Five minutes with Alex Batchelor, Co-founder & CEO at PicturaBio

PicturaBio Ltd moved into R&D labs at the Trust’s Wood Centre for Innovation to commercialise a novel approach of detecting pathogens by the unique light signature they emit when made fluorescent. 

Anyone who has waited for a medical test, such as a PCR, will know three things. They take a long time, they’re expensive (for the payer) and they normally only check for a single infection.  

PicturaBio is hoping to change this. Its imaging technology uses artificial intelligence (AI) to detect viruses, bacteria and other pathogens behind a wide range of infectious illnesses. Rather than test for each possible infection individually, their system will take one sample from a patient and reveal if it finds any pathogen known to cause an infectious disease. 

The result is available within one to two minutes, allowing a healthcare professional to immediately offer the most appropriate treatment. This is important because waiting for results is not only time-consuming, it means the patient is no longer in front of a doctor, so the chance to intervene early is lost.  

The test relies on a fluorescent label which, unlike conventional tests is not designed to cause a chemical reaction to reveal a specific pathogen is present. It instead causes all pathogens in a sample to glow when a light is shone at them. Each reacts differently, creating a unique light signature, a little like a fingerprint. PicturaBio’s CEO, Alex Batchelor, explains the novel approach means many diseases can be tested for at the same time. 

“Most tests are like baiting a line for a particular fish, you have to use a reagent for each pathogen you want to test for, so it’s expensive and time-consuming,” he says. 

“A doctor has to have a very good idea of what they need to look for or run several different tests to cover different infections. If they’re right, they tend to find out when the patient is no longer in front of them. Our technology gets away from that by using the equivalent of using a single hook to catch any kind of fish. We don’t need to look for a specific pathogen, the system knows what the light signature for any known pathogen that causes infectious illnesses looks like, and flags them up if they’re present in the sample.” 

The team is programming its AI software with test samples that contain pathogens behind infectious diseases so they are stored in its library of light signatures to look out for. Due to viruses continually changing, this database will always need to be constantly updated to ensure, as the world saw with Covid, the latest variant is always recognised.  

The use of the fluorescent label that allows pathogens to be imaged was discovered by PicturaBio co-founder, Assistant Professor Nicole Robb. Nicole moved from being a virologist at the University of Oxford to the Physics Department to widen her research horizons. The new direction in her academic career led to the discovery of a new way of detecting pathogens. She found that rather than adding specific labels for specific pathogens, a universal label that lights up all pathogens under a microscope worked instead. The labelled pathogens can then be recognised and told apart by an artificial intelligence-based computational algorithm.

The novel approach worked, and prompted the formation of PicturaBio to commercialise the technology. The business spent much of 2021 generating data and seeking investment before announcing pre-seed funding of £2.6m in February 2022. The round was led by IQ Capital and Ahren Innovation Capital with the participation of Science Creates Ventures. 

This AI imaging part of the system was developed by co-founder, Nicholas Shiaelis, a PhD student in the Physics department of the University of Oxford, specialising in finding new ways to detect infection-causing pathogens. He is also a part-time employee in PicturaBio where he works on both software and assay development of the test. 

With the system working well in the lab, Alex Batchelor reveals the rest of the year is going to be spent testing if results can be repeated at outside healthcare facilities. These hospitals and clinics will run samples they know to be infected with a particular disease to see if the system detects them. 

Once PicturaBio’s work can be repeated in a third-party setting, the final stage before rolling out the technology, will be clinical trials. These are expected to take place during 2024 with the possibility the technology could become commercially available the same year. 

“We need to start by getting other clinics to show our system can detect the pathogen in their samples, and then we need to have it working in clinical trials with tests from real patients,” explains Batchelor. 

“The aim is for us to get it down to the size of a shoe box and have each test put into a cartridge that is then inserted into the equipment. We’re aiming to get it in hospitals from the end of 2024. We’ll then be looking to get it into GP practices before, hopefully one day reaching local pharmacies.” 

Before then, the business is spending the latter half of March getting settled in at the Oxford Trust’s Wood Centre in Headington, Oxford. It chose the location because it offered collaboration between team members working in their office on the AI imaging software and those working in an adjoining laboratory. Access to both laboratory and office space in the same building is not always possible, Batchelor explains, but getting access to both in adjoining rooms is even rarer.