Robotics and AI platform for minimally invasive Neurosurgery (RAIN)

Introduction Endonasal surgery, one of the best examples of endoscopic neurosurgery, allows skull-base tumour removal through the natural orifice of the nostril.

Although a promising alternative to invasive transcranial approaches, it comes with its limitations, mainly the lack of instrument articulation within the constrained operative spaces.

The same practical constraints apply to many other procedures such as Endoscopic Sinus Surgery, Endoscopic Ventriculostomy and Endoscopic Spine Surgery.

Thus, surgical innovation is needed since these surgeries are limited to specialised centres, reducing patient access and straining hospitals with high referral costs and readmission rates.

We are proposing an innovative solution, the Robotics and AI platform for minimally invasive Neurosurgery (RAIN), which deploys handheld robotic instruments to increase the surgeon s dexterity, and AI-assistance software to support their decision-making. Our platform can make endoscopic neurosurgery safer, more effective, and feasible in more clinical cases.

Methods Currently, our RAIN platform consists of a core-set of robotic instruments with flexible tips and preliminary AI-assistance software around surgical-phase recognition and anatomy segmentation.

With this project, we aim to expand our instruments portfolio to include specialised articulated instruments incorporating vision, suction and energy. Simultaneously, we will further develop our machine-learning algorithms, to produce a fully-integrated platform.

The RAIN platform will be evaluated for its safety and feasibility during an IDEAL Stage 1 first-in-human study, comprising six patients with skull-base tumours.

Timelines for delivery Due to its instinctive handheld control and affordable cost, the RAIN platform has the potential to be quickly adopted by neurosurgeons. A two-year plan dictates the activities of this project.

Within the first 16 months, we will develop our instrument portfolio and machine-learning software, and spend the remaining time for integration into a single RAIN platform.

To conduct our first-in-human study, MHRA regulatory approval for use of a non-certified device will be acquired within 19 months, allowing 5 months for delivery.

This feasibility study will generate clinical data to support subsequent fundraising, and wider studies that will lead to multi-region certification within the next 2-years after the project ends.

Dissemination Patients involvement has been pivotal since the inception of our technology suggesting the use of handheld robotics and playing an active role in shaping this proposal. Moreover, a patient advisory group will be consulted throughout this project.

Thus, we understand the importance of disseminating our project outputs to patients, the scientific community and to the general public. We intend to do so via short videos, infographics, events, and peer-reviewed journal articles.

Discussion In this project, a Robotics and AI platform for minimally invasive Neurosurgery will be developed and evaluated during a first-in-human clinical study.

Our disruptive solution can democratise endoscopic neurosurgery by enhancing it with the benefits of robotic surgery, only at a fraction of its cost.

This will allow more surgeons to perform keyhole procedures expanding the number of patients that can enjoy their benefits and cutting down referral costs.

Additionally, the occurrence of repeat-procedures due to complications or incomplete tumour resection will be reduced, freeing up hospital beds and resources.