Scanner agnostic image-guided biopsy platform employing advanced computer vision and computer graphics to enable non-invasive registration and dynamic projection mapping

Illuminant Surgical Inc. received a $1,107,943 Project Grant from the National Cancer Institute under the Cancer Detection and Diagnosis Research Program (CFDA 93.394) awarded on September 12, 2025, with a completion date of August 31, 2027. This Phase II effort focuses on the continued development and validation of Skylight, a mixed-reality navigation platform designed to improve the accuracy and safety of image-guided biopsy procedures. The platform employs advanced computer vision and computer graphics technologies, utilizing non-invasive adhesive fiducials (referred to as "skindots") and real-time projection mapping to guide biopsy instruments with millimeter-scale precision during CT-guided procedures targeting lung, liver, and spinal lesions.

The primary deliverables under this award include: (1) development and testing of software algorithms for real-time respiratory motion tracking and compensation using external markers and preoperative imaging; (2) creation of cadaveric models with physiologic respiratory movement to simulate realistic anatomical conditions; and (3) validation of the Skylight system's targeting accuracy across common anatomical biopsy targets. The project is intended to reduce diagnostic delays and procedural complications by enabling dynamic, spatially-aware visualization of subsurface anatomy directly on the patient's body, thereby improving the safety, efficiency, and diagnostic yield of CT-guided biopsy procedures. Research will be conducted at Illuminant Surgical's facility in Palo Alto, California.

Abstract

Accurate targeting during image-guided biopsy procedures is critical for the diagnosis and treatment of cancers such as lung, liver, and spine tumors. However, current CT-guided biopsy workflows rely on static 2D imaging and manual interpretation, which can lead to misdiagnosis, repeat procedures, and increased patient risk-especially in anatomically complex or mobile regions affected by respiration. This project proposes the continued development and validation of skylight™, an innovative mixed-reality navigation platform that uses non-invasive adhesive fiducials (skindots™) and real-time projection mapping to guide biopsy tools with millimeter- scale accuracy. The long-term objective is to reduce diagnostic delays and procedural complications by enabling dynamic, spatially aware, and intuitive visualization of subsurface anatomy directly on the patient's body. This phase II effort will advance skylight's capabilities by implementing and validating a real-time respiratory motion compensation algorithm using tracked external markers and preoperative imaging. Cadaveric models ventilated with programmable respirators will be used to simulate realistic anatomical motion and evaluate targeting accuracy across lung, liver, and spinal lesions. The specific aims are: (1) to develop and test software algorithms for tracking and compensating for respiratory motion; (2) to develop a cadaveric model with physiologic respiratory movement for system evaluation; and (3) to validate skylight's accuracy in guiding biopsies across common anatomical targets. The proposed system has the potential to significantly improve the safety, efficiency, and diagnostic yield of CT-guided procedures, aligning with NIH's mission to improve health through innovative, patient-centered technologies.