This researcher is involved in two project areas using MSI:
- Parallel Simulations for Computer Architecture and Computer Aided Design:
For several emerging applications such as wearables, internet of things, and sensor networks, energy efficiency is of utmost importance. While custom ASICs have higher energy efficiency, general-purpose embedded processors are the preferred solution for many such applications due to the evolving nature of these applications and the high costs of custom IC design. The Sartori group discovers and exploits new opportunities for improving energy efficiency in general purpose embedded processors. They are focusing on new opportunities for energy efficiency enabled by detailed co-analysis of the design-level description of a processor and an application binary. Traditionally, co-analysis of the low-level hardware and details for a system has not been performed due to prohibitive costs. However, the group has developed automated analysis tools that perform unique analyses and expose new opportunities for energy efficiency.
The researchers have created a tool that identifies the parts of a processor that can never be exercised by a particular application. As such, they can identify paths in a processor that can never be exercised for a particular workload. This information can be used to enable several novel application-specific optimization and design techniques. The new techniques being created require detailed analysis of a system's hardware and software. This detailed analysis relies on high-throughput parallel simulation methodologies to be performed in a reasonable amount of time, necessitating the use of MSI computing resources.
- Machine Learning for Improving Healthcare:
The electronic medical record (EMR) is an integral part of our health care system designed to improve the quality and efficiency of medical care provided to patients. Despite its advantages in terms of improving patient care, the means of data entry into the EMR can actually detract from the care that patients receive from a medical provider. Clinical providers typically did not choose the field of medicine because they wanted to interact with a computer. Yet, multiple studies have shown that doctors are increasingly spending more time in front of a screen, navigating the EMR, than face to face with their patients. Frustration with the EMR is consistently cited as a contributor to the epidemic of physician burnout. The process of documenting the details of an encounter is tedious and time consuming, and when done in the patient room, often leads to doc-in-a-box lines of questioning, as providers attempt to fill out an electronic form in real time. Physicians who use medical scribes – assistants who listen to the conversation between patient and provider and translate into the required aspects of a clinical note in the EMR in real time – show greater job satisfaction, decreased burnout, and increased efficiency. This project aims to leverage the growing capability of speech recognition technology, artificial intelligence, and deep learning to create a virtual scribe that can automatically navigate and populate the fields in an EMR based on the conversation between a health care provider and patient. The researchers are also investigating the use of natural language processing and deep learning based on conversational data and structured EMR data to perform clinical decision support that aids physicians in critical care decisions to improve the quality of care received by patients. HPC resources are used to process massive medical record datasets and create machine learning techniques to classify EMR data.