About

The Hawaii Digital Health Lab is dedicated to research at the intersection of healthcare, human-in-the-loop AI, human-centered computing, deep learning, and data science. Our mission is to develop innovative computational tools that improve healthcare delivery and patient outcomes.

We focus on creating practical solutions that can be deployed in real-world healthcare settings, from AI-powered diagnostic tools to crowdsourcing platforms that gather valuable health data from diverse populations.

Our research is driven by the belief that technology, when thoughtfully designed and rigorously tested, can make healthcare more accessible, equitable, and effective for all communities, particularly those in the Pacific region.

By training the next generation of digital health researchers and collaborating with healthcare providers and institutions, we aim to build a lasting infrastructure for health technology innovation in Hawaii and beyond.

The Hawaii Digital Health Lab is directed by Dr. Peter Washington, a faculty member at the University of Hawaii at Manoa. Dr. Washington brings extensive expertise in artificial intelligence, machine learning, and their applications to healthcare and precision medicine.

Under Dr. Washington's leadership, the lab has secured significant funding including a $2.18 million NIH New Innovator Award, demonstrating the caliber of research being conducted. His vision guides the lab's focus on developing AI tools that are both effective and equitable.

Dr. Washington has assembled a diverse research team that includes postdoctoral researchers, PhD students, masters students, undergraduate researchers, and high school students, creating a collaborative environment that fosters innovation and mentorship.

His research interests span machine learning for healthcare, crowdsourcing for health data, digital diagnostics, and ensuring fairness and equity in AI-powered health tools, positioning the lab at the forefront of responsible health technology development.

The lab employs a wide range of cutting-edge technologies including deep learning frameworks, computer vision systems, natural language processing tools, and multimodal data analysis platforms. These technologies form the foundation of our AI-powered diagnostic and intervention systems.

Machine learning models developed in the lab are trained on diverse datasets to ensure they perform accurately across different demographic groups. This includes convolutional neural networks for image analysis, transformer models for sequential data, and self-supervised learning approaches for handling limited labeled data.

Cloud computing platforms from providers like Amazon Web Services and Google Cloud enable the lab to process large-scale health datasets and deploy models at scale. This infrastructure supports both research experimentation and the development of production-ready health tools.

The lab also develops human-in-the-loop systems that combine automated AI predictions with human expert input, creating hybrid approaches that are more accurate and trustworthy than either humans or machines working alone.

The lab conducts several major research projects across digital health domains. One flagship project focuses on developing fair breast cancer prediction models using mammogram images, working in partnership with the AI Precision Health Institute to ensure these tools work equitably across diverse patient populations.

Another significant project involves creating digital diagnostic tools for mental health conditions and developmental disorders. These tools leverage smartphone-based data collection and AI analysis to enable earlier identification and intervention.

The lab has also been involved in building benchmark datasets for health AI research, addressing a critical need in the field for standardized, well-annotated data that can be used to train and evaluate diagnostic models.

Research on data missingness and its impact on AI model performance represents another important focus area, developing methods to handle incomplete health records and ensure that AI tools remain reliable even when working with imperfect data.

Health equity is a core principle guiding all research at the Hawaii Digital Health Lab. The lab actively investigates and addresses biases in AI health tools, ensuring that these technologies perform accurately for people of all backgrounds, ethnicities, and socioeconomic statuses.

The lab's participation in the NIH AIM-AHEAD program, where the team placed 3rd in the 2023 Health Equity Data Challenge, demonstrates their commitment to developing AI systems that serve underrepresented and historically marginalized communities.

Located in Hawaii, the lab is uniquely positioned to address health equity challenges facing Pacific Island communities, Native Hawaiian populations, and other diverse groups in the region. This geographic context informs their research priorities and methodological approaches.

By developing and sharing open-source tools, benchmark datasets, and research findings, the lab contributes to a broader scientific community effort to make health AI more inclusive and equitable worldwide.