Article

The Global Rethinkers: Financial Times interview with Bea Bakshi

3 April 2023 | 5 minute read

While working on the frontline of the NHS, Dr Bea Bakshi developed an innovative AI solution that can help diagnose cancer earlier.

“For me, it was simply a case of… you have to do this,” says Dr Bea Bakshi. While working long hours in the NHS – first as a hospital doctor and later as a GP – she developed game-changing new healthcare tech. On a time-management level alone this was, unsurprisingly, “exceptionally hard”, but the reward is huge: “This is software that can profoundly change the narrative for survival in cancer patients.”

Bakshi’s software is called C the Signs, and it can automatically scan medical records for a confluence of cancer risk-factors and recommend the action to be taken. She worked to co-create it with a friend and former colleague, Dr Miles Payling. “We both had a love of technology and this burning desire to improve health outcomes,” says Bakshi. “Miles developed the first wireframe and the first level of the algorithm. We then had funding from NHS England, which enabled us to hire more engineers and data scientists, in order to build out.” Today, C the Signs has close to 40 employees and is being used increasingly across the NHS. “So far, we have diagnosed over 13,000 cancers.”

C the Signs is such a significant innovation in the fight to increase cancer survival rates that it caught the attention of the Obama Foundation

C the Signs is such a significant innovation in the fight to increase cancer survival rates that it caught the attention of the Obama Foundation, which named Bakshi as one of its ‘Obama Foundation Leaders’ in Europe. Bakshi met President Obama at a foundation summit in Copenhagen in 2021: “Having such courage and confidence bestowed upon you by the president of the United States is a moment that stays with you for life.”

Software created by doctors, for doctors

“The technology we built is integrated with electronic medical record systems. It’s that first touchpoint when a patient sees a healthcare professional,” explains Bakshi. “Not necessarily just a GP, but a pharmacist, a dentist, a nurse. We have over 1,000 GP practices using C the Signs across the country and, through that, we have a vast array of healthcare professionals.”

As for how the software works: “Harnessing the power of AI, we are able to pull together all the different signs, symptoms and risk factors that have been academically validated to have a predictive value for cancer,” explains Bakshi. “We use more than 1,000 data points and combine them. When you take this symptom, and you combine it with that factor, and another, and another, it can then take us over a risk threshold so we can say, ‘This patient is at risk of cancer.’” Bakshi gives some simple examples: “Indigestion alone, or a cough – someone might think that’s insignificant. But when combined with other factors, it could mean we should be investigating.”

A life-changing encounter

As a hospital doctor in 2015, Bakshi saw a patient called Joe. “He had previously been very fit and well but, over three months, he had been slowly deteriorating,” she says. “He had seen his doctor several times, tried various treatments – nothing seemed to be working. Then he presented in A&E, significantly unwell. I had the horrible responsibility of diagnosing him with metastatic pancreatic cancer.”

A lot of the symptoms and signs overlap with very common conditions that are quite benign

Joe’s reaction to his diagnosis affected her profoundly. “His response was, ‘Why was it picked up so late? I did all the right things, I went to see my doctor as soon as I started having symptoms.’ He died three weeks later. And it just stayed with me. What if something could have been done earlier?”

When Bakshi later became a GP, she was keenly aware of how hard it can be to spot a cancer such as Joe’s. “For me, it just seemed obvious to ask why we are relying on clinical experience alone when it’s so challenging to recognise cancer in the early stages. A lot of the symptoms and signs overlap with very common conditions that are quite benign. I was spending a lot of time on Google trying to find technology to help improve early cancer recognition. And I just couldn’t find anything. That’s really when C the Signs was born.” By 2021, running C the Signs had become a full-time commitment for Bakshi.

Helping more patients beat cancer

There is a clear financial incentive for NHS healthcare bodies to license C the Signs in order to help detect and cure cancers earlier; it gives a good return on investment. “Early cancer detection is half the cost of a late-stage cancer detection, partly because you avoid the very expensive late-stage chemotherapy treatments and drugs,” explains Bakshi. “When you diagnose early, not only are you having patients survive at least 10 years from their cancer but it’s also 50 per cent cheaper to treat. So it has a huge benefit to the NHS, especially in the current climate. Through C the Signs, we’re able to demonstrate in real time what the cost benefit is.”

When you diagnose early, not only are you having patients survive at least 10 years from their cancer but it’s also 50 per cent cheaper to treat

Plans in the works include developing a patient-facing version of C the Signs, and fine-tuning the healthcare professionals’ version to be even more sophisticated when it comes to identifying the different demographics of patients. “I see there is a problem to solve, both in the UK and globally,” says Bakshi. “And until that problem is solved, for me, this is my life journey now. With this software, I’ve gone from being able to have an impact on thousands of patients to potentially millions of patients.”