Canadian Healthcare Technology Logo
  • Issues
    • Current Print Issue
    • Print Archive
  • Advertise
    • Publishing Schedule
    • Circulation
    • Unit Sizes and Rates
    • Mechanical Requirements
    • Electronic Advertising
    • White Papers
  • Subscribe
    • Print Edition
    • e-Messenger
    • White Papers
  • Events
  • Vendors
  • About Us

GE Revolution Ascend

GE Revolution Ascend

Enovacom EPC

Enovacom EPC

Imaging News

Not as much data needed as thought for AI

January 15, 2019


Charles Kahn 2018CHICAGO – Deep Learning has tremendous potential for radiology, but the roadblock at the moment is a shortage of quality data to build the applications, says Dr. Charles E. Kahn (pictured), professor and vice chair of radiology, at the University of Pennsylvania’s Perelman School of Medicine in Philadelphia.

Dr. Kahn is also editor of the new RSNA online journal, Radiology: Artificial Intelligence. He spoke at a session at the annual RSNA meeting in Chicago, which was held last November.

“Most people who have done work in this area have discovered that about 70 to 80 percent of the work that you do is not building the model or testing it,” he said. “It’s curating, cleaning and massaging the data to get it into shape.”

However, new research has reached startling conclusions about the optimal number of images needed to train an algorithm.

A study in Radiology that looked at the automated classification of chest radiographs found that the Deep Learning model’s accuracy improved significantly when the number of images used to train the algorithm jumped from 2,000 to 20,000. But accuracy improved only marginally when the number of training images increased from 20,000 to 200,000.

“That’s actually a useful thing, that maybe we don’t need to have millions of images in order to train the system,” Dr. Kahn said. “Maybe having a modest number would be a good start, along with other approaches that you could perhaps superimpose on top of that.”

PreviousNext

SteraMist (Feb)

SteraMist (Feb)

News and Trends

  • RACE streamlines patient journey
  • Healthcare supply chain needs a re-think, observers say
  • EDI spots pricing anomalies in Ontario’s healthcare supply chain
  • AI centres of excellence and companies collaborate on apps
  • Talking Stick: New hope for Indigenous mental healthcare
More from the Print Edition

Subscribe

Subscribe

Free of charge to Canadian hospital managers and executives in nursing homes and home-care organizations. Learn More

Follow us on Social Media!

Follow us on Social Media!

Nihi Data [Winter 2023]

Nihi Data [Winter 2023]

WP

WP

Advertise with us

Advertise with us

Sectra One Cloud

Sectra One Cloud

Change Healthcare [2]

Change Healthcare [2]

Infoway [Feb2023]

Infoway [Feb2023]

Zebra

Zebra

CHT print-200×400

CHT print-200x400

SteraMist (Feb)

SteraMist (Feb)

Advertise with us

Advertise with us

Sectra One Cloud

Sectra One Cloud

Change Healthcare [2]

Change Healthcare [2]

Infoway [Feb2023]

Infoway [Feb2023]

Zebra

Zebra

CHT print-200×400

CHT print-200x400

Contact Us

Canadian Healthcare Technology
1118 Centre Street, Suite 207
Thornhill, Ontario, Canada L4J 7R9
Tel: 905-709-2330
Fax: 905-709-2258
info2@canhealth.com

  • Quick Links
    • Current Print Issue
    • Print Archive
    • Events
    • Vendors
    • About Us
  • Advertise
    • Publishing Schedule
    • Circulation
    • Unit Sizes and Rates
    • Mechanical Requirements
    • Electronic Advertising
    • White Papers
  • Subscribe
    • Print Edition
    • e-Messenger
    • White Papers
  • Resources
    • White Papers
    • Writers’ Guidelines
    • Privacy Policy
  • Topics
    • Administrative Solutions
    • Clinical Solutions
    • Companies
    • Continuing Care
    • Diagnostics
    • Education & Training
  •  
    • Electronic Records
    • Government & Policy
    • Infrastructure
    • Innovation
    • People
    • Privacy and Security

© 2023 Canadian Healthcare Technology

The content of Canadian Healthcare Technology is subject to copyright. Reproduction in whole or in part without prior written permission is strictly prohibited. Send all requests for permission to Jerry Zeidenberg, Publisher.

Search Site

Error: Enter a search term

  • Issues
    • Current Print Issue
    • Print Archive
  • Advertise
    • Publishing Schedule
    • Circulation
    • Unit Sizes and Rates
    • Mechanical Requirements
    • Electronic Advertising
    • White Papers
  • Subscribe
    • Print Edition
    • e-Messenger
    • White Papers
  • Events
  • Vendors
  • About Us