Adrian Florea is a patent agent in the firm’s IP Procurement and Portfolio Management practice group. Prior to joining the firm, Adrian served as a graduate research fellow at Columbia University, where he developed a stable diffusion machine learning model conditioned on discriminative medical BERT embeddings to generate realistic, prompt‑accurate heart murmur signals for downstream classification tasks. His research also included comprehensive reviews of emerging innovations in smart cardiac sensing and artificial intelligence for digital stethoscopes. In addition, Adrian integrated a noise‑robust audio representation model for heart sound segmentation and fine tuned an audio‑based large language model that achieved state‑of‑the‑art diagnostic accuracy in identifying rare, clinically significant acoustic heart murmur features.
This is a temporary biography. A complete biography will be available soon.
