Vasiliki Bikia
Stanford University
bikia@stanford.edu
Bio
Vasiliki (Vicky) Bikia is a Postdoctoral Researcher at Stanford University, working with Prof. Roxana Daneshjou at the Institute for Human-Centered Artificial Intelligence (HAI). She earned her Ph.D. in Biomedical Engineering from the Swiss Federal Institute of Technology in Lausanne (EPFL) under the guidance of Prof. Nikolaos Stergiopulos. During her doctoral studies, she developed and validated clinical AI-enabled tools to enhance non-invasive cardiovascular monitoring. Prior to that, Vicky obtained her Advanced Diploma in Electrical and Computer Engineering from the Aristotle University of Thessaloniki (AUTH). Moreover, she was a Visiting Postdoctoral Fellow at the Byers Center for Biodesign at Stanford, collaborating with Prof. Oliver Aalami on software pipelines designed to improve data accessibility and interoperability in digital health applications. Vicky is an active member of several research communities, contributing to numerous successful research proposals and publishing clinical studies that drive innovation in the application of AI in medicine.
Areas of Research
- AI for Healthcare and Life Sciences
Multimodal AI for Improved Health Monitoring and Patient Communication
The landscape of healthcare is rapidly evolving, driven by the increasing complexity of health data and the growing demand for personalized patient care. Health data is inherently multimodal, encompassing inputs such as textual reports, medical images, wearable sensor data, and patient histories. Accurately interpreting this data requires tools capable of integrating diverse data types to inform clinical decisions. Recent advancements in AI, particularly large language models (LLMs) and large multimodal models (LMMs), offer promising solutions to these challenges. These models can synthesize complex data from multiple sources, providing healthcare professionals with more accurate predictions and actionable insights. Additionally, they can enhance patient communication by automating responses to common queries, helping patients understand their health conditions and make informed decisions. My work so far addressed the need for non-invasive tools to assess cardiovascular health by developing and validating AI-driven models for key biomarkers, including blood pressure, cardiac parameters, and arterial stiffness. These methods have shown improved performance compared to existing techniques while reducing costs and operational dependencies. Extending my research into digital health, I focused on leveraging data from smartphones and wearables to design open-source pipelines that enhance data accessibility and interoperability. Looking ahead, I aim to develop advanced representations of multimodal health data to improve biomarker and outcome prediction. I am also enthusiastic about creating patient-facing chatbots that can interpret imaging results and hospital discharge instructions. By utilizing advanced prompt engineering and retrieval-augmented generation techniques, these chatbots may provide accurate, unbiased communication, enhancing patient understanding and engagement. We are at a pivotal moment in history, with the potential to make a profound impact on medicine. By combining these AI advancements with robust monitoring systems, we can push the boundaries of patient care, making healthcare more accessible, effective, and personalized for everyone.