The problem of accurately identifying which pancreatic cancer patients will respond to certain kinds of chemotherapy is especially pressing in the midst of the COVID-19 pandemic. The tracer used to bring pancreatic cancer patients to benefit is called a neuroendotoxin.

Artificial intelligence has granted artificial intelligence (AI) researchers a new tool to simulate risk of receiving a particular drug and this tool helped the researchers derive a tool that clarifies the most simplistic way of drawing the many dots connecting different types of tumors.

In May Sridhar Ramasamy a postdoctoral fellow in electrical and computer engineering at Imperial College London the team posted a proof-of-principle demonstration online.

Previously doctors had to estimate risk of toxicity while drawing hundreds of different cancer tumor connections using similar methods which is time-consuming and expensive.

However the new paper explores a more complex simulation by modelling the effect of a given tumor on the success rate of any given chemotherapy.

The paper also gives Anand Swaroop the lead author on the paper explaining:

There are many effects of chemotherapy that patients can have which are very difficult to assess including nerve damage gastrointestinal problems and hypertension. The human body is however a very complicated machine and AI gives us a technology to do away with the uncertainty of modelling these effects

Having such broad and complex information available to target and compare treatments gives a whole new field of medicine where our collective judgement is much more complete adds Robert Piper a consultant in the UKs Royal College of Surgeons of England.