Ocular Uganda

Automating Microscopy

Ocular Uganda

Automating Microscopy

Ocular Uganda

Automating Microscopy

Ocular Uganda

Automating Microscopy

ABOUT US

Leveraging existing technologies like smartphones and the widespread availability of microscopes in health centres, we developed a 3D printable adapter that attaches to the microscope's eyepiece and slots into a smartphone for image capture. The smartphone adapter is designed to enable a user to correctly align a smartphone of nearly any model with the focal point of the eyepiece. This mechanism is achieved by using sliders and side-holders, which enable positioning of the phone. The locking mechanism is used to lock the smartphone in position once the appropriate alignment has been made. Once the adapter is locked in alignment, the user can easily slide the smartphone in and out of the adapter at any time without compromising the pre-set alignment. Through computer vision, the technology processes images, guiding microscopists to identify pathogens in malaria, tuberculosis and cervical cancer more efficiently. Ocular offers this customised adapter as well as a mobile and web application to health units as a decision support tool.

MISSION

Ocular commits to providing accessible and efficient disease diagnosis through cutting edge AI-powered mobile microscopy.

VISION

Ocular envisions a future where every individual, regardless of location or resources, has access to reliable and timely disease
diagnosis through artificial intelligence-supported mobile microscopy.

Our Methods

Data Collection

Image representing a computer downloading data

A team led by microscopists designs protocols to capture diverse images of parasites across different fields of view, tools, demographics, and regions. This ensures the dataset reflects variation in parasite expression and species prevalence.

Expert Labelling

Image representing a computer downloading data

Medical experts meticulously label the captured images, providing accurate ground truth for model training.

Algorithm Development

Image representing a computer downloading data

Our tech team designs and trains Machine Learning algorithms using the labeled data. These algorithms learn to extract key features and patterns, enabling them to detect and count parasite species with high accuracy.

Pilot & Proof of Concept

Image representing a computer downloading data

Ocular's initial focus is on malaria, aiming to establish a successful pilot and demonstrate the proof of concept for wider application.

Supported by:

Google Organization's LogoNational Institutes Of Health LogoLacuna Fund Organization's logo

Partners:

Makerere University's Artificial Intelligence Lab's LogoUganda Cancer Institute LogoMakerere University LogoKirudu National referral Hospital LogoMulago National Referral Hospital Logo