Removing the data security barriers and concerns to accelerate speech recognition usage in radiology

[London, 30 August 2022] – Augnito cloud-based speech recognition technology highlights why data security is a concern for radiologists and NHS Trusts.

While the healthcare sector and the NHS face increased pressure to adopt new technology, Scribetech (UK) Ltd, developers of Augnito, a clinical AI-powered speech recognition solution, reports widespread concerns over security and patient data.

According to Scribetech, security is a key barrier to adopting speech recognition – despite its potential to accelerate reporting time, drive accuracy, and enable radiologists to deal with a significant backlog.

Shiraz Austin, Managing Director at Scribetech (UK) Ltd and Co-Founder for Augnito, stated: “The security question is one we encounter a lot. The way radiologists work needs to evolve, but Trusts have understandable concerns about any new technology and how it handles sensitive patient data. That’s why security and compliance has been key to Augnito’s development from the very start.”

The most common concerns from Trusts around new technology cross three key areas:

1. The transfer of data to the cloud – specifically how data is protected against unauthorised access in storage and
2. How data is used to improve AI – and how information is anonymized and encrypted before it is fed into any machine
learning algorithm
3. Consistency in compliance – including where data is transferred during processing, particularly where this may
include multiple regulatory jurisdictions

Austin continued: “It’s not enough for cloud technology providers to simply say they are secure. We’ve taken proactive steps to embed security at every level of Augnito, so it’s a solution that Trusts – and radiologists – can really feel confident about using.”

Augnito is a next-generation speech recognition solution, available on desktops, mobile devices and web browsers, or integrated with existing clinical systems. This diverse range of application/s is supported by industry leading security and compliance, including:

– Data security, quality control, and management processes certified under DSPT, Cyber Essentials+, UKAS ISO 9001,
ISO 14001, BSI – ISO 27001
– A speech recognition engine hosted by Amazon Web Services (AWS) in the UK
– Anonymised, encrypted data at rest, stored exclusively in the UK and compliant with GDPR
– Automated backup options in AWS to prevent data loss

As a result, radiologists can adopt speech recognition without compromising on the safety of data and, ultimately, patients. However, Austin also reports measurable improvements to data integrity, made possible by Augnito: “Augnito keeps data safe as standard. But it also brings new levels of accuracy and consistency to reports, through both our 99.3% accurate AI and templates. Compared to manually typed reports, which are prone to errors and mistakes, Augnito is actually driving higher report quality. All while lessening the burden on radiologists and helping key diagnostic decisions be made sooner.”


Note to editors
About Scribetech and Augnito
Scribetech (UK) Ltd. is a clinical voice-AI solution and disruptive technology provider. Founded in 2001, the company has accumulated two decades of experience as an approved supplier of clinical transcription services to the NHS and private healthcare providers. This expertise has been combined with eight years of AI research and development partnerships established with engineers, analysts, and clinicians. The result is Augnito: an innovative, secure, cloud-based clinical speech recognition solution. Augnito brings seamless speech-enablement to daily workflows. All with 99.3% at-the-cursor accuracy, support for multiple medical specialities, and no need for voice profile training.

Visit for the full blog post relating to this media alert.  Contact [email protected] for more product details.

For further press-media information, interview opportunities or accompanying graphics please contact: Georgina Pavelin, Mixed Media – [email protected]