MaxFOV 2

MaxFOV 2 is a software option for GE CT scanners that utilizes a deep learning-based algorithm to extend the display field of view up to 80cm, improving visualization of anatomy outside the standard scan field. This technology is primarily intended to enhance radiation therapy planning and aid in imaging large patients by providing more complete anatomical information from head and whole-body CT scans.

MaxFOV 2 is an advanced image reconstruction software option for GE's CT scanners, designed to extend the display field of view (DFOV) beyond the standard 50cm scan field of view (SFOV) up to 80cm, depending on the CT system's bore size. Its primary purpose is to visualize patient anatomy, including the skin line and CT numbers of tissues, that may be truncated due to large patient habitus or off-center positioning. This capability is particularly beneficial for radiation therapy planning, allowing for more accurate simulation and treatment planning for cancer patients, but it can also be used for routine CT imaging to visualize anatomy outside the SFOV for patients of all ages, including bariatric individuals. The software is compatible with Axial, Helical, and Cine acquisition modes and operates on various GE CT scanner platforms with bore sizes ranging from 70cm to 80cm. The core technology of MaxFOV 2 is a deep learning algorithm, specifically a Convolutional Neural Network (CNN), which was trained on multiple CT scanners to enhance the extended field of view reconstruction process. This represents an improvement over the predicate device's analytical model, which allowed visualization up to 65cm. Non-clinical testing using phantoms confirmed the system's image quality performance, including patient contour accuracy and CT number accuracy, across different CT platforms. Furthermore, a clinical reader study involving 49 CT exams, evaluated by five external clinical readers, demonstrated the system's effectiveness in depicting the patient's skin surface, tissue densities in the extended FOV region, and overall image quality, supporting its performance claims for its intended applications in head and whole-body CT scans.

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 MaxFOV 2 deep learning algorithm compare to traditional analytical EFOV reconstruction methods in terms of specific image quality metrics like spatial resolution and noise in the extended FOV region?; What are the potential limitations or artifacts introduced by the deep learning algorithm in MaxFOV 2, particularly in challenging cases with extreme patient off-centering or unusual anatomy outside the SFOV?; How robust is the MaxFOV 2 algorithm across different CT scanner models and acquisition protocols (e.g., varying kVp, mAs, slice thickness) not explicitly covered in the provided testing summary?; What was the range of patient body mass index (BMI) or habitus included in the clinical reader study, and were there observable differences in MaxFOV 2 performance for extremely large patients?; Beyond visual assessment by readers, were quantitative metrics used to evaluate the accuracy of CT numbers and geometric fidelity of anatomical structures reconstructed by MaxFOV 2 in the extended field of view, especially for radiation therapy planning?