AIRA uses artificial intelligence to analyze CT images and measure kidney volume, offering clinicians an automated tool for renal assessment.
FDA Authorization
| Item | Information |
| Device | Artificial Intelligence Renal Assessment (AIRA) Version 1 |
| Manufacturer | Aramis Global AI Consultancies, LLC |
| Manufacturer Location | Dubai, United Arab Emirates |
| FDA Clearance Date | September 15, 2026 |
| 510(k) Number | K261677 |
| Regulatory Pathway | Traditional 510(k) |
| Classification | Class II |
| Regulation | 21 CFR 892.2050 |
| Regulation Name | Medical Image Management and Processing System |
| Product Code | QIH |
| Medical Specialty | Radiology |
| Predicate Device | Ceevra Reveal 3 (K222676) |
FDA determined that AIRA Version 1 was substantially equivalent to a legally marketed predicate device, permitting its marketing in the United States subject to applicable medical device regulatory requirements.
Device Overview
Artificial Intelligence Renal Assessment (AIRA) is medical imaging software designed to automatically measure kidney volume using computed tomography (CT) images. Kidney size and volume can provide useful information when clinicians evaluate changes in renal anatomy, including kidney enlargement or loss of tissue volume.
AIRA processes CT images to identify kidney structures and calculate renal parenchymal volume, the volume of functional kidney tissue. The software generates measurements for the left and right kidneys and produces a downloadable report. It also creates image overlays showing the identified kidney structures, which medical professionals can review and modify when necessary.
The device is intended for adult patients requiring kidney volume measurements as part of further clinical assessment. It is designed for use by medical professionals, including radiologists, nephrologists, and urologists.
Technology and Innovation
AIRA uses artificial intelligence and deep-learning algorithms to automate image segmentation, a process that identifies and separates anatomical structures within medical images. The software analyzes CT image data, identifies kidney boundaries, and calculates the volume of the renal parenchyma.
Its principal functions include:
- Automated segmentation of kidney structures from CT images.
- Calculation of left and right renal parenchymal volumes.
- Generation of segmentation overlays for visual review.
- Optional manual adjustment of segmentation results and recalculation of measurements.
- Creation of downloadable PDF reports containing kidney volume measurements.
The software operates as Software as a Medical Device (SaMD), using non-adaptive machine-learning algorithms. By automating measurements that would otherwise require manual image analysis, AIRA may help improve the efficiency and consistency of renal volume assessment.
The clearance is another example of AI-enabled software being incorporated into established medical imaging workflows.
Performance Evidence
The 510(k) submission included a retrospective standalone performance study evaluating AIRA’s ability to segment kidney structures accurately. The study collected 119 CT cases from three geographically distinct U.S. clinical sites. Of these, 113 cases were included in the analysis, with the dataset enriched to include at least 25% of cases involving atrophied kidneys.
Two radiologists established the reference segmentations used to evaluate the software, and performance was assessed using the Dice Similarity Coefficient (DSC), which measures how closely the software-generated segmentation matches the reference segmentation. A value of 1.0 represents perfect agreement.
| Performance measure | Right kidney | Left kidney |
| Mean Dice Similarity Coefficient | 0.9092 | 0.9114 |
| Mean Hausdorff Distance | 9.97 mm | 9.94 mm |
| Mean 95th-percentile Hausdorff Distance | 3.77 mm | 3.92 mm |
The results demonstrated substantial agreement between AIRA-generated kidney segmentations and the reference segmentations.
These findings support the software’s ability to perform its intended image-processing and measurement functions. However, the reported study evaluated segmentation performance rather than demonstrating improvements in clinical outcomes or independent diagnostic accuracy.
Regulatory Significance
Several aspects of this clearance are noteworthy for medical device and regulatory professionals.
AI-enabled Software as a Medical Device. AIRA demonstrates the continued use of the traditional 510(k) pathway for AI-enabled medical imaging software. Its algorithms perform defined image-processing and measurement functions rather than independently diagnosing kidney disease.
Predicate device selection. FDA cleared AIRA based on substantial equivalence to Ceevra Reveal 3 (K222676), a previously cleared medical imaging system incorporating machine-learning-based anatomical segmentation. Although the predicate supports broader anatomical applications and imaging functions, both devices share relevant image-processing and measurement technologies.
Software performance validation. The submission relied on quantitative segmentation performance measures, including Dice similarity and Hausdorff distance, supported by retrospectively collected clinical CT images. These methods illustrate how developers can evaluate the accuracy of AI-based anatomical segmentation.
Defined intended use. AIRA’s cleared application is limited to CT-based kidney image processing and volume measurement in adults. The clearance should not be interpreted as authorization for autonomous diagnosis, disease prediction, or other clinical applications outside the stated indications for use.
Software lifecycle and change control. The FDA clearance letter specifically references the Quality Management System Regulation (QMSR), including design controls, nonconforming product, corrective and preventive action, and documentation of design and production changes. The FDA database also indicates that no Predetermined Change Control Plan (PCCP) was authorized with this submission.
For manufacturers developing AI-enabled medical devices, this clearance illustrates the importance of clearly defined intended use, appropriate predicate selection, documented software validation, and ongoing control of software changes.
FDA Sources and Supporting Documentation
FDA 510(k) Premarket Notification Database — K261677
Device classification, applicant information, submission type, and substantial equivalence determination.
Regulatory Snapshot
Device: Artificial Intelligence Renal Assessment (AIRA) Version 1
Manufacturer: Aramis Global AI Consultancies, LLC
Indication: Automated kidney volume measurement from CT images in adult patients
Technology: AI/deep-learning kidney segmentation and renal parenchymal volume calculation
Regulatory Pathway: Traditional 510(k) clearance (K261677), Class II
FDA Clearance: September 15, 2026
Predicate Device: Ceevra Reveal 3 (K222676)
Performance Evidence: Mean Dice similarity of 0.909 (right kidney) and 0.911 (left kidney) across 113 CT cases
Key Limitation: Validated for image segmentation and volume measurement, not autonomous diagnosis or improved clinical outcomes
FDA Panel to Examine Pathogen-Agnostic Sequencing
November 19 advisory committee meeting will explore the scientific, clinical, and regulatory challenges of diagnostic technologies designed to detect known and potentially unknown pathogens.On November 19, 2026, the U.S. Food and Drug Administration (FDA) will convene...
A New Approach for Infectious-Disease Detection
Pathogen-Agnostic Sequencing and the Future of Infectious Disease PreparednessDetecting the Unknown. Connecting the Signals. Preparing for the Future. What if we could recognize an emerging infectious disease before we knew what pathogen was causing it? Advances in...
FDA Approves Wilserdly for Advanced Breast Cancer
Genentech’s oral estrogen receptor antagonist significantly extends progression-free survival, accompanied by FDA authorization of a liquid biopsy companion diagnostic.On November 16, 2026, the U.S. Food and Drug Administration (FDA) approved giredestrant (Wilserdly,...
FDA Pivots from Artificial Intelligence to Super Intelligence
A Regulatory Solution in Search of a ProblemFDA's implementation of a presidential terminology mandate raises questions about scientific accuracy, regulatory consistency, and the responsible use of limited public health resources.Beginning November 16, 2026, the U.S....
Science, Politics, Special Interests, and FDA Independence
Former Insiders Challenge the Administration's Health AgendaFormer FDA Commissioner Marty Makary and former Kennedy adviser Calley Means allege that political and commercial pressures interfered with federal health priorities. Their accounts raise questions about...
When Preventable Diseases Return
Diseases once brought under control through vaccination have returned with devastating consequences. Historical outbreaks in the former Soviet Union and Japan offer important lessons as the United States confronts its largest measles resurgence in decades.For...