• No results found

Radiology without Walls:

N/A
N/A
Protected

Academic year: 2021

Share "Radiology without Walls:"

Copied!
34
0
0

Loading.... (view fulltext now)

Full text

(1)

Radiology without Walls:

Building the Practice of the Future

Se s s ion # 94, Augus t 11, 20 21

VP, Product Management - Radiology Worldwide Healthcare and Life Sciences Physician Lead

GE Healthcare Amazon Web Services

Timot hy Ros e Lynds e y Burt on, MD

(2)

#HIMSS21 2

Welcome

Worldwide Healthcare and Life Sciences Physician Lead

Amazon Web Services

Lynds e y Burt on, MD

Vice President, Product Management, Radiology GE Healthcare

Timot hy P. Ros e

(3)

Agenda

• Current Challenges, Industry Dynamics and Market Trends

• Key Enablers for Creating the Radiology Practice of the Future

• What’s Next for the Radiology

Practice of the Future

(4)

Learning Objectives

• Discuss the current challenges and industry dynamics facing radiology organizations

• Recognize essential elements for

creating the radiology practice of the future

• Evaluate the opportunity for cloud based diagnostic radiology and workflow

solutions

• Hypothesize the future of radiology in the

cloud

(5)

The following content in this presentation is for illustration and demonstration purposes only.

General Electric reserves the right to make changes in specifications and features , or discontinue the product or service described at any time, without notice or obligation. This does not constitute a

representation or warranty or documentation regarding the product or service featured. Illustrations are provided for informational purposes, and your configuration may differ.

General areas represent ongoing product research and development efforts. These research and development efforts are not products and may never become products. Not for sale. Not cleared or approved by the FDA or any regulatory body for commercial availability.

Any descriptions of future functionality reflect current product direction, are for informational purposes only and do not constitute a commitment to provide specific functionality. Timing and availability are subject to change and applicable regulatory approvals.

GE, the GE Monogram, Centricity and Imagination at Work are trademarks of General Electric Company.

©2021 General Electric Company

(6)

6

Current Challenges

Industry Dynamics

and Market Trends

(7)

Productivity challenge:

Evolution of imaging models

Radiology as a silo Connected Radiology Radiology is a network

Spe cia lis t Clinic

READING GROUP

Pa t ie nt s Aca de mic Ima ging

X-hos pit a l s e rvice s

Aca de mic Communit y

Priva t e ima ging Spe cia lis t

(8)

8

Traditional Data

NUMBERS

5

KB/RECORD

TEXT

500

KB/RECORD

2D IMAGES

1,500

KB/IMAGE

ULTRASOUND

20,000

KB/WAVEFORM

3D IMAGES

35,000

KB/STUDY

VIDEO

5,000,000

KB/MOVIE

HI-RES

202,375,168

KB/GENOME

50 Petabytes of data per hospital

1

1,2Source: IDC & EMC Study -https://www.cycloneinteractive.com/cyclone/assets/File/digital-universe-healthcare-vertical-report-ar.pdf

3Source: https://icitech.org/wp-content/uploads/2016/01/ICIT-Brief-Hacking-Healthcare-IT-in-2016.pdf

New Data

GROWTH AT

48%

PER YEAR2

of healthcare systems have

81%

been compromised by one or more cyberattacks

within the last year3

(9)

CLOUD ENABLED FLEXIBLE HOSTING

Market Trends and Priorities Drive New Requirements

Cloud will Improve access and increases speed of adoption of new technologies Precision Health

Diagnostics Care Pathways Complexity ↑ Demand ↑ vs.

shortage of rads Mergers ↑ and

Regional projects

Market Trends

Enterprise Imaging Directions

Customer Priorities

Modernized Clinician Viewing

AI enabled· platform agnostic server rendering· new advanced tools

Streamlined Architecture

Relevant data aggregation· holistic patient record· one store for all· cloud ready

Enhanced Smart Workflows

Structured reporting· PACS driven· X enterprise· beyond boundaries

Productivity, Efficiency &

Resource allocation Access to patient data

across systems Aiding in Definitive

Differential Diagnosis

(10)

10 10

Creating the Radiology

Practice of the

Future

(11)

Key Enablers

Intelligent Workflows Native AI Integration Cloud Enabled

(12)

#HIMSS21

Why employ AI in Radiology workflows?

• AI can ‘intelligently’ automate manual tasks, freeing time for the radiologist to focus on supporting informed decision-making and personalized care.

• AI can help provide decision support to radiologists – notifying clinician of condition and providing some sort of intelligence to back

that up. Both criticality of read (what exam¹ to read next) as well as information to support a more narrow and

accurate differential diagnosis.

• AI can help detect patterns in image data that are too subtle or complex for humans

to easily see.

JB72396XX 12

¹ Triaging is based on the regulatory approved intended use of the AI algorithm

(13)

• Implementing AI tools effectively is complex and difficult - resulting in slowed AI adoption

• Integrating multiple vendors/algorithms is risky and could lead to errors

• Using AI clinical apps is a manual & disjointed process

• Radiologists/organizations need a way to experiment with AI algorithms in their

workflows

• Prefer to acquire AI enablement from a single trusted and known vendor

Challenges in Utilizing AI Applications in Imaging

Sources: Quantitative Market Research, MarketVision Research, PACS AI, VNA AI, and AI

(14)

#HIMSS21

Enable AI Adoption

14

Integrate

Enable

Grow

(15)

Consuming AI Results

Seamless and uniform UX for AI applications

CADSR Or Secondary Capture

Viewer Worklist Support

Exam Prioritization

Diagnostic Hub

Reported Results

(16)

#HIMSS21

Why Cloud?

16

Cos t Sa vings (TCO)

St a ff Product ivit y

Ope ra t iona l Re s ilie nce

Bus ine s s Agilit y

Typical

Focus Most Compelling

Cloud Benefits

TCO: Total cost of ownership, SLA: Service-level agreement, AZ: Availability Zones, DR: Disaster Recovery

(17)

Ways to Work Differently in the Cloud

Containerization, HA and scalability

Security

Enabled as a service

Deployment

Log management and monitoring

Managed services

(18)

#HIMSS21 18

Software deployment / application configurations

On premise Installation ~ 24hrs

Single tenant Deployment & config

<8hrs

Multi tenant

Deployment & config

<1hrs

Trained for HW and SW

Working and helps remotely (Customer & FE)

Huge SW Troubleshooting knowledge

Available regionally, less multiplicity (<100)

Technical Installation Specialist

Trained, mostly enterprise engineers with HW skills

May travel onsite, installs HW and SW

Small SW Troubleshooting knowledge

Thousands of FEs worldwide

Online Engineer

Expert cloud knowledge, deployment

Working remotely

Fully automated deployment and configuration

Cloud Engineer

(19)

As the world moves towards a more virtualized and distributed care delivery model with home

care, remote patient management, and the increased use of AI to

support radiologists and other

clinicians, GE Healthcare and AWS are working to make healthcare

more efficient and personalized.

(20)

#HIMSS21 20

• Over 1 million active customers across 190 countries

• 2,000+ government agencies

• 5,000+ educational institutions

• 25 regions (+7 Planned)

• 81 availability zones with 3+ data centers per zone

• 245 countries and territories served

Delivered securely and reliably across a

global infrastructure

(21)

There’s no compression algorithm for experience

Years of experience, on average, for our team leaders

in the healthcare and life sciences industry

15+ 8+

Years with dedicated healthcare and life sciences

cloud technology practice

18+

Years as the world’s most comprehensive and broadly

adopted cloud platform

1,0 0 0 ,0 0 0 + 2,0 0 0 + 5,0 0 0 +

Educational institutions Government agencies

Active customers

(22)

#HIMSS21

of customers have used machine learning on AWS

of TensorFlow projects in the cloud run on AWS

85%

of deep learning in the cloud runs on AWS

81%

AWS holds the top spots on Stanford’s deep learning benchmark, DAWN, for

fastest training time, lowest cost, lowest inference

latency

More machine learning happens on AWS than anywhere else

22

(23)

Purpose-built HIPAA-eligible services Solutions for everyone from expert practitioners to application developers

AI services, fully managed ML services, and support for popular frameworks

Broa de s t a nd de e pe s t s e t of AI a nd ML s e rvice s

Single IDE for the entire ML workflow At least 54% lower TCO

Up to 10 times improvement in productivity

Acce le ra t e your a dopt ion of ML wit h

Sa ge Ma ke r

Built on t he mos t compre he ns ive

cloud pla t form

Highly secure, reliable, fully featured data store

The strongest set of compute, storage, security, database, and analytics

capabilities to build upon

85% of TensorFlow project in the cloud runs on AWS

Why AWS for ma chine le a rning?

(24)

#HIMSS21

Security infrastructure built to satisfy healthcare, finance,

and other high- sensitivity organizations Benefit from AWS

industry leading security teams 24/7,

365 days a year Over 50 global

compliance certifications and

accreditations, including HIPAA and

HITRUST

Leverage security enhancements from

1M+ customer experiences

Priva cy, re lia b ilit y a nd s e curit y is a t our core

24

(25)

Introducing Edison True PACS¹

Edison™ True PACS is a modern PACS solution helping to support intelligent workflow

automation and AI-based decision support for increased productivity and diagnostic accuracy

www. ge he a lt hca re .com/ wit hout wa lls

(26)

#HIMSS21

Platform deployment automation

GE Healthcare’s approach with Edison™ True PACS

CloudFormation²/TerraForm:

Infrastructure as code framework

Provides repeatable infra & software deployment

Maintains the platform desired state

One tool to deploy the platform and software

What CloudFormation/TerraForm builds:

Networking: Subnet, security groups, load balancers, routes

Storages: FSx (SMB) and S3

Active Directory cluster

EC2 instances, (virtual machines), containers and kickstarts software install and configuration

Stacks:

Automation templates are called stacks. Stacks contain definitions for the

infrastructure components and can be aggregated together to form the targeted infrastructure

26

CloudFormation Framework Environment 1 (common services)

customer 2

Svcs VMs Support VMs VPN VMs Streaming VMs DB Server Archive VMs

Storage (FSx) Directory Services

Network

Console VMs Viewer VMs

Storage (S3) Network

customer 1

Console VMs Viewer VMs

Storage (S3) Network

¹ Edison™ True PACS is a solution which contains Universal Viewer, Enterprise Archive, Centricity Universal Viewer ZeroFootprint, Open AI Orchestrator and 3rdparty AI applications

² Edison™ True PACS cloud solution is only available in USA

(27)

Platform

deployment automation

Example: GE Healthcare leverages Amazon Web Services to support automation

(28)

28

What’s Next for

the Radiology

Practice of the

Future?

(29)

Enhancing Advanced Visualization

Supporting a Definitive Differential Diagnosis Aiding in Speed

and Accuracy

(30)

#HIMSS21

GE Healthcare’s Vision

Grant rapid access to advanced

visualization applications provided by a horizontally scalable, geographically distributed, multi-tenant, SaaS solution, which is quickly releasable and

represents an attractive reduction in Total Cost-of-Ownership when

compared to an on-premise solution.

¹ Technology in development that represents ongoing research and development efforts. These technologies are not products and may never become products.

Not for sale. Not cleared or approved by the U.S. FDA or any other global regulator for commercial availability.

(31)

AWS Region closest to customer

AW Web Access VPC

Availability Zone Availability Zone

X11 applications 1 APP / 1 POD

POD POD POD POD POD POD POD POD POD POD

IF applications 1 APP / n PODs

Edison services 1 APP / N PODs

POD POD POD POD POD POD

AW AW AW AW AW AW AW AW AW AW AW AW

Hybrid approach

(VM & Containers)

• Multi tenant

• EKS based

(32)

#HIMSS21 32

Bottom Line:

Cloud Enabled

Radiology Breaks Down Walls

www. ge he a lt hca re .com/ wit hout wa lls

(33)

Questions

(34)

#HIMSS21

Thank You

34

Vice President, Product Management, Radiology linkedin.com/in/tim-rose-2829878a

Timot hy P. Ros e

Worldwide Healthcare and Life Sciences Physician Lead linkedin.com/in/lyndseyburtonmd

Lynds e y Burt on, MD

Sca n t o s che dule a me e t ing or de mo wit h GE He a lt hca re

Or vis it Boot h 3832

www. ge he a lt hca re .com/ wit hout wa lls

References

Related documents

Though vaccination is a most successful public health approach in protecting and controlling HPV infection [14], for successful implementation of the vaccine, it is

La España sesentista está dominada por el Barroco, cuyo rasgo predominante es la presencia de un conflicto de fondo o de una oposición de principios que

In our experiment a significant correlation was found between the self-identification scoring and the P200 component related with self-recognition mechanisms: the more the

Although an equal amount of maxi- mum force exertion throughout the workspace cannot be achieved (unless the manipulator is not a PP type of mechanism), an optimization criteria

Broj ispitanika koji gotovo nikada nije smatrao da im je prehrana nezadovoljavajuća zbog problema sa zubima, ustima, čeljustima ili zbog protetskog rada je 78 (22,1%),

En particulier, Nippold et ses collaborateurs (1988) constatent que des enfants présentant des troubles du langage ont des performances inférieures à celles

Like most programming languages, Haskell has library facilities to imple- ment pseudo-random number generators that can be used to generate random numbers; see the standard