The tech behind the test
How PinPoint unlocks the data in blood to predict your risk of cancer
The PinPoint Test is a regulated ‘Software IVD’ (invitro diagnostic), UKCA marked for use as a tool for triage and stratification of suspected cancer patients.
PinPoint represents an intelligent upgrade to traditional pathology. Rather than simply looking down a list of analytes and identifying which are out of range, PinPoint does what a human cannot: it analyses 33 data points all at once, assessing linkages and interdependencies between them. This is only made possible by machine learning.
We are known as an
‘AI-driven blood test’, but what does that mean?
PinPoint uses a type of AI called machine learning to develop the test. It’s a technology that can analyse massive amounts of data to identify patterns and ‘rules’ that help us predict future outcomes.
For us, that means understanding how information in the blood, a person’s age and their sex relate to their risk of cancer.
The language around AI can be confusing. In this video, our Chief Scientist and Chief Technology Officer explain what those terms really mean.
Rapid AI update cycle
The PinPoint Test was trained using a retrospective dataset of 371,799 patients. Based on rules derived from the training data, a patient’s bloods are processed against past cases, making it possible to predict the probability of cancer now.
The more tests are processed, the more accurate future versions of the algorithm can become.
This process of AI improvement is already underway as new, more powerful generations of our algorithms are developed with newly available real-world data.
In this video, we hear about how AI can be applied to healthcare and the potential benefits of this exciting new technology.
PinPoint algorithms are not capable of ‘continuous learning’ and that is not the intention. At every stage there is human oversight and intervention to gradually develop and improve the test.
Like any piece of regulated medical software, the process of developing new updated versions is one which involves generating evidence, testing, conducting risk analyses and regulating the final product for approved use.