
MagnifEye allows standardisation of test reporting, thereby increasing confidence in LFD testing, reducing user error and increasing overall accuracy of an AI LFD programme.1
AI-powered reading and analysis of diagnostic tests
The COVID-19 pandemic has increased the need for accurate reading of point-of-care diagnostic tests for clinical decision support and disease surveillance across large populations.
MagnifyEye utilises Sensyne’s expertise in developing novel deep learning approaches and AI algorithms to automate and improve the accuracy of reading diagnostic tests across a wide range of indications including COVID-19.
MagnifEye API
A deep learning, cloud-based API that can read lateral flow test results beyond the human visible spectrum.
MagnifEye COVID app
Assists users and health professionals to perform diagnostic tests and share results. Available as an Android, iOS or web application.
Test to Go
Powered by MagnifEye
Test To Go by Excalibur Healthcare Services reads Excalibur Rapid SARS-CoV-2 Antigen Test results rapidly, reliably and accurately.
Potential benefits
Improves reading accuracy and consistency
Reduces the variability, subjectivity and possible error associated with manual, unassisted image reading.
Epidemiological insight
Analysis of large image datasets may help inform disease and public health management.
Speed and flexibility
Results are returned in under 2 seconds, and available for analysis immediately, improving user adoption and supporting decision making.
Secure
Data is encrypted and securely held in the cloud with robust cyber security.
A proven tested technology
NHS Digital conducted a pilot study in April 2021 that tested MagnifEye on over 100,000 real world cases. The results of the pilot study1 show:
MagnifEye increased sensitivity (the ability to identify positive tests correctly) to 97.6% and specificity (the ability to identify negative tests correctly) to 99.99% in reading lateral flow tests, as compared to a human reader.
Results
A significant proportion of positive cases that were previously missed by users or trained lateral flow test readers were detected by MagnifEye. At Assisted Testing Sites (ATSs) the algorithm identified 24.4% of true positive tests that were missed by trained operators.
Ongoing use of this algorithm is anticipated to support the Government in identifying positive cases with low viral loads which might otherwise be missed via human read and improve objective reading of tests by supporting the visually impaired or anxious in interpretation of their COVID lateral flow results.
MagnifEye is aimed at
Diagnostic test providers and manufacturers requiring faster, more accurate reporting of test results.
Public health agencies requiring rapid, accurate test reporting and analysis for epidemiological insights.
MagnifEye in the news
FAQs
MagnifEye is a new smartphone software application that uses deep learning to automate the accurate reading, beyond the human visible spectrum, of diagnostic tests including lateral flow tests.
MagnifEye draws on research & development at Sensyne over the past two years in applying Artificial Intelligence to the analysis of medical images and clinical data across large patient populations.
MagnifEye is available as a complete app or alternatively, the photo capture and/or image processing algorithm can be provided as an SDK (Software Development Kit) or as an API (Application Programming Interface) for use on both Android or iOS phones, to be incorporated into existing apps.
The interpretation of diagnostic tests often relies on the human eye and is therefore limited by the human visible spectrum, is subjective in nature, and can be open to error, misinterpretation or fraud. In the case of lateral flow tests, dedicated readers can overcome some of these limitations but they limit the scale and speed of lateral flow testing, and can increase costs i.e. for mass, self-administered COVID-19 lateral flow tests it is not practical to use readers.
MagnifEye offers a solution to these limitations by utilising a smartphone camera to take a photo of the test which is then processed using a cloud-based deep learning algorithm to accurately interpret the result within a few seconds in accordance with the manufacturer’s indications.
This means that no additional hardware is needed other than a standard smartphone. Tests can be read at home, or at the point of care without the need to be returned to a central point and results immediately transferred to a secure data storage for aggregated data analysis.
MagnifEye is a business-to-business commercial product currently being offered to lateral flow test manufacturers and healthcare providers that use lateral flow tests for disease surveillance across large populations, including in international markets.
For lateral flow test manufacturers, MagnifEye allows a seamless connection between test and reader that can potentially improve the quality and accuracy of test results, and provides a highly portable, powerful and cost-effective alternative to existing bench-top readers which are used widely across the sector.
The solution is also intended for use by public health bodies deploying lateral flow tests, for whom we can deliver Big Data insights along with reading the test, for epidemiological purposes.
By using MagnifEye to interpret diagnostic test readings (for example the COVID-19 lateral flow test) individuals can confirm the test result within seconds, and then communicate the result as required and receive clinical guidance following a positive test.
Clinicians and public health officials can monitor cases at local, regional and national levels and initiate rapid disease containment measures as needed.
MagnifEye can be used across multiple diagnostic test types, including semi-quantitative lateral flow tests where the deep learning algorithm, once trained on the particular test type and medium, can interpret the line intensity to determine how much of a substance is present.
Lateral flow testing is a well-established technology that has been used in the diagnostic sector for decades (examples include: pregnancy testing, drugs of abuse (DOA) testing, etc using a variety of biological matrices from urine, to saliva, to sweat, etc). Lateral flow tests are used across many sectors including agriculture, food and environmental safety.
Compared to conventional lateral flow test readers MagnifEye provides the following advantages:
MagnifEye also reduces the variability associated with human interpretation of lateral flow tests and may improve accuracy.
Unlike most other smartphone readers which employ classical algorithms trained on optical character recognition (OCR), MagnifEye uses deep-learning, and has been trained and validated on 1200 tests across a range of titrated (COVID-19) viral loads. When tested on 166 independently sourced images on different (COVID-19) viral dilution levels visible to the human eye, the algorithm performed with 100% accuracy. At a viral load not readable by the human eye, performance was above 75%.
The product can also be configured to read and report QR/barcodes, to help manage supply chains and support fraud detection.
- Improved accuracy of reading – avoiding human error and reading beyond the human visible spectrum to reduce the number of false negative test results
- Image capture guidance – supports users to take high quality images alongside clear advice and instructions for use
- Fast response – provides lateral flow test results in under 2 seconds.
- Reduced risk of fraud – identifying tests that have been tampered with
- Automated provision of validated test results – securely linking to electronic health records
- Faster generation of epidemiological insights – using Big Data analytics with the capacity to analyse the data across very large populations
- Improved security – encrypted and securely held data transfer and storage with links to healthcare providers and public health officials
- Flexible – easily portable and available in iOS, Android or web app formats, as an API (application programming interface) or SDK (software development kit)
MagnifEye may support the feasibility of asymptomatic testing programmes by potentially improving accuracy and reliability, for example by identifying positive tests that might otherwise be missed using the human eye alone. These individuals can then take action to limit the risk to family, friends and colleagues.
Reducing the likelihood of false negatives provides individuals with greater certainty over their test status. This is especially important for people who wouldn’t otherwise be tested, but still go to work/have children at school (in the case of mass asymptomatic testing). Where these people receive a positive test as part of a mass testing programme, they’ll go into self-isolation, which can help break a chain of transmission which would have otherwise been missed.
In addition, if test results can be interpreted with a greater level of accuracy, asymptomatic testing may be extended to enable the lessening of restrictions on, for example, air travel.
Lateral flow tests for COVID-19 have only come onto the market recently (i.e. approved for use). They are relatively easy to manufacture.
What is not easy is achieving a good enough sensitivity and specificity which depends on the target (in this case the COVID-19 antigen or part of it, the famous spike protein).
Right now COVID-19 rapid Lateral Flow Tests are being used for specific at-risk groups including NHS staff, university students, and for care home visitors, amongst others. The use cases are rapidly growing, and may be extended to mass population testing in the near future.
MagnifEye can potentially improve the quality and accuracy of reading lateral flow test results. In testing, the algorithm performed with accuracy levels above 75% at a viral load not readable by the human eye. We believe this additional level of detection may help to improve the efficacy of a mass asymptomatic testing programme using lateral flow tests. It can also be a powerful tool for enabling testing surveillance, the purpose of which is to understand how the epidemic is spreading, and so inform public health measures, rather than to identify individual cases.
Some lateral flow tests use immunofluorescence and require a special reader to provide a result, but this isn’t typical. A significant number of antigen/antibody tests are simple lateral flow tests requiring a swab placed in a buffer solution, then applied to the test device which creates lines for control/test. The test line may or may not be visible to the human eye depending on the viral load present. MagnifEye can potentially improve the accuracy of the results by using deep learning to read and interpret faint lines that are beyond the human visible spectrum.
Lateral flow tests are usually based on an immunological reaction that happens between an antigen and an antibody. The need for a reader depends on how that reaction is developed; for example, if the development of the Ag-Ab reaction is developed by a gold-conjugate, this would typically reveal a red line that would be visible to the human eye; if on the other hand the manufacturer decides to use a fluorescent particle to show the reaction between Ag-Ab then an instrument that reads in the fluorescent spectrum would be needed. There are cases where the two different development technologies can be used and some other cases where only one achieves the desired results.
COVID-19 has rapidly accelerated the adoption of digital technologies in healthcare systems around the world. Sensyne has been able to respond quickly to this increased demand, for example by offering our award-winning diabetes in pregnancy app, GDm-Health free to the NHS for one year.
Similarly, COVID-19 has also accelerated the use of rapid antigen testing to help governments and public health bodies contain the spread of the disease while at the same time ensuring essential services can continue to operate. MagnifEye aims to play an important role alongside lateral flow tests by potentially improving their accuracy, and providing big data analysis of results in real time to support effective disease surveillance.
Image data and any other data captured as part of the test reading process is transferred via a secure API link to secure cloud data storage.
Organisations receive a real-time, aggregated view of the results, helping them understand the status of cases at any point, and to support appropriate notification to public bodies.
Public health officials could monitor cases at local, regional and national levels and, in the case of COVID-19, implement and tailor rapid disease containment measures as needed.
We protect the data stored in MagnifEye in a number of ways:
- Sensyne Health conform to data protection legislation and no data is shared with third parties.
- All data is always encrypted (whether at rest or in transit).
- Patient Identifiable Data i.e. data which can identify an individual is kept separate to operational data i.e. your blood glucose readings
- We continuously monitor the system to protect against any data breaches
- Cloud storage meets all Information Governance standards, including ISO 27001
- MagnifEye was developed under Sensyne Quality Management System which meets requirements of ISO 13485:2016 standard for medical devices. Sensyne is also working towards compliance to the new Digital Technology Assessment Criteria (DTAC)
We have built MagnifEye but we will not access any personal data unless instructed to do so by NHS organisations or other healthcare providers. In these cases, we specifically ask for permission according to GDPR and Data Protection Act 2018.