Shin Kitaoka

Fellowship Bio

I am a specialty registrar in Clinical Radiology in Southend, subspecialising in Head and Neck and PET. I previously studied medicine at the University of Cambridge and Imperial College London.

Fellowship Project

My fellowship project has focused on the clinical evaluation, governance and implementation of artificial intelligence (AI) tools within radiology. Under the supervision of Dr Tilak Das, consultant neuroradiologist and clinical AI lead, I have contributed to several projects assessing commercially available AI systems and developing practical resources to support their safe introduction into clinical practice.

I was primary co-author of a multicentre evaluation of AI tools for large vessel occlusion detection on CT angiography in patients with suspected anterior circulation stroke. This study reviewed 2,917 examinations across seven sites, with my role including data collection, curation and analysis. A key finding from this work was the importance of understanding the limitations of AI systems, including cases where outputs were difficult to interpret/explain and uncertainty around the precise anatomical boundaries assessed by the algorithms. This inspired the creation of a tool summary document for reference by users of any AI tool being deployed, clearly outlining intended use, known limitations and other practical considerations.

Further projects have included designing and performing data preparation for head-to-head evaluations of CT brain and fracture detection tools, and preparation of clinical safety case documentation (DCB0160) and user-facing tool summaries for an AI tool for assessing multiple sclerosis on MRI. I have also reviewed the current radiology AI landscape through a comparison of international AI registries across Europe, the USA and the UK, with this work accepted for conference presentation and currently being prepared for publication.

Fellowship Testimonial

The NHS Fellowship in Clinical AI has provided an invaluable opportunity to develop a comprehensive understanding of the clinical, technical and regulatory aspects of artificial intelligence. Through structured teaching, workshops, e-learning and practical project work, I have gained insight into both the opportunities AI presents and the challenges involved in ensuring its safe and effective implementation. The roadshows and workshops with key stakeholders in the evaluation and deployment of AI tools, has taught me that successful adoption is not solely based upon technical performance, but also clear understanding of the tool’s limitations, appropriate governance and engagement with end-users. My project work has strengthened my skills in clinical evaluation, data curation and analysis, regulatory documentation and communication of AI-related information to clinicians. It has also highlighted areas where further innovation is needed, particularly around explainability, transparency and efficient evaluation for post market surveillance. The fellowship has allowed me to teach and present research at a national level and contribute to multicentre projects. I am truly grateful for the experience and hope to continue supporting the safe and evidence-based use of AI in radiology throughout my career.