Manual Fine Needle Aspiration (FNA) biopsy is often used to diagnose breast cancer in Sub-Saharan Africa (SSA). While performed with ultrasound guidance in the United States, most SSA healthcare centers lack imaging resources, rendering clinicians unable to consistently sample breast lesions with adequate cellularity, leading to high false-negative rates, repeat patient visits, increased financial burden, and delayed treatment initiation. Our solution, an automated handheld device, has shown promise in improving sample cellularity as compared to manual FNA.

Our project is a new boardwalk design for the Stream Trail in Gwynns Falls Leakin Park. It will replace an existing strip of boardwalk that has deteriorated over time due to washouts and tree damage. In collaboration with FOGFLP the boardwalk would be re-built to the design specifications provided by JHU students using materials provided by BCRP and volunteers associated with FOGFLP. This L-shaped boardwalk design is elevated on pressure-treated wood posts to account for the small ravine it will span, which has caused wet soil conditions and occasional flooding in the past.

Our device is a single-microneedle attachment that administers liquid nitrogen in a top-down manner. By delivering the cryogen 2-3 mm below the surface of the skin, this novel top-down intralesional cryogen administration allows for keloid tissue remodeling below the surface, resulting in a more effective cryotherapy treatment while still reducing additional skin trauma.

Neurological disorders impact a large percentage of the global population and are a vast area of research in the clinical field. However, state of the art diagnostic measures such as MRI and CT scans are invasive and expensive. Saccades, rapid fixations in eye movements, are a promising but underutilized non-invasive biomarker for neurological abnormalities due to limited publicly available data and privacy concerns. To address this, we developed a pose-guided video generation model that produces synthetic saccades of three types: normal, bilateral hypermetria, and bilateral hypometria that mimic real eye movement patterns observed in clinical settings. We trained an MViT-V2 video classification model on the synthetic data as a baseline and tested its performance on clinical saccade data. Our approach demonstrates the potential of synthetic data to enable accurate and scalable saccade-based diagnostics, reducing the dependency on invasive imaging.