Sepsis is a life-threatening condition characterized by a dysregulated host response to infection. Early prediction is crucial for improving outcomes but remains a challenge using traditional models. Leveraging data from 11,325 adult ICU patients in the MIMIC-III dataset, we developed a transformer-based model to predict sepsis onset and outcomes using clinical features and high-frequency physiological waveforms (e.g., ECG, PPG, ABP). Clinical and waveform data were tokenized into learnable embeddings, enabling the model to capture temporal dynamics through cross-attention mechanisms. Performance evaluation showed strong predictive ability at shorter lead times, with reduced accuracy at longer horizons, highlighting the need for richer input features. To address this, we built a waveform tokenization pipeline to distill informative segments from raw signals. Token importance analysis revealed distinct physiological and clinical markers associated with sepsis. Future work will focus on fusing clinical and waveform embeddings to improve accuracy and interpretability for early, actionable sepsis prediction in critical care.

As institutions of higher education become increasingly conscious of their carbon footprints, understanding the emissions generated by professional activities is essential. This project investigates the environmental impact of large academic events by analyzing attendance at conferences organized by the Society for Industrial and Applied Mathematics. Using conference program data, we identified speakers and attendees, determined their home institutions and locations, and used this information to generate heat maps illustrating attendee counts and travel-related emissions. An interactive dashboard presents this data in an accessible format, allowing users to select from three analyzed conferences and view attendee numbers, emissions, and travel distances for participating institutions on a world map.

Characterizing soft tissue mechanics is of interest across biomedical disciplines, including in regenerative medicine and prosthetics development. In the context of skin cancer, there is a well-elucidated relationship with tissue stiffness; however, mechanical properties of skin are not factored into skin cancer screening and diagnostics because of a lack of established benchmark values for skin stiffness. Existing technologies for soft tissue mechanical characterization are ill-suited for assessing the elastic modulus of potential skin cancers—they are ex vivo, do not provide the resolution needed for skin lesions, and/or are not sufficiently maneuverable for in situ use. Here is reported a proof-of-concept for a novel device for the measurement of skin mechanics in situ, which is handheld, portable, and operates on the scale of skin lesions. Testing on polymers shows the device is able to reliably differentiate between materials that have elastic moduli in the range of human epidermal/dermal tissue.

PurrCare is a smart cat health monitoring gadget for busy tech-savvy cat owners who prefer to feel safe from a distance when they’re not around. The smart collar and smartphone app monitor physical movement, vital signs, and moods of a cat in real time. With remote voice control and health reminders added, PurrCare allows owners to be proactively involved in nursing their pets—over distance. It is particularly appealing to city workers who already wear health monitors on themselves and will pay for top-of-the-line preventive care on their pets.