Wearable Corneal Imaging Device for the Home- Based Monitoring of Patients on Blenrep™ Therapy

A novel method that allows quantifying corneal clouding caused by keratopathy and monitor disease progression over time. A quick and reliable, acceptable, and accessible means of quantifying corneal clouding, objective quanti=cation of the haze seen by clinicians via an image processing technology.

How it Works

Wearable IR Camera

Head-worn, highly sophisticated iris recognition camera that is small and adapted to achieve rapid, reliable and standardised objective imaging in a wide variety of patients while minimising artefactual interference in image quality.

Reliable Eye Imaging

The use of multiple wavelengths (incl. near- infrared) allows penetration of even severe levels of corneal haze; the iris camera rapidly takes several images that are then combined and pre-processed for analysis.

Image Analysis

Propriatary image analysis algorithm is then applied to pre-processed images from the iris sensor; Pixel intensity map is generated and corneal opaci=cation measure (COM) score is generated; the change in corneal opaci=cation is measured over time.

Clinical Evidence and Data

Validity:

Figure 2 demonstrates in graphical form a clear and strong relationship between the values of corneal opacity score obtain by our algorithm and the clinical grading of participants. Linear regression (SPSS V.14) con=rmed a strong relationship with a coeQcient of p<0.0001.
In addition to the 18 eyes of patients with anterior segment pathology who were analysed by the COM measure, an additional 18 normal eyes were

COM scores of zero.

assessed by the algorithms. All 18 normal eyes gave Figure 1:

Plot of corneal opaciﬁcation measure (COM) score against clinical grading of corneal opacity. This
ﬁgure is produced in colour in the online journaldplease visit the website to view the colour ﬁgure.

Reliability:

Seventeen subjects eyes (included four normal eyes) had imaging repeated by the same examiner on two occasions, at least an hour apart to assess the reliability of the COM measure. A Blande-Altman plot (=gure 2) demonstrates no evident systematic bias and narrow coeQcient of repeatability of 1.21 (95% limits of agreement). The intraclass correlation coeQcient (mixed model) for average measures (SPSS V.15.0) was 0.997 (0.989e0.999).

Figure 2:
Blande-Altman plot of mean of corneal opaciﬁcation measure (COM) repeat measures against difference. This ﬁgure is produced in colour in the online journaldplease visit the website to view the colour ﬁgure.

Development Approach

Stage 1: Working Prototype Model
D U R A T I O N : 4 - 5 M o n t h s B U D G E T : £ 3 6 0 , 0 0 0

image representations of patients with disease are captured

Decision Gate: The prototype is able to repeatedly take high- quality/precision/repeatable

Stage 2: Algorithm Development (n=20)
D U R A T I O N : 2 - 3 M o n t h s B U D G E T : £ 1 8 0 , 0 0 0

The =rst stage of the project will Iteratively Adjust the imaging be the development of a system and an image analysis

volunteers in order to develop the engineering and software analysis

corneal images. SuQcient quality algorithm are able to detect for the analytics algorithm to keratopathy
process. Varied enough to
combine and compensate for any artefacts

Stage 3: Validating the Algorithm (n=90)
D U R A T I O N : 6 M o n t h s B U D G E T : T B D
Acquire a bigger sample of images to enhance and adjust

wearable IRIS scanner integrated software to automatically detect the sensitivity and reliability of into a standard VR. And iterative and grade corneal abnormalities. the diagnostics system and adjustment of the imaging We will use a a small number of image analysis software. system until the highest quality patients with Blenrep corneal

disease (20 patients) and healthy Decision Gate: Demonstrate

relationship between the camera automatic pathology score as well as remote expert analysis of images with grading obtained by

Decision Gate: Demonstrate that clinical grading of participants. the proposed system and (n=90)

Stage 4: Prepare for the Clinical Trial (n=TBD)
D U R A T I O N : T B D B U D G E T : T B D
Engineering and development of the pilot-ready medical device (incl cert) in parallel with design and set up of the clinical study

Decision Gate: TBD

20 Air Street, London Phone: +44 7873 942426
E-mail: stan@givevision.net

KEY FEATURES:

Precision: Multiple images of each eye with different illumination angles and frequencies of light including near- Infrared wavelength illumination penetrating through any severe level of corneal haze.
Scalability: doesn't rely on in-clinic visits to support clinical decision-making (initiating/reducing treatment) as well as advanced research and clinical trials.
Consistency: Wearable device allows for the =xed distance between the eye and the sensor distance as well as consistent low light environment to maximise image quality and minimise variability.
Accessibility: Rapid, simple to use and non-invasive diagnostics allows for unsupervised use by patients with even the most severe physical or mental disabilities.

Product Evolution
Form Factor: Once we have validated the clinical eQcacy of the proposed system (Stage 3) using a VR headset as a platform for the imaging sensor we will commence (Stage 4) engineering work to miniaturise and re=ne the device into a hardware solution with a form factor of a small and comfortable pair of electronic glasses.
Architecture: Wearable device (a) with an IRIS imaging system (b) on board equipped with a pair of high-resolution displays for the Visual Acuity and Contrast Sensitivity test (c) connected via a smartphone app (d) to the secure cloud where images are processed and analysed (d).

The application of IRIS biometric systems for ophthalmic diagnostics was pioneered in 2011 by Professor Tariq Aslam - a Consultant Ophthalmologist at Manchester Royal Eye Hospital and a Professor of Ophthalmology and Interface Technologies at the University of Manchester.
The studies initially published in 2012 and further publication up to 2019 (as well as ongoing work) have shown that their algorithm analysing corneal IRIS images:
Provided an objective measure of corneal clouding
Has good repeatability and reliability
Was faster, more practical and objective than a full ophthalmic examination at quanti=cation of corneal clouding (including subjective assessment of corneal clouding with a slit-lamp and use of more complex corneal imaging equipment)

Company Background
Sight.ED™ is a clinically validated, wearable electronic sight aid that enables patients with severe sight loss to see clearly again, while at the same time, providing AI-powered continuous remote sight monitoring including:
Visual acuity and contrast testing: an inbuilt high- resolution display allow the user to perform a visual acuity test from the comfort of their home.
Retinal Imaging: in-built retinal imaging and analytics system detects the changes in the retinal thickness as a proxy for disease deterioratio
Вернуться к редактированию

Company Background

Содержание

Похожие презентации