Abstract

Project Clara: AI Deployment in Healthcare

Although deep learning models have already demonstrated great capabilities when applied to medical image analysis tasks, their deployment and usage in clinical workflows is often inefficient and cumbersome. In this talk we discuss NVIDIA's solution to this issue: Project Clara. Clara is a platform built to facilitate deployment of DL algorithms in an efficient, scalable, and safe manner. We have designed Clara to tackle healthcare specific tasks and therefore ensure reliability and compatibility with current medical systems. 


Session No: SIL8128
Speaker: Eliron Amcir
Type:

AI in Healthcare

Date: Thursday - October 18, 2018 03:00 PM - 03:45 PM
Location: Hall M
Topic: AI in Healthcare
Industry: Healthcare & Life Sciences

Although deep learning models have already demonstrated great capabilities when applied to medical image analysis tasks, their deployment and usage in clinical workflows is often inefficient and cumbersome. In this talk we discuss NVIDIA's solution to this issue: Project Clara. Clara is a platform built to facilitate deployment of DL algorithms in an efficient, scalable, and safe manner. We have designed Clara to tackle healthcare specific tasks and therefore ensure reliability and compatibility with current medical systems. In a joint effort with Nuance, our partner in this development, we have built an AI marketplace powered by Clara which allows developers to make their algorithms accessible and doctors to process medical data in a simple and straightforward manner. Clara implements the "PLASTER" paradigm which sets the standard for programmability, latency, accuracy, size, throughput, energy efficiency and rate of learning of modern inference systems and therefore is the ideal tool to deploy healthcare oriented AI methods.