Dr. S. Divya is an AI researcher with a PhD in Computer Science and Engineering and over two decades of experience spanning academia and industry. She is currently a Postdoctoral Research Fellow in the Faculty of Medicine at the University of Southampton, where she leads AI-driven research on pulmonary imaging for COPD as part of an Asthma + Lung UK–funded project. Her work focuses on developing advanced machine learning and computer vision methods to support clinically meaningful biomarkers and improve respiratory diagnostics.
Dr. Divya has led and contributed to multiple UKRI-funded projects, including HOMEAI-ENABLE and OTWISE-AI, delivering impactful AI solutions for healthcare and housing accessibility. Her research portfolio spans medical imaging, ethical and lightweight AI systems, decision support platforms, and social-impact technologies, with publications in high-impact international journals and conferences. A Senior Member of IEEE, she is an active reviewer and assessor for UKRI and international research forums. Her contributions have been recognised through national and international awards, invited talks, and innovation accolades, reflecting her strong commitment to advancing AI for societal benefit.
AI for Social Good: Transforming Healthcare and Housing Accessibility
Public health outcomes are shaped not only by timely diagnosis and effective treatment, but also by the environments people live in and the accessibility of support services. This keynote presents a practical AI for social good framework that connects trustworthy clinical AI with inclusive, user centred decision support in social care, with the aim of improving early detection, reducing inequities, and strengthening system level consistency.
Three applied case studies illustrate pathways from research to real world impact. In medical imaging, robust preprocessing, segmentation, and calibrated classification reduce variability across scanners and readers, improving diagnostic reliability and supporting earlier, more consistent decision making. In housing accessibility, decision support using natural language processing and structured knowledge interprets functional needs, maps them to standardised vocabularies, and generates transparent recommendations that help make home adaptation pathways fairer and faster for older adults and disabled people. In respiratory health, automated CT vascular analysis and topology informed biomarkers support detection of subtle early signals of conditions such as COPD, enabling proactive risk stratification and prevention oriented care.
Throughout, responsible AI in practice is emphasised, including transparency, interpretability, bias monitoring, governance, and workflow alignment to support safe and equitable deployment. The keynote concludes with a forward agenda for integrating multimodal data and cross sector partnerships so that AI innovation translates into measurable improvements in population health and lived experience.