Rayvolve PTX-PE

Rayvolve PTX-PE is a software-only medical device designed to assist healthcare professionals by using artificial intelligence algorithms to passively identify suspected pleural effusion and pneumothorax on chest x-rays. Its purpose is to facilitate worklist prioritization and triage, with its methodology validated through comprehensive testing demonstrating high sensitivity, specificity, and AUC values for these critical findings.

Rayvolve PTX-PE is a software-only device utilizing artificial intelligence algorithms to analyze chest x-ray images (Postero-Anterior or Antero-Posterior views) from adult patients (18 or older). Its core purpose is to identify suspected critical findings, specifically pleural effusion and/or pneumothorax, by providing study-level outputs that facilitate worklist prioritization or triage. This device integrates with DICOM node servers like PACS, functioning as a passive notification tool within the standard clinical workflow without providing proactive alerts, highlighting specific image areas, or serving as a standalone basis for clinical decision-making, emphasizing its role as an assistive tool rather than a replacement for medical professionals.The device's efficacy has been demonstrated through comprehensive software verification, validation, and bench testing, confirming its substantial equivalence to the previously marketed Lunit INSIGHT CXR Triage device. This determination is supported by shared intended use, similar technological characteristics, and operational principles. Test results indicate Rayvolve PTX-PE effectively detects pneumothorax and pleural effusion with high sensitivity, specificity, and Area Under the Curve (AUC) values, confirming its intended performance. Minor differences from the predicate device, such as the specific patient age range or image modality, were found not to introduce new questions of safety or effectiveness.

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 AI algorithm architecture and training data were used to achieve the reported high sensitivity and specificity for pneumothorax and pleural effusion detection?; How does the integration with existing PACS systems and the passive notification approach impact clinical workflow efficiency and patient outcomes in real-world settings?; Are there plans for future iterations of Rayvolve PTX-PE to include detection of other critical chest pathologies beyond pleural effusion and pneumothorax, or to expand to different imaging modalities?; What is the device's performance in challenging or atypical cases, such as subtle findings, superimposed artifacts, or imaging from patients with complex medical histories?; How do the high sensitivity and specificity translate into the positive and negative predictive values for a typical clinical prevalence of these conditions, and what is the false positive/negative rate in practice?