NHS hospitals use artificial intelligence to practise brain surgery

Six NHS hospitals are piloting an artificial intelligence (AI) system designed to make brain surgery safer for patients.

Using a 3D representation of a patient’s brain, the software enables surgeons to practise procedures for treating brain aneurysms. These procedures, which involve inserting a tiny stent into blood vessels, are complex and delicate.

The PreSize technology, developed by startup Oxford Heartbeat, has already been used successfully at Leeds Teaching Hospitals. Professor Tufail Patankar, leading consultant interventional neuroradiologist and chief investigator for the study, carried out the first virtual practice operation.

The software works by creating an accurate 3D model of the patient’s vessel tree, indicating where a stent should start and finish. This enables the surgeon to select the most appropriate stent for the procedure. In a study of 250 real stenting procedures, the software was found to have a precision rate of nearly 97%.

Normally when carrying out brain surgery to treat aneurysms, there has been a risk that surgeons would choose the wrong stent. Brain aneurysms are so small (just a few millimetres in size) that selecting the most appropriate stent is difficult. Choosing the wrong one, however, could lead to fatal consequences.

By using the 3D model of the brain to guide them, the hope is that surgeons will be able to have greater confidence that they are making the right decision, patients will have better outcomes and the NHS will save money.

Patankar said: “Better informed preparation and technological support can help us clinicians deliver the best treatment choice personalised for each patient, making treatment of aneurysms, especially complex ones, safer and quicker.”

The pilots are being supported through the NHS AI lab.