Aurora

Aurora is a hybrid SPECT-CT imaging system for nuclear medicine procedures that detects radioactive tracer uptake and provides CT images for attenuation correction and anatomical localization. It is intended for general nuclear medicine imaging, head/body/cardiac/vascular CT applications, and aids in detecting, diagnosing, and assessing diseases including cardiovascular disease, neurological disorders, and cancer.

The proposed Aurora adds a deep-learning Automatic Kidney Segmentation algorithm to the predicate's Real-time Time Activity Curve software feature for Dynamic Renal Scintigraphy images. The algorithm automatically localizes and outlines kidneys to reduce manual steps, though the user must confirm the output. All other specifications—dual NaI-based detectors, Revolution Ascend CT subsystem, patient table, and scan ranges—remain identical to the predicate device (K243605).

IEC 60601-1 and applicable Collateral and Particular Standards including IEC 60601-1-2, 60601-1-3, and 60601-2-44; NEMA NU-1 for performance testing; ISO 13485 quality system standards.

The kidney segmentation algorithm does not change indications for use, intended use, energy type, operating principles, or control mechanisms from the predicate. Bench testing on 70 diverse clinical renal studies achieved DICE scores above predefined acceptance criteria, and clinical evaluation by three qualified U.S. readers confirmed acceptable segmentation quality with minimal user interaction. Design verification and risk management identified no new hazards, unexpected results, or adverse effects, demonstrating the device is as safe and effective as the legally marketed predicate.