The application of Artificial Intelligence (AI) in healthcare has triggered strides in diagnostics, treatment personalisation and delivery times, equipment maintenance, health monitoring and more1. Many of these fields rely on AI in imaging and computer vision to process information that cannot be seen by the naked eye or to learn patterns that couldn’t be or would take far longer to be analysed and interpreted by a human.
Such technology is also being applied in surgery to improve accuracy and reduce operating times and costs. AI imaging can assist in both in vivo and ex vivo surgical procedures. AI for imaging and augmented reality can be used to improve visual feedback during minimally invasive laparoscopic surgeries, including robot-assisted surgeries.
AI imaging for in vivo procedures
EnAcuity is a spin out company from Imperial College London and UCL that has developed the first hardware-free hyperspectral imaging system for surgery without needing expensive hyperspectral sensors/cameras. Hyperspectral imaging divides the electromagnetic spectrum into more bands than the human eye can see2, revealing subtle colour discriminations capable of highlighting tissue perfusions and structures such as vessels, nerves and even cancers to the surgeon. EnAcuity’s promising technology is currently going through clinical validation. We look forward to seeing how their work continues3.
Not just real-time imaging can be used to enhance surgical performance and surgeon experience. Pre-operative images, such as computed tomography (CT), magnetic resonance imaging (MRI) and ultrasounds, can be used to superimpose information and structures onto the surgeon’s view of the operating field. This can be done on a monitor when performing telesurgical robot-assisted-surgery (e.g. using the da Vinci) or rendered in a head-mounted-display.
Augmedics has developed such a headset for spinal surgery. The system, xvision, allows the surgeon to ‘see through’ a patient’s external anatomy and get a clear view of the spine without performing open surgery. It works by using pre-operative CT scans of the spine to segment the spine into vertebras. This information is combined with fluoroscopy images taken intraoperatively to perform registration of each vertebral body. This method accommodates for any changes to structural placements, for example due to patient positioning. A 3D anatomical view of the spine, as well as 2D stereoactic views of the spine are superimposed onto the surgical field via the head-mounted display. This allows more surgeries to be performed laparoscopically by providing a simulation of ‘x-ray vision’4.
Xvision is the first U.S. Food and Drug Administration approved augmented reality system for surgery. Earlier this year Augmedics announced that 10,000 patients have been treated using the system, a major milestone for both the company and the technical field.
These developments all provide visual feedback that can make minimally invasive surgeries safer for little extra cost and allow for more procedures to be performed laparoscopically, reducing recovery times and tissue damage for patients, as well as reducing radiation exposure for both patients and surgeons and saving costs that can be expended on treating more patients.
AI imaging for ex vivo procedures
Although a lot of surgery takes place on and in the body, essential steps such as sample analysis during surgery must also take place. Speeding up such processes is desirable to shorten surgery times, during which the time a patient is exposed to risks such as infection and general anaesthesia5.
One such procedure is cancer margin assessment. Removing cancerous tissue with a safe margin of healthy tissue around the cancer to ensure no cancer cells are left in the body is a primary treatment method; however, it can be a challenge to identify a tumour’s boundaries in order to do this. Pathologists evaluate the edges of excised tissue specimens to determine if the edges contain cancerous cells. This is highly skilled, detailed work, which takes roughly 20 minutes intraoperatively but can take longer.
To speed up this process 3Sonic have developed an AI-based cancer tumour imaging platform, which performs ultrasound scans of resected tumours and uses AI to assess whether the surgical margin is sufficient. The system has been tested on oral cancer patients with promising results. 3Sonic’s solution is not only faster than a pathologist having to manually perform the margin assessment but also requires no registration of the specimen, is cost-effective by reducing surgery times, and is designed for portability so can be integrated into almost any space and workflow6.
Conclusion
AI in surgical imaging is making great strides in providing safer and more efficient surgeries. The rapid development in this field is also reflected in the patent landscape. The number of patent filings for AI assisted MedTech solutions is increasing quickly, as is further analysed and discussed in our report on how AI is shaping the future of MedTech. With new technologies in AI becoming available, medical research is ever more focussed on innovation, making the need to consider IP strategy early in development all the more pressing. We look forward to seeing how these technologies progress as AI continues to develop and medical research finds new ways to make use of it.
References:
- https://www.mdpi.com/2075-4418/14/14/1472
- https://en.wikipedia.org/wiki/Hyperspectral_imaging
- https://www.imperial.ac.uk/news/265673/enacuity-wins-grant-bring-ai-driven-surgical/, https://www.imperial.ac.uk/news/251276/enacuity-develop-visualisation-technology-minimally-invasive/,
https://www.enacuity.com/home
- https://augmedics.com/xvision/,
https://today.uconn.edu/2025/04/augmedics-completes-10000th-augmented-reality-spine-surgery-performed-at-uconn-health/,
https://www.massdevice.com/augmedics-launches-new-features-xvision-spine/
- https://madeforthismoment.asahq.org/anesthesia-101/types-of-anesthesia/anesthesia-risks/
- https://www.3sonic.com/about
https://copenhagenhealthinnovation.dk/opslag/3sonic-a-novel-3d-imaging-device-to-improve-precision-in-cancer-surgery/
