Limbus Contour

Limbus Contour is a software medical device designed to automate the contouring of CT and MRI DICOM images to support radiation treatment planning for trained oncology professionals. It achieves this by employing an automated segmentation methodology, validated through rigorous benchtop testing comparing its outputs to expert-created contours using the Dice Similarity Coefficient, consistently meeting performance requirements with good geometric alignment.

Limbus Contour is a stand-alone software medical device designed for the automatic contouring (segmentation) of CT and MRI DICOM images. Intended for use by trained radiation oncologists, dosimetrists, and physicists, the software aims to derive optimal contours essential for radiation treatment planning. It facilitates the creation, transformation, and modification of contours, seamlessly integrating with existing radiotherapy treatment planning systems and DICOM image viewers by importing image files and exporting segmented RT-Structure Set files. This software assists in localizing and defining anatomical structures for treatment evaluation and planning, but it is not intended for use with digital mammography.The device has been found substantially equivalent to a previously marketed version of the Limbus Contour software. While maintaining identical indications for use and a similar software-only design, the current version expands the range of anatomical structures available for automatic contouring and has a slightly increased RAM requirement. Performance was validated through rigorous software verification and benchtop testing, where contours generated by the software were compared to expert-created contours using the Dice Similarity Coefficient. The results demonstrated that the device's outputs consistently meet specified requirements and exhibit good geometric alignment, requiring minimal editing, thus confirming its comparable performance to the predicate.

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. What is the specific clinical impact of the "minimal editing" required for the automatically generated contours on treatment planning efficiency and accuracy?; How does the expanded range of anatomical structures in the new version compare in accuracy and reliability to the structures available in the predicate device?; Are there any specific types of image artifacts or patient anatomies that Limbus Contour performs less optimally with, and how are these handled or mitigated?; What are the long-term data on the clinical outcomes of patients whose treatment plans utilized contours generated by Limbus Contour?; Beyond Dice Similarity Coefficient, what other metrics or clinical endpoints were used to validate the performance and equivalence of the software, particularly concerning dosimetric impact?