficiency without risk to patients.

Learning objectives

The course will: 
  • Introduce the core AI concepts and terms that laboratory managers meet in supplier claims and published studies. 
  • Show where AI is used at the pre-analytical, analytical, post-analytical and operational stages of laboratory work. 
  • Explain the regulatory, quality and data protection requirements that apply to AI in UK laboratories. 
  • Give you a structured method for evaluating, validating and monitoring an AI tool. 
  • Cover the ethical, workforce and accountability issues that come with automated decision support.
Write your awesome label here.

Learning outcomes

By the end of the course, you will be able to: 

  1. Explain the key AI concepts and performance metrics, and calculate PPV from sensitivity, specificity and prevalence. 
  2. Identify at least one AI use case at each stage of the testing process and judge whether it would suit your own laboratory. 
  3. Describe the requirements that apply to an AI tool under the UK Medical Devices Regulations, ISO 15189:2022, DCB0160 and UK GDPR. 
  4. Assess a supplier’s AI product using a structured set of evaluation questions and the NICE Evidence Standards Framework. 
  5. Plan a local validation and monitoring approach, with clear go-live criteria and KPIs. 
  6. Recognise risks of bias and automation bias and set out safeguards that keep professional accountability for results.

Created with