# CRCS held Computing for Population Maternal and Child Health Workshop April 26, 2024 In April 2024, the Center for Research on Computation and Society held a workshop titled “Computing for Population Maternal and Child Health”. Organized by Niclas Boehmer (Postdoctoral Fellow in Computer Science) and Puneet Sharma (Clinical Fellow in Pediatrics at Boston Children’s Hospital), the goal of the workshop was to bring together experts to discuss pressing issues in maternal and child health as well as opportunities to use innovative technological solutions to address these challenges. Invited attendees included clinicians, computer scientists, and public health researchers. The focus was on interdisciplinary collaboration to enhance healthcare delivery and outcomes, particularly through the integration of Artificial Intelligence (AI), machine learning, and data analytics in addressing global health challenges. Interdisciplinarity was a central theme of the workshop, underscoring the necessity of combining knowledge from various fields to develop holistic healthcare solutions. By fostering a collaborative environment through short talks followed by panel discussions, the workshop enabled participants to share insights and leverage their distinct expertise to highlight the multifaceted nature of maternal and child health research. The discussions aimed to bridge gaps between technology and clinical practice, inquiring into how innovations can be effectively implemented and beneficial in diverse healthcare settings. High-level topics covered during the workshop included the challenges and opportunities of applying AI in clinics and community health settings, the development of mobile health tools for maternal and child health, and strategies for improving data collection and usage in health systems. Discussions also explored the integration of AI models into health systems and the ethical considerations of deploying AI in low-resource settings. These topics highlighted the potential (and limitations) of advanced technologies to transform healthcare practices, improve patient outcomes, and reduce health inequities on a global scale. **Interdisciplinary questions** Given the interdisciplinary nature of the gathering, there were a number of questions that came out of the panel discussion that highlighted the value of collaborating across medicine, public health, AI, and human-computer interaction. From conception to data collection to design/development to implementation, we found that AI systems with meaningful and ethical impact required combined expertise. One major theme of discussion was community engagement. Far before building AI systems, there were key questions that needed to be answered, which required the expertise of not just researchers but also those who were not in the room—policymakers, health workers, and care recipients. For example, in many sessions, there was extensive discussion of local data ownership and sovereignty, alignment with community aspirations, and sustainability of tools within health systems. We posed questions such as how we can develop strong partnerships with local governments and other stakeholders, ensure that tools address relevant issues and do not exacerbate inequities, and create pathways to transition tools and data to local ownership. Community engagement also has implications for implementation—for example, we asked how can strong partnerships help integrate tools into health systems, particularly when in public health, it is programs that are implemented, not just tools? Methods in public health and HCI, such as community-based participatory research and co-design, could prove useful here. There were also a set of questions around equity that could clearly benefit from strong connections between clinicians, AI developers, and HCI researchers. There were overarching concerns such as how we can be mindful of the impact tools have on worker wellbeing or on patients’ experiences with an already complex health system. There were also technical concerns such as how can we ensure that predictions are actionable or how might we set particular thresholds to target treatments while centering health equity and effective management of resources? Finally, there were questions of limits and refusal, such as which data should be off limits for use in AI systems. Given that equity shows up within many stages of conceptualization, design, development, and implementation, it was clear that interdisciplinary collaboration and researchers developing a shared vocabulary and values was essential. Delving deeper into the technical, there were a set of questions around data collection and tool creation that came out of the discussion, but even here, it was clear that an understanding of AI needed to be paired with other expertise. For example, it was clear that there was potential for data-driven tools and AI to be used in care coordination, diagnostics, understanding maternal mental health, and outreach to high-risk populations. However, key questions such as what data to collect or outcomes to predict, how to work with the size and quality of datasets, or how to protect data privacy required collaboration between clinicians and AI and HCI researchers. --- Attachments- [ image workshop\_1\_crop.jpg ](/sites/g/files/omnuum6171/files/workshop_1_crop.jpg) --- See also:- [ Harvard news ](/news/harvard-news) Share on:- [ Facebook ](#) - [ Twitter ](#) - [ Linkedin ](#)