Page banner

#TestingMethods2020

Submitted
Being Assessed
Responded
Thanks for your vote
This voting milestone has now ended, and the idea has progressed to a private milestone
Thanks for approving the idea
This milestone has now ended, and the idea has progressed to a private milestone

AVPaT - Automated Virus Pathology (Dry) Testing in the Field using Artificial Intelligence neural networks and nano microscope

B: One to watch

The AVPaT is a dry testing device approximately the size of a desktop printer, making it ideal for use in the field (Hospital wards, GPs etc). It combines Artificial Intelligence and a Nano microscope to provide rapid testing (results in under 5 seconds) for antigens or antibodies. AVPaT can tell operators whether the individual is infected, or has been, and with that what stage of the infection they are at, or the level of immunity they have. It has a 98+% specificity for antigen testing, and 99+% specificity for antibodies.

The AVPaT system also has uses past COVID-19. It is able to test for other pathogens and will be able to detect the development of unknown/new pathogens that could lead to further pandemics. There is also potential for use in the NHS for rapid on site diagnostics for an increasing numbers of ailments.

Advantages
Moving from Wet Chemistry to solid state imaging means no more consumables (save for the test slide) or logistics, making testing cheaper and quicker. Results take less than 5 seconds to complete rather than hours, and are displayed for the operator on the display, telling them if specific antigens or antibodies are detected, and stage of the infection or level of immunity.

Fully digitised data recording of results allows for analytical analysis of samples by the AI, enabling the early detection of new viruses and mutations before outbreaks occur.

System Features

  • The AVPaT features an auto slide insertion process
  • Less than 5 seconds from scan entry to results.
  • WiFi updates to enable it to test for an ever growing range of infections and illnesses
  • Anomaly detection for surveillance early warning (unknown cell types) 

  • Integrated data structuring for epidemiology analytics 

  • Nanoscale multispectral imaging and AI recognition @ 25nm 

  • Automated operator display depicting antigen or antibody detection, and confidence level of results.
  • Degree of infection (cell count)
  • Stage of immune response (immune cell type) 

  • Degree of immunity (antibody quantities) 

  • Supports multiple types of sample slides i.e. blood, mucus, saliva etc. (other less intrusive samples are being looked at for development)

How quickly could this be deployed and what are the dependencies?

Delivery on a mass scale is expected to be possible by late July / early August. The dependencies will be on the supply and manufacture of the key components and the delivery times of them and the finished article.

What is the likely production volume?

An enquiry for tens of thousands of the devices has been received from America and further mass orders should pose no problem providing the key scientific components can be provided.But we anticipate no issues.

What are the risks and barriers to using this at scale?

Using at scale should only benefit the device given the machine learning capabilities of the AI neural network. The main barrier to scale will be ensuring the production of slides and key components for the device, though this is viewed as a minimal risk given the parties involved. Additionally it will need a power source and to be connected to the database (local or central) that enables the AI.

Who are you already partnering with on this?

This is a partnership between iAbra, Dell and Intel, with HISTATE representing them in the UK.

edited on Apr 24, 2020 by HISTATE
HISTATE

Attach files

@

Bev Matthews 8 months ago

The idea has been progressed to the next milestone.

Reply 0

Bev Matthews 8 months ago

Status label added: B

Reply 0

Bev Matthews 8 months ago

The idea has been progressed to the next milestone.

Reply 0

Bev Matthews 8 months ago

Thank you for your submission, which has now been reviewed. We are interested to hear back from you nearer the time you have set out above please. Bev

Reply 0

HISTATE 8 months ago

Thank you Bev, once we have a few more prototypes we will be in touch. What is the best way to contact you?
J Bassett
HISTATE

Reply 0

Bev Matthews 8 months ago

Thank you! Please reply on here. Thanks Bev

Reply 0

HISTATE 8 months ago

There are some forms of prototype that you can see now, but given commercial sensitivities we would appreciate further discussions about them directly rather than by public forum

Reply 0

Tom Jordan 6 months ago

Thanks for this HISTATE and apologies to not have gotten back to you on your previous query. I'd be keen to hear on progress now we are into July.

We have just added a feature to privately share updates on the platform in the "Authors and moderators" comments tab which you can click just above where you enter comments. Only you and we (the admin / review team) will see information you submit here.

If it helps to clarify anything on this do let me know.

Tom
Testing Methods team

Reply 0

Share