(53 days)
CORI is indicated for use in surgical knee procedures, in which the use of stereotactic surgery may be appropriate, and where reference to rigid anatomical bony structures can be determined. These procedures include unicondylar knee replacement (UKR) and total knee arthroplasty (TKA).
CORI is indicated for use with cemented implants only.
CORI is a computer-assisted orthopedic surgical navigation and surgical burring system. CORI uses established technologies of navigation, via a passive infrared tracking camera, to aid the surgeon in establishing a bone surface model for the target surgery and in planning the surgical implant location. Based on intraoperatively-defined bone landmarks and known geometry of the surgical implant, CORI aids the surgeon in establishing a bone surface model for the target surgery and planning the surgical implant location.
CORI software controls the cutting engagement of the surgical bur based on its proximity to the planned target surface. The cutting control is achieved with two modes:
- Exposure control adjusts the bur's exposure with respect to a guard. If the surgeon encroaches on a portion of bone that is not to be cut, the robotic system retracts the bur inside the guard, disabling cutting.
- . Speed control regulates the signal going to the drill motor controller itself and limits the speed of the drill if the target surface is approached.
Alternatively, the surgeon can disable both controls and operate the robotic drill as a standard navigated surgical drill.
The acceptance criteria and study proving device performance are described below, based on the provided text for the Real Intelligence Cori (CORI) device.
1. A table of acceptance criteria and the reported device performance
The document mentions "Comprehensive performance testing demonstrated that the system meets required design inputs" and "The comparative results of the cut-to-plan accuracy data is acceptable and equivalent to the predicate devices." However, specific quantitative acceptance criteria and detailed reported performance metrics are not explicitly stated in the provided text. The submission focuses on substantial equivalence to predicate devices (K193120 and K191223) rather than presenting new, specific acceptance criteria with distinct performance numbers for this iteration of CORI.
It can be inferred that the acceptance criteria for accuracy are implicitly tied to demonstrating equivalence to the predicate devices, which are also navigational systems for orthopedic surgery.
2. Sample size used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)
The document primarily describes bench testing and states, "No human clinical testing was required to determine the safety and effectiveness of CORI." Therefore, the concept of a "test set" in the context of clinical data provenance (country, retrospective/prospective) does not directly apply here.
The performance data consisted of "physical performance test for all system components and system accuracy testing." The sample size for these non-clinical tests is not specified.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g. radiologist with 10 years of experience)
Since "No human clinical testing was required" and the testing primarily involved "physical performance test for all system components and system accuracy testing," the concept of "experts" establishing ground truth for a clinical test set is not applicable in the traditional sense. The "ground truth" for accuracy testing would have been established through validated measurement techniques (e.g., precise optical metrology, CMM measurements) under controlled laboratory conditions, not by human expert assessment of medical images or outcomes.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set
Given that clinical data from human subjects was not used for this submission, an adjudication method for a clinical test set is not applicable.
5. If a multi reader multi case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance
No MRMC comparative effectiveness study was conducted, as stated: "No human clinical testing was required." The CORI system is an orthopedic stereotaxic instrument for surgical guidance, not an AI-assisted diagnostic imaging tool that would typically involve human readers.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
The document describes the CORI device as a "computer-assisted orthopedic surgical navigation and surgical burring system" that "assists the surgeon" and provides "software-defined spatial boundaries." It also states, "Alternatively, the surgeon can disable both controls and operate the robotic drill as a standard navigated surgical drill." This indicates that the device is designed to be human-in-the-loop, providing assistance to a surgeon.
However, "Performance data consisted of physical performance test for all system components and system accuracy testing," which implies that the accuracy of the algorithm's guidance and burring control was tested in a standalone, bench setting, separate from a surgeon's subjective usage. The "cut-to-plan accuracy data" would represent the standalone performance of the algorithm and hardware.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)
For the non-clinical "system accuracy testing" and "cut-to-plan accuracy data," the ground truth would have been established through precise, objective physical measurements (e.g., using metrology equipment) of the planned bone resection versus the actual bone resection created by the device on a test medium. This is not expert consensus, pathology, or outcomes data.
8. The sample size for the training set
The document refers to the device as a "computer-assisted orthopedic surgical navigation and surgical burring system" that uses "established technologies." While it mentions "software verification testing" and development in accordance with IEC 62304, there is no specific mention of a "training set" in the context of machine learning. If machine learning models were used, details about their training setup are not provided in this regulatory submission summary. The equivalence is primarily based on the functional and technological similarities to predicate devices and performance in physical accuracy tests.
9. How the ground truth for the training set was established
As no "training set" in the context of machine learning is explicitly mentioned, how its ground truth was established is not provided. If the system is primarily rule-based or model-based on known geometry and intraoperative data rather than a data-driven machine learning model, then a traditional "training set" with established ground truth would not be applicable. The "ground truth" for the system's underlying models would be based on engineering specifications and anatomical models.
§ 882.4560 Stereotaxic instrument.
(a)
Identification. A stereotaxic instrument is a device consisting of a rigid frame with a calibrated guide mechanism for precisely positioning probes or other devices within a patient's brain, spinal cord, or other part of the nervous system.(b)
Classification. Class II (performance standards).