aPROMISE X

aPROMISE X is a software platform intended for healthcare professionals and researchers to enhance the review, quantification, and reporting of PSMA PET/CT medical images in oncology. Its methodology centers on DICOM image management, machine-learning assisted hotspot detection for regions of interest, and user-validated automated quantitative analysis to improve consistency and efficiency in oncological assessments.

aPROMISE X is an updated software platform designed for healthcare professionals and researchers to manage and process digital medical images, specifically focusing on Prostate-Specific Membrane Antigen (PSMA) Positron Emission Tomography/Computed Tomography (PET/CT) scans for oncology. It functions as a medical image management and processing system, facilitating the acceptance, transfer, storage, display, manipulation, quantification, and reporting of these images. The system, accessible via a web interface, can be deployed on cloud infrastructure or a local server and supports the DICOM 3 standard for image data, allowing uploads via file selection or direct transmission from a Picture Archiving and Communication System (PACS). It presents studies in a 4-panel view (PET, CT, PET/CT fusion, and maximum intensity projection) and offers image visualization and analysis tools for users to identify and mark regions of interest. The core methodology involves users reviewing image data and marking regions of interest (ROIs) either by selecting from pre-defined hotspots, which are identified by a machine-learning algorithm detecting high local intensity areas in the PET series, or by manual voxel selection. These selected hotspots undergo automatic quantitative analysis, calculating metrics such as Standardized Uptake Values (SUV-max, SUV-mean, SUV-peak), volume, Lesion Index, and Intensity-weighted Tissue Lesion Volume. Users then review these quantitative results to determine which hotspots represent suspicious lesions for inclusion in electronically signed reports, exportable as JPG or DICOM files. Performance data from bench testing confirmed its accuracy and linearity using a digital phantom, equivalent performance to its predicate for marking and quantitative assessments, and an analytical study demonstrated that aPROMISE X enables more reproducible and efficient automated quantification of tracer uptake compared to manual methods, while maintaining physicians' diagnostic performance.

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 machine-learning model for hotspot detection in aPROMISE X differ from the analytical model used in its predicate, and what specific improvements in sensitivity or specificity for lesion detection were observed in comparative studies?; What are the specific validation metrics from the 'Analytical Performance in Clinical Study' that quantify the increased reproducibility and efficiency of automated quantification with aPROMISE X compared to manual methods by clinicians?; Can aPROMISE X be customized or trained further with local institutional data to improve its hotspot detection or organ segmentation accuracy for specific patient populations or scanner protocols?; What are the computational requirements (e.g., processing time, memory) for aPROMISE X when handling large PSMA PET/CT datasets or multiple simultaneous users, particularly for the cloud-deployed configuration?; Beyond the listed SUV metrics and volume, are there plans to incorporate more advanced radiomic features or textural analysis within aPROMISE X for a more comprehensive characterization of identified lesions?