iTnews
  • Home
  • News
  • Technology
  • Software

Macquarie Uni races tech giants to build AI medical scribe

By Matt Johnston on Dec 9, 2019 1:40PM
Macquarie Uni races tech giants to build AI medical scribe

Calls for data to beat out Google, Microsoft.

A team from Macquarie University has set out to build its own version of an artificial intelligence-powered scribe in a race against global tech giants like Google, Microsoft and Amazon.

Researchers from the Australian Institute of Health Innovation (AIHI) at Macquarie are currently hunting for data to build an automated speech-to-text capability that can accurately record medical jargon for integration with health records.

While AI scribes have become much more popular and powerful in recent times, they’re still far from perfect (often with hilarious or frustrating results).

And although AIHI didn't delve too deep into it, the security of consultation recordings could be a major concern for doctors and patients alike, especially when processed by large multinationals known for harvesting data to sell to unknown third parties.

In a paper recently published in the journal Digital Medicine, the researchers said that apart from the sensitivity of the data, there are five main challenges in producing an effective scribe.

One of the main problems is that AI technology needs a lot of data on which it can be trained, which is “extremely hard to come by,” Dr Juan Quiroz, one of the researchers working on the project, told Macquarie’s Lighthouse publication.

“We need recordings of tens of thousands of conversations between patients and doctors so the AI can learn patterns from the data.

“Doctors and clinicians rarely record their consultations,” he said, noting that asking them to do so is yet another imposition for professionals already operating with a late on their plate.

Even if a clinician does record a consultation, many AIs struggle to transcribe recordings with multiple voices, accents or ambient noise.

Training the AIs is no easy task either due to the unstructured nature of many patient consultations and taking down notes to compare human and AI outcomes takes a great deal of time and clinical expertise.

“The currently closed environment for sharing sensitive data means that the few research teams with access to data are the only ones that can make advances, further slowing progress by impeding open science,” the researchers said.

Nevertheless, a localised AI transcriber could prove incredibly useful at mitigating the current adverse effects of clinical documentation, which has been associated with clinician burnout, information loss, distractions during consultation and increased cognitive load.

Got a news tip for our journalists? Share it with us anonymously here.
Copyright © iTnews.com.au . All rights reserved.
Tags:
aiamazongooglehealthhealthcarehealthcareitmacquarie universitymicrosoftsoftwareustralian institute of health innovation

Partner Content

Avoiding CAPEX by making on-premise IT more cloud-like
Promoted Content Avoiding CAPEX by making on-premise IT more cloud-like
Why Genworth Australia embraced low-code software development
Promoted Content Why Genworth Australia embraced low-code software development
How to turn digital complexity into competitive advantage
Promoted Content How to turn digital complexity into competitive advantage
Security: Understanding the fundamentals of governance, risk & compliance
Promoted Content Security: Understanding the fundamentals of governance, risk & compliance

Sponsored Whitepapers

Extracting the value of data using Unified Observability
Extracting the value of data using Unified Observability
Planning before the breach: You can’t protect what you can’t see
Planning before the breach: You can’t protect what you can’t see
Beyond FTP: Securing and Managing File Transfers
Beyond FTP: Securing and Managing File Transfers
NextGen Security Operations: A Roadmap for the Future
NextGen Security Operations: A Roadmap for the Future
Video: Watch Juniper talk about its Aston Martin partnership
Video: Watch Juniper talk about its Aston Martin partnership

Events

  • CRN Channel Meets: CyberSecurity Live Event
  • IoT Insights: Secure By Design for manufacturing
  • Cyber Security for Government Summit
By Matt Johnston
Dec 9 2019
1:40PM
0 Comments

Related Articles

  • CSIRO uses AI to crunch a trillion genomic data points
  • Microsoft stops selling emotion-reading tech, limits face recognition
  • UK financial regulators to directly oversee cloud services
  • Edtech vendors invaded student privacy: Human Rights Watch
Share on Twitter Share on Facebook Share on LinkedIn Share on Whatsapp Email A Friend

Most Read Articles

Qantas calls time on IBM, Fujitsu in tech modernisation

Qantas calls time on IBM, Fujitsu in tech modernisation

Service NSW hits digital services goal two years early

Service NSW hits digital services goal two years early

SA Police ignores Adelaide council plea for facial recognition ban on CCTV

SA Police ignores Adelaide council plea for facial recognition ban on CCTV

NBN Co says TPG tie-up could help Telstra sidestep spectrum limits

NBN Co says TPG tie-up could help Telstra sidestep spectrum limits

Digital Nation

Crypto experts optimistic about future of Bitcoin: Block
Crypto experts optimistic about future of Bitcoin: Block
The security threat of quantum computing
The security threat of quantum computing
Integrity, ethics and board decisions in the digital age
Integrity, ethics and board decisions in the digital age
IBM global chief data officer on the rise of the number crunchers
IBM global chief data officer on the rise of the number crunchers
COVER STORY: Operationalising net zero through the power of IoT
COVER STORY: Operationalising net zero through the power of IoT
All rights reserved. This material may not be published, broadcast, rewritten or redistributed in any form without prior authorisation.
Your use of this website constitutes acceptance of nextmedia's Privacy Policy and Terms & Conditions.