AWS IMAGINE Grant Children’s Health Innovation Award

Brightpoint's Home Visiting and Doula Programs support high-risk young women and families to improve maternal and child health. By leveraging AWS generative AI tools, Brightpoint aims to enhance these programs' reach and effectiveness - generating personalized care plans, dynamic content, automated data exchange, and AI-enhanced staff training and decision support. Harnessing generative AI, Brightpoint seeks to amplify its vital community-based services and serve more vulnerable mothers and children to advance greater equity in maternal and infant health. 

As Canada's leading mother-and-child hospital, CHU St. Justine aims to develop an end-to-end AI workflow on AWS to enhance the usability of their centralized data lake (containing data from over 25+ hospital systems and more than 2.5 million patients) by extracting insights from the unstructured data. Focusing first on chronic conditions in children with neurodevelopmental and psychiatric disorders, they seek to characterize and analyze patterns of multimorbidity, bridging the gap between mental and physical conditions and unlocking the potential for precision child health.

Children's Mercy is establishing a living lab that will advance digital clinical pathways by leveraging smart home technology. By enabling proactive notifications, remote monitoring, and enhanced insights on triggers, clinicians will be better equipped to intervene early with at-risk patients. Built on Amazon Web Services (AWS), this scalable approach aims to benefit pediatric patients by providing families real-time insights and clinicians data-driven tools for more personalized, proactive care. The first condition to be advanced will be pediatric asthma. 

The Association of Public Health Laboratories (APHL), in partnership with Ruvos, developed Detor, an Electronic Test Order and Result (ETOR) Intermediary Solution built on AWS, which aims to improve the exchange of test orders and results between public health laboratories and healthcare providers. Detor has the potential to enhance capabilities across the public health landscape, starting with improving timeliness and accuracy for newborn screening, where early interventions are crucial.

The Elizabeth Glaser Pediatric AIDS Foundation (EGPAF) is building a solution to provide clinicians with real-time insights into HIV patient risks for children in Kenya, Malawi, and Zimbabwe on AWS. By integrating generative AI with existing electronic health and medical records, EGPAF aims to enhance healthcare providers’ ability to predict and prevent HIV treatment failures, enabling more personalized and proactive care for patients, and improving long-term patient data and feedback.

Great Ormond Street Hospital (GOSH) is pioneering the development of a generative AI tool to transform the care for children with wrist fractures – the most commonly fractured bone in childhood. This innovative system will detect and classify the fracture type, align treatment recommendations with the latest evidence-based guidelines, and automatically draft radiology reports and referral letters for doctors and patients to ensure expedited and improved patient care through the clinical pathway. Supported by AWS, GOSH is leveraging a substantial, multicentric UK dataset of pediatric radiographs to train and validate this prototype, setting a new standard for digital innovation in pediatric fracture management.

Memorial Sloan Kettering Cancer Center (MSK) is leveraging generative AI capabilities to extract valuable insights from MSK’s expansive clinicogenomic, harmonized oncologic real-world dataset which combines molecular alteration data from >125,000 tumor samples with natural language processing-derived clinical information. MSK is also developing an AI-assisted conversational agent which will integrate the MSK clinicogenomic data with large language models to provide personalized cancer outcome predictions and treatment recommendations. This project aims to enable advancements in biomarker and therapeutic development to empower clinicians and researchers to improve adult and pediatric cancer care while also driving scientific discoveries.

The Rady Children's Institute for Genomic Medicine (RCIGM) is developing a pediatric genomics large language model using AWS. They aim to empower physicians with better access to their extensive database of over 10,000 critically ill children's genomic sequences. RCIGM is pioneering two AI solutions: JericaBot, a generative AI platform for doctors to query patient genomic data in natural language, and Gtrx, an AI-assisted pipeline that automates the curation of rare genetic disease treatment options. These tools seek to enhance the utilization of pediatric genomic information in clinical care globally.

Texas Children's is launching an accelerator program focused on high-performance compute (HPC) and advancements in AI and data science via graphics processing units (GPUs) and complex graph database analysis. This tech-forward program will amplify development of new techniques, creating potential breakthroughs in pediatric research for rare disease drug development, genomics and personalized medicine, critical care surveillance, and image-based diagnostic accuracy.