ENT EM

The ENT EM system provides image-guided planning and navigation for Ear, Nose, and Throat procedures, utilizing electromagnetic tracking technology to continuously localize instruments relative to patient-specific medical images. Its core science involves sophisticated software running on mobile IGS platforms, featuring multiple registration methods including an AI/ML-based Convolutional Neural Network for automated landmark detection to aid surface matching.

The ENT EM system is an image-guided planning and navigation tool designed to assist in Ear, Nose, and Throat (ENT) procedures. It operates by linking a patient's anatomical structures, identified via a patient reference, with preoperative medical images such as CT, CTA, X-Ray, MR, MRA, and ultrasound. This linkage allows for the continuous localization of medical instruments and the patient's anatomy in real-time, facilitating precise surgical interventions. The system utilizes electromagnetic (EM) tracking technology to achieve this, with software installed on mobile Image Guided Surgery (IGS) platforms (Kick 2 Navigation Station or Curve Navigation 17700) that display tracked instruments within the patient's image data, guiding the surgeon through various preoperative and intraoperative steps. A key technological aspect of ENT EM is its software, which includes modules for registration, instrument handling, and navigation. Notably, it incorporates an artificial intelligence/machine learning (AI/ML) algorithm, specifically a static, supervised learning-based Convolutional Neural Network (CNN), to aid in the patient registration process by providing pre-registration guide points or landmarks for surface matching. The system is indicated for any medical condition where a reference to a rigid anatomical structure can be identified relative to images, specifically for intranasal structures and paranasal sinus surgery (including functional endoscopic sinus surgery (FESS), revision, and distorted anatomy cases) and anterior skull base procedures. Performance data demonstrated that the system achieves a mean positional error of ≤ 2 mm and a mean angular error of ≤ 2°, meets usability requirements for the intended users, and complies with standards for electrical safety, electromagnetic compatibility, and biocompatibility/reprocessing of associated instruments.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. How does the performance of the AI/ML-based landmark detection for pre-registration vary across different qualities and modalities of input medical images (e.g., CT vs. MR, noisy vs. clear scans)?; What was the demographic and anatomical diversity of the patient dataset used for training and testing the Convolutional Neural Network, and how might this impact its accuracy in atypical anatomies or populations not well-represented in the training data?; Beyond the stated system accuracy (≤ 2 mm positional, ≤ 2° angular) in controlled bench testing, what is the reported end-to-end clinical accuracy of the ENT EM system when used in complex surgical scenarios, particularly those involving distorted anatomy?; How does the integration of AI/ML for guide point suggestion in the registration workflow quantitatively impact surgical setup time or the number of attempts required for successful registration compared to purely manual or conventional atlas-based methods?; What are the specific mechanisms or safeguards within the ENT EM system to ensure continued performance and safety if the electromagnetic tracking field is perturbed by other metallic instruments or electromagnetic interference commonly found in an ENT operating room environment?