Neural Decoding of Smell with Generative AI Masters of Science in Data Science Thesis Proposal
School of Professional Studies, Northwestern University
Abstract
Olfaction (sense of smell) and related structural abnormalities in the olfactory bulbs are among the first observed symptoms corelated with Alzhiemer’s disease (Esiri and Wilcock 1984; Thomann et al. 2009). The functional relationship between olfaction and the olfactory bulbs is poorly understood (Weiss et al. 2020). In the area of visual decoding, advances in machine learning using generative adversarial networks (GANs) have provided an unprecedented insight into neural decoding or the mapping of an individual brain’s responses and performance (Seeliger et al.
Generating SOAP notes from audio using Microsoft Azure Abstract
SOAPgen is a healthcare AI and LLM project focused on converting recorded clinical conversations into structured SOAP notes: Subjective, Objective, Assessment, and Plan. The goal is to reduce manual documentation time for clinicians and make patient records easier to organize, review, and share.
The pipeline uses Microsoft Azure services for speech processing and downstream text handling, followed by an LLM to organize the content into the standard SOAP format.
Partial image credit Why Go? Static site generators such as Hugo were developed recently using Go. Go is a compiled language designed for today’s multi-processor, scalable, high-performance systems. Other platforms have inherited dependency chains that can lead to infeasible build times. Realizing these benefits, Cloudfare recently migrated from Gatsby (a web framework based on JavaScript).
Demo website I first created a demo website using the Hugo Winston theme. This theme does have a Live Demo which made it simple to deploy onto Netlify.
