DELTAVIEW MODEL 2.1

DeltaViewTM is a radiological image processing system designed to improve the detection of lung nodules by generating a residual image from current and prior patient chest X-rays. It employs advanced image registration, normalization, and bone suppression techniques to highlight temporal changes, and clinical studies demonstrated its efficacy in enhancing radiologists' diagnostic performance.

DeltaViewTM is a post-processing software application designed to assist clinicians in detecting changes in lung nodules by comparing current and prior chest X-ray images from the same patient. The system operates by registering the two radiographic images and then generating a "residual" or "difference" image that highlights areas of temporal change. This is achieved through advanced image processing and pattern recognition techniques, including image normalization and bone suppression—a feature also present in its predicate device—to effectively obscure static anatomical structures and emphasize evolving pathologies, thereby automating and enhancing the traditional manual comparison process undertaken by radiologists. A multi-reader multi-case (MRMC) reader study was conducted to evaluate the clinical performance of DeltaView. This study assessed the ability of radiologists to detect actionable lung nodules using standard prior and current PA/AP X-ray image pairs both with and without the adjunctive DeltaView image. Utilizing Localized Receiver Operating Characteristic (LROC) analysis, the results demonstrated a statistically significant improvement in radiologists' diagnostic performance when aided by DeltaView. Specifically, the area under the LROC curve increased from 0.477 to 0.536, representing a 12.4% improvement (p=2.9E-05). DeltaView is intended for use by trained professionals, such as physicians and radiologists, as an adjunctive tool for patients at risk of having lung nodules; it is not intended for pediatric use and serves as a secondary image, not a replacement for the original radiographs.

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 specific image registration algorithms are employed by DeltaView, and how robust are they to variations in patient positioning and X-ray acquisition parameters between the prior and current images?; The reader study showed a 12.4% increase in AUCLROC; what was the impact on sensitivity and specificity independently, especially for nodules of varying sizes or subtlety?; How does DeltaView's performance compare specifically to other computer-aided detection (CAD) systems for lung nodules that also utilize temporal comparison?; Are there known limitations or types of image artifacts (e.g., from motion, implants) that significantly degrade the quality or reliability of the DeltaView residual image?; What is the typical processing time required by DeltaView to generate the residual image, and how does this integrate into a standard radiology workflow?