(29 days)
The Elekta Medical Linear Accelerator is indicated to assist a licensed practitioner in the delivery of radiation to defined target volumes (e.g. lesions, arterio-venous malformations, malignant and benign tumors), whilst sparing surrounding normal tissue and critical organs from excess radiation for treatment that includes but is not limited to, malignant and benign brain tumors, brain metastases, spine lesions treated using SRS, squamous cell carcinoma of the head and neck, lung, breast, pancreatic, hepatic malignancies treated using SBRT, prostate, and bone metastases.
The Elekta Medical Linear Accelerator system is an image guided Radiation Therapy device to assist a licensed practitioner in the delivery of ionizing radiation to a defined target volume. The system consists of components of the accelerator, such as, beam shaping, with imaging and accessories for patient positioning and set-up to deliver therapeutic treatments. The Elekta Medical Linear Accelerator System is currently available in the following model variants – Precise Treatment System, Elekta Synergy Platform, Elekta Infinity and Versa HD.
This document (K182138) is a 510(k) Premarket Notification for a Medical Linear Accelerator, which is a radiation therapy device. The document primarily focuses on demonstrating substantial equivalence to a predicate device, rather than detailed performance study results against specific acceptance criteria for a new AI/software component.
Therefore, many of the requested details, such as specific acceptance criteria for AI performance, sample sizes for test/training sets, expert qualifications, ground truth establishment methods for a new AI, or MRMC studies, are not applicable or not provided in this document. This submission appears to be about software updates (control software, Integrity™) to an existing device, which mostly focuses on safety and existing functionality, rather than introducing a new AI-powered diagnostic or therapeutic capability that would require such extensive AI performance validation.
However, I can extract the information that is present and explain why other information is missing.
Acceptance Criteria and Device Performance (Based on the provided document)
Since this submission is for software changes to an existing medical linear accelerator and focuses on substantial equivalence, the "acceptance criteria" discussed are primarily related to general device safety, performance, and adherence to regulatory standards, rather than specific diagnostic or therapeutic efficacy metrics that would be evaluated for a novel AI algorithm.
| Acceptance Criteria (Inferred/General) | Reported Device Performance |
|---|---|
| Conformance to FDA Quality System Regulation (21 CFR §820) | Met - "Testing in the form of module, integration and system level verification was conducted in accordance with FDA Quality System Regulation (21 CFR §820)" |
| Conformance to ISO 13485 Quality Management System standard | Met - "ISO 13485 Quality Management System standard" |
| Conformance to ISO 14971 Risk Management Standard | Met - "ISO 14971 Risk Management Standard" |
| Conformance to IEC 62304 Software life-cycle processes | Met - "IEC 62304 Software life-cycle processes" |
| Conformance to FDA recognised consensus standards (e.g., IEC 60601-1, etc.) | Met - "and the other FDA recognised consensus standards which includes but is not limited to IEC 60601-1, IEC 60601-2-1, IEC 60601-1-6, IEC 62366-1." |
| Software Verification Testing (for Major Level of Concern - Class C) | Met - "Software verification testing was conducted and documented in accordance with FDA's 'Guidance for the Content of Premarket Submissions for Software Contained in Medical Devices' for devices that pose a major level of concern (Class C per IEC 62304)." |
| Basic safety and essential performance | Met - "Basic safety and essential performance have been satisfied through conformance with device-specific recognised consensus standards, as well as applicable general and collateral safety and essential performance standards for medical devices." |
| Conformance to applicable technical design specifications | Met - "Results from verification and validation testing demonstrate that conformance to applicable technical design specifications have been met" |
| Safety & Effectiveness (overall) | Met - "...and safety & effectiveness has been achieved." and "The results of verification, validation and safety standard testing demonstrate that the Elekta Medical Linear Accelerator system is substantially equivalent to their predicate device." This refers to overall device safety and effectiveness as being substantially equivalent to the cleared predicate, not new performance for a novel AI. |
| Tighter error detection (for the new Integrity™ software) | Achieved - "Elekta has introduced changes to the control software, Integrity™, primarily to provide tighter error detection..." (This is a design goal for the software change, not a quantified acceptance criterion here). |
| Merged codebase to support current hardware platform (for Integrity™) | Achieved - "...and to merge the codebase to provide a single release that supports the current hardware platform." |
Detailed Study Information (Based on the Provided Document):
-
A table of acceptance criteria and the reported device performance:
- As detailed in the table above, the acceptance criteria are generally qualitative and refer to adherence to regulatory standards (e.g., QSR, ISO standards, IEC standards) and the successful completion of verification and validation testing. The "reported device performance" is a confirmation that these standards were met and testing was successfully completed, leading to a determination of substantial equivalence.
- Specific quantitative metrics for an AI algorithm's performance (e.g., AUC, sensitivity, specificity for a diagnostic task) are not present because the submission is for software updates to a linear accelerator, not a new AI diagnostic/therapeutic algorithm.
-
Sample sizes used for the test set and the data provenance:
- N/A. This document describes "module, integration and system level verification" and "validation of the integrated system under clinically representative conditions." This is typical for a software update to an existing medical device. It does not mention a "test set" in the context of an AI algorithm's performance evaluation on a separate dataset of patient cases.
- Data Provenance: Not specified as it's not a dataset-driven AI validation. The testing is described generally as "clinically representative conditions."
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- N/A. Ground truth establishment for an AI test set is not applicable here as the submission is for control software updates, not AI performance validation.
- Validation was performed by "competent and professionally qualified personnel," but their specific number or qualifications are not provided for the purpose of "ground truth" establishment in an AI context.
-
Adjudication method (e.g. 2+1, 3+1, none) for the test set:
- N/A. Not applicable for this type of submission.
-
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. This type of study is not mentioned as it's not relevant to the nature of this submission (software update for a linear accelerator control system, not an AI assisting human interpretation).
-
If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
- N/A. No specific AI algorithm performance in a standalone capacity is evaluated or described. The software changes are to the control system of a device that assists a licensed practitioner.
-
The type of ground truth used (expert consensus, pathology, outcomes data, etc):
- N/A. The concept of "ground truth" in the context of an AI algorithm's performance is not discussed here. The validation described is against technical design specifications and safety/performance standards for a medical device's control system.
-
The sample size for the training set:
- N/A. No AI training set is mentioned or implied for this submission. The software changes are described as "tighter error detection" and codebase merging, which likely comes from software development and testing cycles rather than machine learning training.
-
How the ground truth for the training set was established:
- N/A. Not applicable as there is no mention of an AI training set. Grounds for validation are based on established engineering principles, regulatory standards, and clinical representativeness as opposed to a data-driven AI training methodology.
§ 892.5050 Medical charged-particle radiation therapy system.
(a)
Identification. A medical charged-particle radiation therapy system is a device that produces by acceleration high energy charged particles (e.g., electrons and protons) intended for use in radiation therapy. This generic type of device may include signal analysis and display equipment, patient and equipment supports, treatment planning computer programs, component parts, and accessories.(b)
Classification. Class II. When intended for use as a quality control system, the film dosimetry system (film scanning system) included as an accessory to the device described in paragraph (a) of this section, is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 892.9.