AI in Imaging - beyond just detection

5.00 pm
Coffee and registration

5.30 pm       
AI Prioritisation in Imaging: From Detection to Pathway Transformation
Dr Rhidian Bramley, Consultant Radiologist, Christie NHS Foundation Trust, Clinical Lead for AI and Digital Diagnostics with the Greater Manchester Cancer Alliance and Greater Manchester Imaging and Pathology Networks     

Talk synopsis:

Artificial intelligence in imaging is often framed as computer aided detection. In practice, its impact depends on how it is integrated into clinical workflow.

This lecture will explore real world experience of AI prioritisation in stroke imaging and within the Greater Manchester chest X-ray AI programme. A key lesson was that prioritisation alone does not improve outcomes. Early deployment demonstrated that unless reporting workforce and workflow are aligned to prioritised studies, the intended benefits are not realised.

Dr Bramley will describe how simulation modelling was used to determine the optimal proportion of studies to prioritise, and to understand the potential unintended consequences for non prioritised patients. The principle underpinning this work is risk stratification. The same stratification framework was then applied to evaluate negative predictive value, support pathway redesign, reduce unnecessary referrals, and build a defensible clinical and economic case for change.

Finally, he will discuss how identification of a high confidence normal cohort may provide a structured and governable route towards supervised and future autonomous reporting in selected chest X-ray pathways.

Learning Points:

1. Understand that AI prioritisation requires workflow redesign, not just algorithm deployment. Effective clinical impact depends on aligning staffing, reporting processes and operational models with AI outputs.
2. Recognise how simulation modelling can optimise prioritisation strategies and mitigate unintended harms. Over prioritisation can adversely affect non prioritised patients, and modelling helps determine safe thresholds.
3. Appreciate how risk stratification and negative predictive value can support pathway redesign and future supervised AI reporting. Identifying low risk cohorts enables safe service transformation and underpins the case for carefully governed increases in AI autonomy.

6.30 pm
Radiology Reimagined at Scale: Delivering a fast and accurate diagnosis through clinician-led technology
Dr Kiran Patel, Regional Commercial Lead, Hexarad

Talk synopsis:

Radiology bottlenecks have become a system-wide constraint, delaying diagnosis, disrupting patient flow, and increasing pressure across hospitals. This talk explores why many digital and AI initiatives fail to translate into real performance gains, and why technology alone is not the answer. Drawing on real-world NHS experience, it argues that the most effective use of AI is not to write reports, but to fix workflows, surface operational intelligence, and better support radiology teams. By rethinking how work is organised and measured, radiology services can unlock capacity, strengthen governance, and deliver faster, higher-quality diagnosis at scale.

Key learning points:

1. Radiology bottlenecks are system problems, not reporting problems
2. AI delivers the greatest value when applied to workflow and intelligence, not report writing
3. Clinician-led, data-driven systems are essential for sustainable improvement

7.00 pm
Meeting close

This course provides 2 CPD Credits in accordance with the CPD Scheme of the Royal College of Radiologists.
CPD certificates will be awarded to attendees.