The Predicate Ships Hardware. You Don't Have To.

An analysis of all 1,483 AI/ML-flagged records in the Innolitics FDA Device Explorer found 88 AI software-as-a-medical-device clearances that cite a predicate which is not SaMD, and 50 of those are cases where a software-only device was cleared against a physical instrument, console, or hardware-integrated system. The oldest such crossing cleared in 1998 with the Neurometric Analysis System (K974748), and 41 of the 50 cleared since 2020, including CADDIE (K240044), which cited the hardware-integrated GI Genius (DEN200055) as its predicate. The finding matters because a software-only pathway can be roughly 10 to 20 times faster to market than a comparable software-hardware combination, avoiding electrical safety testing, biocompatibility, PCB design, manufacturing, and supply-chain work.

Did you know that even real-time intraoperative endoscopy video AI can be a software-only medical device? In fact, at least 88 AI software-as-a-medical-device clearances cite a predicate that is not SaMD, and 50 of them are cases where a software-only device is predicated on a physical instrument, console, or hardware-integrated system. Seventeen of those operated live, during the procedure or at the bedside: colonoscopy, atrial fibrillation ablation, robotic surgery, ICU hemodynamics. GI Genius DEN200055 https://fda.innolitics.com/device/DEN200055 , the De Novo that created real-time colonoscopy CADe, is a hardware-integrated video box wired into the endoscopy stack. CADDIE K240044 https://fda.innolitics.com/device/K240044 cited it as a predicate and cleared as a cloud-hosted SaMD, adding a cecal landmark feature on top of the same real-time polyp detection claim. FDA treated the missing box as a technological difference, addressed in the comparison table, rather than as a different device. We screened all 1,483 AI/ML-flagged records in the Innolitics FDA Device Explorer https://fda.innolitics.com/ to find every case like it, then split the results by clinical domain to see where the play is still needed. A software-only device is generally faster and easier to bring through FDA clearance and gives you far more distribution flexibility than a software–hardware combination. You can update and improve the software without redesigning, manufacturing, or revalidating the hardware. You also avoid much of the additional work and cost associated with electrical safety testing, biocompatibility, PCB design, manufacturing, and supply-chain management. As a rough planning estimate, the pathway to market for a software-only device may be 10 to 20 times faster than for a comparable software–hardware combination. The exact difference depends on the device, its intended use, and the evidence FDA requires, but the strategic advantage is substantial: less hardware to build and test, fewer operational constraints, and more ways to distribute and update the product. Of the 1,483 AI/ML records, 985 are flagged SaMD and 951 cite at least one predicate. We resolved all 738 unique cited predicate records against the full corpus and kept the pairs where the predicate is not SaMD: 88 clearances, 120 citations. We verified each one using a panel of AI models and kept only the strict cases: a software-only subject citing a hardware/software combo predicate; this left 50 clearances spanning 1998 to 2026. The 34 records with no predicate were De Novos or PMAs yes you can predicate off of a PMA if it gets down-classified to class II , and the three whose citations a parser would have missed, are accounted for in the Method section. The oldest crossing in the set cleared in 1998, when the Neurometric Analysis System K974748 https://fda.innolitics.com/device/K974748 predicated a bunch of EEG machines back when split predicates were still very much in style . 41 of the 50 cleared since 2020. Each of the 50 clearances makes one of four arguments. GI Genius hardware and software seeded four software-only children: SKOUT K213686 https://fda.innolitics.com/device/K213686 , Fujifilm's EW10-EC02 K230751 https://fda.innolitics.com/device/K230751 , Wision AI's EndoScreener K211326 https://fda.innolitics.com/device/K211326 , and CADDIE K252586 https://fda.innolitics.com/device/K252586?doc=summary . Volta's VX1 K201298 https://fda.innolitics.com/device/K201298 cleared as standalone software that flags electrogram dispersion live during ablation, citing the CARTO XP electroanatomical mapping console K093566 https://fda.innolitics.com/device/K093566 , an entire navigation platform. Medtronic's Touch Surgery Aide tracks instruments in real time during robotic-assisted surgery K253984 https://fda.innolitics.com/device/K253984 against the HyperSnap Surgical System K250268 https://fda.innolitics.com/device/K250268 . Latency, interface, and failure-mode questions were answered in the software documentation. Twelve sponsors split the algorithm out of their cleared device and refiled it as SaMD, citing their own hardware. Itamar turned the WatchPAT wrist device K250460 https://fda.innolitics.com/device/K250460 into WatchPAT SW K254042 https://fda.innolitics.com/device/K254042 , a cloud scoring service. Nox did it with DeepRESP K241960 https://fda.innolitics.com/device/K241960 , Leica with the Aperio scanner's new artifact-detection AI K253561 https://fda.innolitics.com/device/K253561 , GE with its CT reconstruction module K212067 https://fda.innolitics.com/device/K212067 . Once the algorithm is a separate device, it iterates on a software timeline, sells without a hardware refresh cycle, and can carry a PCCP, as Ceribell's unbundled seizure software did K241589 https://fda.innolitics.com/device/K241589 . The largest group, 17 clearances, moved the sensor problem onto hardware someone else ships. Healthy.io predicated a home smartphone urinalysis test on a desktop lab analyzer K210069 https://fda.innolitics.com/device/K210069 citing K141874 https://fda.innolitics.com/device/K141874 . NeuroRPM K221772 https://fda.innolitics.com/device/K221772 and Rune Labs K213519 https://fda.innolitics.com/device/K213519 both predicated a proprietary wrist-worn Parkinson's logger and run on an Apple Watch. Butterfly predicated Verathon's dedicated BladderScan hardware for a software bladder-volume tool K200980 https://fda.innolitics.com/device/K200980 . Alife's Embryo Predict K250781 https://fda.innolitics.com/device/K250781 cited a time-lapse microscopy system and processes plain images in the cloud. In each of these files, FDA asked for performance data on the inputs the device actually receives. Twelve clearances replaced a physical measurement with an inference from data the hospital already collects. Riverain's SoftView K092363 https://fda.innolitics.com/device/K092363 produced bone-suppressed chest images from one standard exposure, citing the two-exposure dual-energy option K013481 https://fda.innolitics.com/device/K013481 it replaced. Lifescreen Apnea derived respiration from Holter ECG, citing a bedside multi-sensor sleep unit K042745 https://fda.innolitics.com/device/K042745 , in 2005. CLEW predicated Edwards' arterial-waveform hemodynamic platform for an ICU instability predictor fed by EHR and vitals data K200717 https://fda.innolitics.com/device/K200717 . The precedent for computing a measurement instead of sensing it goes back at least to 2005. The 50 crossings are not spread evenly. Split the 985 AI SaMD records by clinical domain and two regimes appear. In hemodynamic monitoring, EEG, sleep, and endoscopy, many AI SaMD with a predicate cited hardware. These are hardware-native domains: the cleared devices are monitors, headbands, wrist units, and video processors, so the software lineage is thin, and predicating the box is one way to go software-only in a hardware-filled world. In ultrasound, CT and MR processing, radiotherapy, dental, and ophthalmology, the share is much lower, with software-to-software predication being much more common. Those domains already have deep software lineages, and sponsors there predicate software rather than scanners. Sometimes, the De Novo author must get cleared as a software-hardware combination until the hardware platform can be “abstracted away” while still mitigating risks. My hypothesis is that these types of devices start out as software-hardware combinations because standardization amongst hardware vendors has not yet reached a point where software-only devices can argue hardware agnosticism. A similar trajectory has also happened in digital pathology. GI Genius went De Novo as a box, and has spawned 19 software-only descendants devices. On the other hand, not all De Novos need to go hardware initially. Caption Guidance DEN190040 https://fda.innolitics.com/device/DEN190040 went De Novo as software running on third-party probes, and its 8 descendants never touched a hardware predicate. Both lineages cover the same real-time, in-procedure claim class and inherited opposite form factors. Of 101 ultrasound AI SaMD records, 95 cite a predicate and 5 are verified hardware crossings; 62 citations are cross-company software predicates. In 12 clearances the citation ran the other way: probe and console makers predicated a startup's pure-software device. Clarius cited DiA K222406 https://fda.innolitics.com/device/K222406 and Sonio K233955 https://fda.innolitics.com/device/K233955 , Exo cited DiA K232501 https://fda.innolitics.com/device/K232501 , GE's Automated Aortic Stenosis cited Ultromics K254161 https://fda.innolitics.com/device/K254161 . Real-time acquisition guidance runs from Caption's software De Novo through UltraSight K223347 https://fda.innolitics.com/device/K223347 , HeartFocus K242807 https://fda.innolitics.com/device/K242807 , and ThinkSono K260338 https://fda.innolitics.com/device/K260338 with no hardware predicate in the chain. All 46 dental AI SaMD records cite a predicate and none cite hardware. The root is a 1998 PMA for caries-detection software, Logicon P980025 https://fda.innolitics.com/device/P980025 , which Pearl K210365 https://fda.innolitics.com/device/K210365 , Videa K213795 https://fda.innolitics.com/device/K213795 , and Overjet K212519 https://fda.innolitics.com/device/K212519 all cited in 2022. Four years later the imaging OEMs predicate the startups: Dentsply Sirona cites Videa K253009 https://fda.innolitics.com/device/K253009 , Nobel Biocare cites Pearl and Videa K221921 https://fda.innolitics.com/device/K221921 . Overjet reached outside the domain entirely, predicating an MRI denoiser for its dental image-enhancement tool K241681 https://fda.innolitics.com/device/K241681 . Reconstruction and denoising split along the raw-data line. Algorithms that touch projection data stayed with the scanner maker and cleared by unbundling: GE's Deep Learning Image Reconstruction K212067 https://fda.innolitics.com/device/K212067 , United Imaging's HYPER DLR against its own PET/CT systems K193210 https://fda.innolitics.com/device/K193210 , GE's Clarify DL against the StarGuide SPECT gantry K241350 https://fda.innolitics.com/device/K241350 . Algorithms that work on the DICOM image went vendor-neutral and predicate software: SubtlePET K182336 https://fda.innolitics.com/device/K182336 in 2018, Airs Medical's SwiftMR citing SubtleMR K210999 https://fda.innolitics.com/device/K210999 , Foqus citing SwiftMR K241982 https://fda.innolitics.com/device/K241982 . Across 46 reconstruction and denoising SaMD records, every hardware predicate belongs to an OEM. If your model consumes images rather than sinograms or k-space, you are in the software regime. Digital pathology followed the same pattern as ultrasound. Paige Prostate went De Novo as software DEN200080 https://fda.innolitics.com/device/DEN200080 , and Ibex's Galen Second Read predicated it K241232 https://fda.innolitics.com/device/K241232 . The lone hardware crossing is Leica unbundling a QC algorithm from its own Aperio scanner K253561 https://fda.innolitics.com/device/K253561 . Scanner vendors own the acquisition; the diagnostic software lineage is finally starting to become vendor-neutral. The age of VNAs for digital pathology is upon us, and with it, AI SaMD will go mainstream. The 50 SiMD to SaMD substantial equivalence arguments share a similar structure: hold the intended use and claim type steady, and the missing hardware becomes one row of different technological characteristics. What fills the gap is standalone performance on the real input distribution, interface and compatibility documentation for the hardware you now depend on, latency and failure-mode analysis if the output is real-time, and cybersecurity. CADDIE’s K240044 https://fda.innolitics.com/device/K240044 reviewers accepted a software-only cloud architecture for a live in-procedure claim. The obvious objection to CADDIE is latency. A sponsor does not control cloud compute or the hospital's network, and a colonoscopy CADe has to draw its box on a live video feed. The K240044 https://fda.innolitics.com/device/K240044 summary shows how Odin Vision answered the objection, and the answer is architectural rather than evidentiary: the latency-sensitive path never leaves the room, and performance degrades gracefully to standard of care anyway. A small computer on the endoscopy stack sits between the video processor and the monitor. It passes the live feed through to the display with a fixed local pipeline delay, encodes a copy, streams that copy to the cloud for GPU inference, and composites whatever detections have come back onto whichever frame is currently on screen. The special controls in 21 CFR 876.1520 https://www.ecfr.gov/current/title-21/chapter-I/subchapter-H/part-876/subpart-B/section-876.1520 ask for exactly two latency measurements: real-time video delay due to the device, and video delay due to marker annotation. CADDIE reported 32.50 ms for the first on both the Olympus CV-1500 and CV-190, and 32.59 ms and 33.61 ms for the second. That design converts cloud latency from a display-latency problem, which FDA would have to scrutinize, into a detection-timing problem, which the labeling already owns. If the inference round trip is delayed, the polyp detection box lands a few frames late on a polyp that is still on screen. If the connection drops, boxes stop appearing but, crucially, the video keeps running. The failure mode degrades toward unassisted colonoscopy, the baseline standard of care, and the indication makes the gastroenterologist responsible for confirming every finding anyway. This would have been a completely different story if they claimed to be fully autonomous. Inference latency distribution, network jitter, and offline behavior are not in the special controls, and the summary does not characterize them. Bounding the network became a site-installation i.e., post-market quality system DHR matter rather than a premarket regulatory matter. In summary, relying on third-party implementations you do not fully control in a medical device adds risk—that much is undeniably true. However, FDA is not expecting devices with zero risk. They expect you to control risks to an acceptable level where the benefits outweigh the risks. Yes. At least 88 AI SaMD clearances cite a non-SaMD predicate, and 50 are strict hardware-to-software crossings we verified by hand. FDA evaluates intended use and technological characteristics, and a missing enclosure is a technological characteristic addressed in the comparison table. No. A predicate citation does not require the predicate holder's agreement. Only the second half. Your domain already has a deep software lineage, so the efficient predicate is another software device, often from a company that is not your competitor. Use the hardware-predicate plays if a suitable software only predicate does not exist. Source: Innolitics FDA Device Explorer https://fda.innolitics.com/ , AI/ML-flagged cohort, N = 1,483 records as of August 31, 2026. We flagged the 985 SaMD records; 951 cite at least one predicate, and the 34 that do not are accounted: 28 De Novos and 2 PMAs carry no predicate by definition, and 4 records had no machine-readable predicate in their summaries. Citations in nonstandard formats lowercase K numbers, spaces inside DEN numbers were recovered by AI; one recovered record, Sight Diagnostics' X100, was then excluded because the subject ships a benchtop analyzer. We resolved all 738 unique cited predicates against the corpus and kept subject-predicate pairs where the predicate is not SaMD 88 clearances, 120 citations . Each pair was reviewed against the intended use and predicate-change summaries; we excluded subjects that ship hardware, and predicates that are freestanding software, keeping hardware-integrated predicates whose function depends on dedicated acquisition hardware. That left 50 verified crossings. Domains were assigned by keyword matching on device name and intended use, so domain counts are approximate and a record belongs to one domain only. The screen is deliberately conservative and not intended to be an exhaustive list, but rather intended to add support to SIMD-to-SAMD prediction. Every K number, DEN number, and PMA number above links to the primary record in our FDA browser tool.