Search Results

Leksell Gamma Knife® is a teletherapy device intended for stereotactic irradiation of head structures ranging from very small target sizes of a few millimeters to several centimeters e.g.

• Metastatic tumors
• Recurrent glioblastomas
• Trigeminal neuralgia
• Medically refractory essential tremor
• Orbital tumors
• Ocular tumors
• Optic nerve tumors
• Benign diseases (such as meningiomas, vestibular schwannomas, post-surgical pituitary adenomas, craniopharyngioma, hemangioblastomas, schwannomas, arteriovenous malformations, cavernous malformations, chordomas, glomus tumors, hemangiomas)
• Skull base tumors
• Head and neck tumors (such as unknown primary of the head and neck, oral cavity, hypopharynx, oropharynx, nasopharynx, sinonasal, salivary gland)
• Pediatric tumors (such as glioma, ependymoma, pituitary tumors, hemangioblastoma, craniopharyngioma, meningioma, metastasis, medulloblastoma, nasopharyngeal tumors, arteriovenous malformations, cavernous malformations, skull base tumors)
• Refractory mesial temporal lobe epilepsy in adults
• Refractory epilepsy associated with structural changes such as hamartomas, cerebral cavernous malformations, and arteriovenous malformations (adult and pediatric)

Leksell Gamma Knife® (available models Elekta Esprit, Icon™, and Perfexion™) is a radiosurgery system for use in the stereotactic irradiation of head structures. Surgery is achieved by delivering a prescribed dose as one or more shots of ionizing radiation to the exact site of the target.

The provided FDA 510(k) clearance letter pertains to the Leksell Gamma Knife® (Elekta Esprit, Icon™, Perfexion™) and describes the basis for its substantial equivalence to a predicate device (K222047).

It's important to note that the document outlines the non-clinical performance testing and explicitly states that no animal or clinical tests were performed to establish substantial equivalence for this specific 510(k) submission. This is because the submission primarily focuses on the replacement of obsolete hardware components and a new control system software version to support these components, along with the addition of new indications (Refractory mesial temporal lobe epilepsy in adults and Refractory epilepsy associated with structural changes) that are deemed to fall within the existing intended use and have a favorable benefit-risk profile based on established knowledge of Gamma Knife treatment for certain neurological conditions.

Therefore, the typical structure for describing acceptance criteria and a study proving device performance (especially for AI/ML-driven devices with specific performance metrics like sensitivity, specificity, etc.) is not fully applicable in this context. The acceptance criteria here are more related to functional performance, safety, and maintaining equivalence with the predicate device after hardware and software updates.

However, I can extract the information relevant to what testing was performed and how the device's continued safety and effectiveness were assured.

Acceptance Criteria and Study to Prove Device Meets Acceptance Criteria

Given that this 510(k) focuses on hardware replacement and a software update to support those replacements, rather than the introduction of a novel AI/ML algorithm with specific diagnostic performance metrics, the acceptance criteria are not in the form of typical clinical performance measures (e.g., sensitivity, specificity). Instead, they revolve around functional validation, safety, and maintenance of equivalence to the predicate device.

Key Acceptance Criteria (Inferred from the document):

• Functional Equivalence: The new hardware and software must maintain the same fundamental scientific technology, clinical workflow, and functionality as the predicate device.
• Safety: The device must continue to meet established safety standards and not introduce new or increased risks.
• Performance: The updated system must perform as intended, ensuring accurate and precise stereotactic irradiation.
• Usability: User interface and labeling updates must support safe and effective use.
• Biocompatibility: All new materials in contact with patients must be biocompatible.

Study that Proves the Device Meets the Acceptance Criteria:

The "study" described is a comprehensive set of non-clinical verification and validation activities.

1. Table of Acceptance Criteria and Reported Device Performance

• "The subject device does not introduce any new clinical requirements or functions."
• "The fundamental technical characteristics... have not changed and are substantially equivalent."
• "The clinical workflow, safety, functionality, and performance is unchanged thus ensuring substantial equivalence." |
  | Safety (Risk Control & Compliance with Standards) | - "The change in design required a new risk control to be added to ensure the same level of safety as predicate device and thus ensuring substantial equivalence."
• Verification testing performed to IEC 60601-1 series, IEC 62304, ISO 10993 series. |
  | Performance (Accuracy & Consistency) | - "Testing in the form of module, integration and system level verification was performed to evaluate the performance and functionality of the new control system and the new hardware against requirement specification."
• "Regression test of unchanged functionalities... to ensure that new and updated functionalities did not introduce any undesirable effects." |
  | Usability | - "Performance testing, including design and usability validation of the system, has been performed by competent and professionally qualified personnel to ensure that the product fulfils the intended use and user needs."
• "The design and usability validation was also made to ensure that the risk control measures associated with functions related to safety (FRS) for the affected functionality... were effective."
• "Updated labeling was in scope of usability evaluation according to IEC 62366-1. The result demonstrate that the labeling support the safe and effective use of the device." |
  | Biocompatibility | - "Biocompatibility testing has been performed to ensure that all materials are biocompatible in relation to the use of the device." |
  | Substantial Equivalence to Predicate Device (Overall) | - "The performance data demonstrate that the subject device is as safe and effective and performs as well as the predicate devices (K222047)."
• "The result of verification and validation as well as conformance to relevant safety standards demonstrate that the Leksell Gamma Knife® meets the established safety and performance criteria and is substantially equivalent to the predicate device." |

2. Sample Size Used for the Test Set and Data Provenance

• Sample Size: Not explicitly stated as a number of "cases" or "patients" in the context of clinical data. The testing involved system-level verification, module testing, integration testing, and regression testing. This implies testing across the functionality of the device.
• Data Provenance: The nature of the testing is non-clinical performance and functional validation. It's conducted by Elekta Solutions AB (Sweden) as part of their design control and quality system processes. It's not based on retrospective or prospective clinical patient data for efficacy or diagnostic performance.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts

• Experts: Not applicable in the sense of clinical experts establishing "ground truth" for disease diagnosis or outcome.
• Qualifications: For performance, design, and usability validation, the document states: "Performance testing, including design and usability validation of the system, has been performed by competent and professionally qualified personnel." No specific number or precise qualifications (e.g., "radiologist with 10 years of experience") are provided within this public summary, as this refers to internal quality and engineering validation teams.

4. Adjudication Method for the Test Set

• Adjudication Method: Not applicable. The testing is focused on system functionality, safety, and performance against defined engineering requirements and standards, not on interpreting images or making clinical judgments requiring adjudicated consensus.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was done

• MRMC Study: No. The document explicitly states: "No animal or clinical tests were performed to establish substantial equivalence with the predicate device." Therefore, no MRMC study comparing human readers with and without AI assistance was conducted for this specific 510(k) submission.

6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) was done

• Standalone Performance: No. This device is a radiation therapy system, not an AI diagnostic algorithm. While it has control software, "standalone performance" in the context of AI metrics like sensitivity/specificity for disease detection is not relevant here. The software's performance is integral to the entire system's functionality and safety.

7. The Type of Ground Truth Used

• Ground Truth: For non-clinical validation, the "ground truth" is typically defined by:
  • Requirement Specifications: The device's performance is measured against its designed functional and safety requirements.
  • Established Standards: Adherence to recognized national and international standards (e.g., IEC 60601, IEC 62304, ISO 10993, IEC 62366-1).
  • Predicate Device Equivalence: The gold standard is the documented performance and safety profile of the legally marketed predicate device (K222047).

8. The Sample Size for the Training Set

• Training Set Sample Size: Not applicable. This 510(k) does not describe the development of a novel machine learning algorithm that requires a distinct training set in the AI sense. The software updates are for system control and enabling new hardware, not for learning from vast datasets.

9. How the Ground Truth for the Training Set was Established

• Ground Truth for Training Set: Not applicable, as there is no mention of a training set for an AI/ML algorithm.

Ask a specific question about this device

Leksell GammaPlan® is a computer-based system designed for Leksell Gamma Knife® treatment planning.

Leksell GammaPlan® is a powerful, computer-based treatment planning system specifically designed for the simulation and planning of stereotactic Leksell Gamma Knife® radiosurgery based on tomographic and projectional images.

The basis of treatment planning is the acquisition and processing of digital images by a computer workstation running the treatment planning application software. The program is capable of handling a range of different imaging modalities. Images from tomographic sources such as Computer Tomography (CT), Magnetic Resonance (MR) and Positron Emission Tomography (PET) scanners can be used as well as projectional images from angiograms (AI). This allows the direct comparison between vascular structures in projectional images and tissue structures in CT and MR.

Digital images can be imported into the system via the computer network.

The treatment planning application has the ability to plan a patient's treatment protocol based on a single target or multiple targets.

The basic elements of treatment planning are:

• defining the cranial target or targets .
• . devising the configuration of the collimators to be used during treatment
• determining the parameters of the radiation shots to be delivered by Leksell • Gamma Knife®.

The provided FDA 510(k) summary for the Leksell GammaPlan® (LGP) (K232854) does not contain specific details regarding acceptance criteria and the comprehensive study results in the way requested for a typical AI/ML device.

This submission is for an updated version of a treatment planning system, which is generally software for medical device operation, rather than an AI/ML algorithm that performs autonomous diagnostic or prognostic tasks. As such, the performance testing focuses on software verification and validation against requirements, rather than clinical performance metrics in terms of sensitivity, specificity, etc., with a ground truth established by experts.

However, I can extract the information that is present and indicate where the requested information is not available in the document.

1. A table of acceptance criteria and the reported device performance

2. Sample size used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)

• Sample Size for Test Set: Not specified in the document. The testing described is software verification and validation, not a clinical trial with a "test set" in the sense of patient data for AI model evaluation.
• Data Provenance: Not specified, as it's not a clinical data-driven study. The document mentions "clinically equivalent device" for validation, implying internal testing and possibly simulated or anonymized data, but details are absent.

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)

• Number of Experts: Not specified.
• Qualifications of Experts: The document mentions "competent and professionally qualified personnel" performed design and usability validation, but no specific professional qualifications (e.g., radiologist, medical physicist) or experience levels are provided. Ground truth in the context of a treatment planning system primarily relates to the accuracy of computational models and adherence to clinical guidelines, rather than expert interpretation of images for diagnosis.

4. Adjudication method (e.g. 2+1, 3+1, none) for the test set

• Adjudication Method: Not specified. It's unlikely this type of adjudication was performed, as the evaluation focused on software functionality and adherence to specifications, not on resolving disagreements in expert clinical assessment.

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

• MRMC Study: No, an MRMC study was not performed. This device is a treatment planning system, not an AI diagnostic or assistance tool intended to improve human reader performance in interpreting medical images. The document states, "No animal or clinical tests were performed to establish substantial equivalence with the predicate device."

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done

• Standalone Performance: The device itself is software for treatment planning. Its "standalone" performance would be its ability to correctly calculate and display treatment plans based on inputs, which is covered by the mentioned verification and validation testing. However, this is not "standalone AI algorithm performance" in the typical sense of a diagnostic AI. The system is inherently "human-in-the-loop" as it requires a clinician to define targets and parameters.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)

• Type of Ground Truth: For the type of software described (treatment planning), the "ground truth" for verification and validation typically involves:
  • Specification Compliance: Verification that the software performs according to its written requirements and design specifications.
  • Industry Standards: Conformance to relevant medical device software (e.g., IEC 62304) and medical physics standards.
  • Phantom/Physical Measurement Comparisons: For dose calculations, comparison with physical measurements in phantoms or known analytical solutions (though not explicitly detailed for this specific 510(k)).
  • Clinical Equivalence/Usability: Validation that the software supports intended clinical use, as performed by "competent and professionally qualified personnel."
    The document does not specify "expert consensus," "pathology," or "outcomes data" as ground truth for this submission, as these are more relevant for diagnostic or prognostic AI.

8. The sample size for the training set

• Sample Size for Training Set: Not applicable and not mentioned. This device is a treatment planning system, not a machine learning model that requires a training set in the conventional sense.

9. How the ground truth for the training set was established

• Ground Truth for Training Set Establishment: Not applicable and not mentioned, as there is no machine learning "training set" for this device.

Ask a specific question about this device

Elekta Unity using Magnetic Resonance Imaging is indicated for radiation therapy treatments and stereotactic radiation treatments of malignant and benign diseases anywhere in the body as determined by a licensed medical practitioner in accordance with a defined treatment plan.

Elekta Unity is intended for use with compatible Treatment Planning and Oncology Information Systems. The Elekta Unity 1.5T MR scanner is a magnetic resonance imaging system that produces cross-sectional images in any orientation of the internal structure of the whole body before, during, and after the radiotherapy treatment.

When interpreted by a trained physician magnetic resonance images acquired before, during, and after the radiotherapy treatment yield information that can be useful in diagnosis and may assist therapy planning, patient positioning and treatment delivery related to radiation oncology.

Elekta Unity is a multifunctional digital linear accelerator designed to assist licensed medical practitioners in the delivery of ionizing radiation to defined volumes (e.g. malignant and benign tumors). Elekta Unity is capable of both intensity modulated radiation therapy (IMRT) and image guided radiation therapy (IGRT).

The Elekta Unity 1.5T MRI scanner is a magnetic resonance imaging sub-system that produces cross-sectional images in any orientation of the internal structure of the whole body before, during, and after the radiotherapy treatment.

When interpreted by a trained physician the images acquired before, during, and after the radiotherapy treatment vield information that can be useful in diagnosis and may assist therapy planning, patient positioning, and treatment delivery related to radiation oncology.

In addition to the motion monitoring with manual interrupt previously cleared with predicate Unity devices, the subject device Elekta Unity has added the following motion management strategies:

• · Anatomic Position Monitoring (APM) with Manual interrupt (also referred to as True Tracking as it 'tracks' in 3D and enables the system to gate using the 3D position of the anatomy's motion)
• · APM with Anatomic Tolerance Check (ATC)
• Adaptive Therapy with optional Baseline Shift (BLS) Recovery.

These modifications are not contained solely within the Elekta Unity Device, as the full clinical benefit is achieved with interoperability of Unity, Monaco RTP, and MOSAIQ OIS.

The provided text is a 510(k) summary for the Elekta Unity MR Linac and describes updates to an existing device rather than a new standalone AI/ML device with a clinical performance study. Therefore, a direct answer to all points regarding "acceptance criteria and the study that proves the device meets the acceptance criteria" in the context of an AI/ML algorithm's clinical performance cannot be fully provided from this document.

Specifically, the document states: "No animal or clinical tests were performed to establish substantial equivalence with the predicate device." This indicates that a human-in-the-loop (MRMC) or standalone (algorithm only) clinical performance study as typically understood for AI/ML diagnostic or treatment planning algorithms was not conducted for this 510(k) submission. The clearance is based on substantial equivalence by demonstrating that the new motion management strategies do not change the fundamental scientific technology or raise different questions of safety or effectiveness.

However, I can extract the information relevant to the device's technical specifications and the non-clinical testing performed to establish substantial equivalence.

Here's a breakdown of the available information based on the provided document:

1. Table of Acceptance Criteria and Reported Device Performance

The document does not present "acceptance criteria" in the traditional sense of a clinical performance study against a ground truth. Instead, it provides a comparison of "Technological Characteristics" between the subject device (Elekta Unity with new features) and its predicate device (Elekta Unity K212114). The "acceptance" here relates to demonstrating that the new features do not negatively impact the established performance and safety characteristics, or that they meet expected performance defined by engineering specifications.

Ask a specific question about this device

MOSAIQ® is an oncology information system used to manage workflows for treatment planning and delivery. It supports information flow among healthcare facility personnel and can be used wherever radiotherapy and/or chemotherapy are prescribed. Users can configure MOSAIQ® for Medical Oncology use, Radiation Oncology use, or the two together. Medical oncology dose calculations are designed to support both adult and pediatric patients. It lets users:

• · Assemble electronic patient charts and treatment plans, order diagnostic tests, and prescribe medications.
• · Generate and keep medication formulary lists and calculate applicable medication dosages for medical oncology.
• · Import, view, annotate, adjust, enhance, manage and archive images.
• · Compare radiation treatment plans and evaluate dose coverage.
• · Design leaf plans for operation with radiotherapy treatment machines that have multi-leaf collimators.
• · Make sure radiation treatment plans imported from treatment planning systems agree with treatment machine constraints. MOSAIQ® reads actual settings from the treatment machine communication interface. It compares these settings with predefined values. If a mismatch occurs between the planned values and the actual machine settings, the system warns the user.
• · View reference images to setup treatment. MOSAIQ® refers to predefined settings to help treatment machine setup and communicates patient and machine setup instructions.
• · Record actual delivered radiation values in an electronic chart to track treatment.
• · Use stereotactic localization to calculate set-up coordinates for treatments.
• · Monitor Intrafractional motion with real time image overlay.

MOSAIQ® is not intended for use in diagnosis.

The MOSAIQ® Oncology Information System (OIS) is an image-enabled electronic medical record system. It manages clinical and administrative workflows within oncology departments and facilitates efficient patient care. It can be configured for Medical Oncology, Radiation Oncology, or both.

The Medical Oncology (MO) configuration is a medical oncology charting solution that includes customizable regimens (Care Plans) that automate chemotherapy orders for labs, procedures, and appointments. Configurable flowsheet views are used for reviewing treatment administration, documents, assessment and lab data. Users can enter medications and screen for drug/drug and drug/allergy interactions. MOSAIQ also performs standard calculations such as Body Surface Area (BSA) and Area Under the Curve (AUC). The Medical Administration Record (MAR) supports all information related to chemotherapy and blood product administration, clinical trial study drugs, dose amounts, infusion time, multiple administration, etc.

The Radiation Oncology configuration is also a charting solution with Computerized Physician Order Entry (CPOE) capability, along with added features for image management, patient setup and positioning, verify and record, plan import, review, and approval, stereotactic localization, and pretreatment checks. MOSAIQ's Radiation Oncology functionality can be used to support a wide variety of treatment modalities including Intensity Modulated Radio Therapy (IMRT), Image Guided Radio Therapy (IGRT), particle therapy, and stereotactic radiotherapy. It can import and store treatment plans from Therapy Planning Systems (TPS) via DICOM import/DICOM RT import.

In addition to these, the current version of MOSAIQ introduces the following modifications for radiation oncology:

• · Anatomic Position Monitoring (APM) with Manual interrupt (also referred to as True Tracking as it 'tracks' in 3D and enables the system to gate using the 3D position of the anatomy's motion)
• · APM with Anatomic Tolerance Check (ATC)
• · Adaptive Therapy with optional Baseline Shift (BLS) Recoverv
• · Care Rules for motion management.

These modifications are not contained solely within MOSAIQ, as the full clinical benefit is achieved with interoperability of Unity, Monaco RTP, and MOSAIQ OIS.

The provided text is a 510(k) summary for the MODAIQ® Oncology Information System. Based on this document, the following information can be extracted regarding acceptance criteria and supporting studies:

1. A table of acceptance criteria and the reported device performance

The document does not explicitly present a table of quantitative acceptance criteria and corresponding reported device performance metrics in the format typically seen for diagnostic or AI-driven systems (e.g., sensitivity, specificity, AUC).

Instead, the "acceptance criteria" are implied to be conformance with various standards and successful completion of design verification and performance testing. The "reported device performance" is essentially a statement that the device met these criteria through testing.

Therefore, a table of stated acceptance criteria and reported performance would look like this, based on the narrative:

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 states: "No animal or clinical tests were performed to establish substantial equivalence with the predicate device." This indicates that no test set involving patient data was used for performance evaluation in the context of clinical outcomes or diagnostic accuracy. The testing was non-clinical.

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 clinical or animal tests were performed, there was no need for experts to establish ground truth for a test set in the conventional sense of clinical performance studies. The "ground truth" for the non-clinical testing would have been established by engineering specifications, design requirements, and standard compliance.

4. Adjudication method (e.g. 2+1, 3+1, none) for the test set

As no clinical test set was used for performance evaluation that would require human expert adjudication of results, no adjudication method was employed.

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. The device is an Oncology Information System and not an AI-assisted diagnostic tool that would typically involve human readers interpreting cases with or without AI assistance.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done

The document describes "Design verification and performance testing" and "Software verification testing." These are standalone evaluations of the device's functionality and adherence to technical specifications and regulatory standards. However, it's not a standalone clinical performance study as would be seen for a diagnostic algorithm. The device's modifications for "Anatomic Position Monitoring" and "Adaptive Therapy" are explicitly mentioned to involve "interoperability of Unity, Monaco RTP, and MOSAIQ OIS," implying a system-level performance rather than a single algorithm's standalone clinical performance.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)

For the non-clinical testing performed, the ground truth was based on:

• Design and risk management requirements: The device's functions were tested against pre-defined specifications.
• Recognized consensus standards: Compliance with standards like ISO 14971, IEC 62304, ISO 62083, etc., served as the "ground truth" for its safety and essential performance.
• Predicate device characteristics: Substantial equivalence was established by demonstrating that the new device's characteristics and performance align with or don't adversely deviate from the predicate device (MOSAIQ OIS, K203172).

8. The sample size for the training set

The document does not mention any "training set." This type of testing is for an updated version of an existing Oncology Information System, focusing on functional verification and regulatory compliance, not machine learning model training.

9. How the ground truth for the training set was established

As no training set is mentioned or implied, this question is not applicable.

Ask a specific question about this device

The Monaco system is used to make treatment plans for patients with prescriptions for external beam radiation therapy. The system calculates dose for photon, electron, and proton treatment plans and in hard-copy, two- or three-dimensional radiation dose distributions inside patients for given treatment plan set-ups.

The Monaco product line is intended for use in radiation treatment planning. It uses generally accepted methods for:

• contouring
• image manipulation
• simulation
• image fusion
• plan optimization
• QA and plan review

The Monaco RTP System accepts patient diagnostic imaging data from CT and MR scans, and source dosimetry data, typically from a linear accelerator. The system then permits the user to display and define (contour) the target volume to be treated and critical structures which must not receive above a certain level of radiation, on these diagnostic images. Based on the prescribed dose, the user, a Dosimetrist or Medical Physicist, can then create multiple treatment scenarios involving the number, position(s) and energy of radiation beams and the use of a beam modifier (MLC, block, etc.) between the source of radiation and the patient to shape the beam. Monaco RTP system then produces a display of radiation dose distribution within the patient, indicating not only doses to the target volume but to surrounding tissue and structures. The optimal plan satisfying the prescription is then selected, one that maximizes dose to the target volume while minimizing dose to surrounding healthy volumes.

The parameters of the plan are output for later reference and for inclusion in the patient file. Monaco planning methods and modalities:

• Intensity Modulated Radiation Treatment (IMRT) planning
• Electron, photon and proton treatment planning
• Planning for dynamic delivery methods (e.g. dMLC, dynamic conformal, Volumetric Modulated Arc Therapy (VMAT))
• Stereotactic planning and support of cone-based stereotactic
• 3D conformal planning
• Adaptive planning (e.g. for the Elekta Unity MR-Linac)
• Monaco basic systems tools, characteristics, and functions:
• Plan review tools
• Manual and automated contouring tools
• DICOM connectivity
• Windows operating system
• Simulation
• Support for a variety of beam modifiers (e.g. MLCs, blocks, etc.)
• Standardized uptake value (SUV)
• Specialty Image Creation (MIP, MinIP, and Avg)
• Monaco dose and Monitor Unit (MU) calculation:
• Dose calculation algorithms for electron, photon, proton planning

Monaco is programmed using C and C++ computer programming languages. Monaco runs on Windows operating system and off-the-shelf computer server/hardware.

The provided text is a 510(k) summary for the Monaco RTP System, an updated version of a previously cleared device. It largely focuses on demonstrating substantial equivalence to the predicate device and does not contain detailed acceptance criteria or a study proving the device meets them in the format requested. The document explicitly states: "No animal or clinical tests were performed to establish substantial equivalence with the predicate device." Therefore, I cannot provide a table of acceptance criteria, reported device performance, or details about a clinical study.

However, based on the non-clinical performance testing described, here's what information can be extracted:

1. A table of acceptance criteria and the reported device performance:

Based on the provided text, specific numerical acceptance criteria and corresponding reported device performance values are not available. The document states that "Design verification and performance testing were carried out in accordance with design controls... against design and risk management requirements at sub-system, integration and system levels." It also mentions "Software verification testing was conducted and documented in accordance with FDA quidance 1 for devices that pose a major level of concern (Class C per IEC 62304)." However, the actual criteria for these verifications (e.g., specific error margins for dose calculation, response times for image manipulation) and the measured performance against those criteria are not detailed.

2. Sample size used for the test set and the data provenance:

• Sample size used for the test set: Not specified. The document mentions "sub-system, integration and system levels" testing, but no specific number of test cases or data sets are provided.
• Data provenance: Not specified. As no clinical studies were performed, the "data" would refer to test cases, models, or simulated data used in the non-clinical verification. The origin of this data is not mentioned.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:

• Not applicable/Not specified. Since no clinical studies or expert consensus activities are described for establishing ground truth on a test set, this information is not available.

4. Adjudication method (e.g. 2+1, 3+1, none) for the test set:

• Not applicable/None specified. No adjudication method is mentioned as there were no clinical studies.

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. The device is a Radiation Treatment Planning (RTP) system, not an AI-assisted diagnostic tool that would typically involve human readers in this context. The document explicitly states "No animal or clinical tests were performed."

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:

• The document describes "Design verification and performance testing" and focuses on the system's ability to calculate dose, perform image manipulation, optimization, etc., which implies a standalone performance evaluation of the software components. However, specific performance metrics for the algorithm only without any human interaction involved in setting up the plan or interpreting the output are not quantified. The mention of "Software verification testing" suggests an evaluation of the algorithm's correctness.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):

• For the non-clinical testing, the "ground truth" would likely be established through:
  • Reference data/models: For dose calculations, comparison against established physics models, phantom measurements, or other validated dose calculation systems.
  • Known input/output pairs: For software functionalities like image manipulation or contouring, where the expected output for a given input is pre-defined.
  • Compliance with standards: The document lists several ISO and IEC standards (e.g., ISO 62083 for radiotherapy treatment planning systems), implying that meeting the requirements of these standards serves as a form of "ground truth" for safety and performance.

8. The sample size for the training set:

• Not applicable/Not specified. The document does not describe the device as employing machine learning or AI that would require a "training set" in the conventional sense for a diagnostic algorithm. It describes a physics-based dose calculation and treatment planning system.

9. How the ground truth for the training set was established:

• Not applicable. As no training set is described, there's no information on how its ground truth would be established.

In summary, the provided document is a 510(k) premarket notification for an updated Radiation Treatment Planning (RTP) system. It focuses on demonstrating substantial equivalence to a predicate device through non-clinical verification and validation testing against design requirements and recognized standards, rather than providing details of a clinical study or specific quantitative acceptance criteria and performance data.

Ask a specific question about this device

Leksell Gamma Knife® is a teletherapy device intended for stereotactic irradiation of head structures ranging from very small target sizes of a few millimeters to several centimeters e.g.

• · Metastatic tumors
• Recurrent glioblastomas
• · Trigeminal neuralgia
• · Medically refractory essential tremor
• · Orbital tumors
• · Ocular tumors
• · Optic nerve tumors
  · Benign diseases (such as meningiomas, vestibular schwannomas, post-surgical pituitary adenomas, craniopharyngioma, hemangioblastomas, schwannomas, arteriovenous malformations, chordomas, chordomas, glomus tumors, hemangiomas)
• · Skull base tumors
  · Head and neck tumors (such as unknown primary of the head and neck, oral cavity, hypopharynx,
• nasopharynx, sinonasal, salivary gland)
  · Pediatric tumors (such as glioma, ependymoma, pituitary tumors, hemangioblastoma, meningioma, metastasis, medulloblastoma, nasopharyngeal tumors, arteriovenous malformations, skull base tumors)

Leksell Gamma Knife® (available models Icon™, Perfexion™ and Elekta Esprit) is a radiosurgery system for use in the stereotactic irradiation of head structures. Surgery is achieved by delivering a prescribed dose as one or more shots of ionizing radiation to the exact site of the target.

The provided text does not contain detailed information about specific acceptance criteria and a study that proves the device meets those criteria as it relates to a specific performance metric (like diagnostic accuracy). Instead, it focuses on the substantial equivalence of a new device model (Leksell Gamma Knife® - Elekta Esprit) to a predicate device (Leksell Gamma Knife® K173789) through non-clinical performance testing.

Here's an analysis based on the provided text, highlighting what is and isn't available:

Overall Conclusion from the Text: The document claims the new device is substantially equivalent to the predicate device because its fundamental technical characteristics are unchanged despite hardware and software updates to the user interface. Therefore, the performance is assumed to be equivalent, and no new clinical studies were conducted for this 510(k) submission.

Missing Information (per the request, but not available in the provided text):

• A table of acceptance criteria and reported device performance for a specific performance characteristic (e.g., diagnostic accuracy, sensitivity, specificity).
• Sample size used for a "test set" in the context of diagnostic performance.
• Data provenance (country of origin, retrospective/prospective).
• Number of experts, their qualifications, and adjudication methods for establishing ground truth.
• Multi-reader multi-case (MRMC) comparative effectiveness study results.
• Standalone (algorithm only) performance.
• Specific type of ground truth used (pathology, outcomes data, etc.) for performance evaluation.
• Sample size for a "training set" for an algorithm.
• How ground truth for a "training set" was established.

Information Available from the Text (interpreted in the context of "acceptance criteria" as general device functionality and safety):

The "acceptance criteria" here are implicitly related to ensuring the device performs as intended and is as safe and effective as its predicate, primarily through non-clinical testing of hardware and software updates.

1. A table of acceptance criteria and the reported device performance:

The document describes non-clinical performance testing to ensure the new model functions as intended and meets requirements.

2. Sample size used for the test set and the data provenance:

• Sample Size: Not specified in terms of patient cases or images, as the testing was non-clinical and focused on system functionality. It refers to "module, integration and system level verification testing."
• Data Provenance: Not applicable in the context of patient data. The testing was conducted internally by Elekta Solutions AB (Sweden).

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:

• Not applicable to this 510(k) submission, as no clinical performance study involving diagnostic accuracy with expert-established ground truth was performed for substantial equivalence. "Design and usability validation... performed by competent and professionally qualified personnel" (Page 5) speaks to the expertise in verifying engineering and human factors aspects, not medical image interpretation.

4. Adjudication method (e.g. 2+1, 3+1, none) for the test set:

• 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, an MRMC comparative effectiveness study was not done. The submission explicitly states: "No animal or clinical tests were performed to establish substantial equivalence with the predicate device." (Page 5). This device is a radiation therapy system, not an AI-assisted diagnostic tool for human readers.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:

• Not applicable in the context of an algorithm's diagnostic performance. The device itself is a treatment system, and its standalone performance refers to its ability to deliver radiation as specified, validated through non-clinical tests.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):

• For the non-clinical testing, the "ground truth" was the device's technical specifications and intended functionality. Verification and validation ensured the device met these engineering and design requirements. There was no medical "ground truth" (e.g., pathology, clinical outcomes) established for diagnostic or treatment efficacy in a clinical setting in this submission.

8. The sample size for the training set:

• Not applicable, as this device is not an AI/ML algorithm trained on a dataset in the sense of image recognition or diagnostic prediction.

9. How the ground truth for the training set was established:

• Not applicable.

Ask a specific question about this device

The Monaco system is used to make treatment plans for patients with prescriptions for external beam radiation therapy. The system calculates dose for photon, electron, and proton treatment plans and displays, on-screen and in hard-copy, two- or threedimensional radiation dose distributions inside patients for given treatment plan set-ups.

The Monaco product line is intended for use in radiation treatment planning. It uses generally accepted methods for:

• contouring
• image manipulation
• simulation
• image fusion
• . plan optimization
• QA and plan review

The Monaco RTP System accepts patient diagnostic imaging data from CT and MR scans, and source dosimetry data, typically from a linear accelerator. The system then permits the user to display and define (contour) the target volume to be treated and critical structures which must not receive above a certain level of radiation, on these diagnostic images. Based on the prescribed dose, the user, a Dosimetrist or Medical Physicist, can then create multiple treatment scenarios involving the number, position(s) and energy of radiation beams and the use of a beam modifier (MLC, block, etc.) between the source of radiation and the patient to shape the beam. Monaco RTP system then produces a display of radiation dose distribution within the patient, indicating not only doses to the target volume but to surrounding tissue and structures. The optimal plan satisfying the prescription is then selected, one that maximizes dose to the target volume while minimizing dose to surrounding healthy volumes.

The parameters of the plan are output for later reference and for inclusion in the patient file.

Monaco planning methods and modalities:

• Intensity Modulated Radiation Treatment (IMRT) planning .
• . Electron, photon and proton treatment planning
• . Planning for dynamic delivery methods (e.g. dMLC, dynamic conformal, Volumetric Modulated Arc Therapy (VMAT))
• . Stereotactic planning and support of cone-based stereotactic
• . 3D conformal planning
• . Adaptive planning (e.g. for the Elekta Unity MR-Linac)

Monaco basic systems tools, characteristics, and functions:

• . Plan review tools
• . Manual and automated contouring tools
• DICOM connectivity .
• . Windows operating system
• . Simulation
• . Support for a variety of beam modifiers (e.g. MLCs, blocks, etc.)
• . Standardized uptake value (SUV)
• Specialty Image Creation (MIP, MinIP, and Avq) •
• . Monaco dose and Monitor Unit (MU) calculation:
• Dose calculation algorithms for electron, photon, proton planning .

Monaco is programmed using C and C++ computer programming languages. Monaco runs on Windows operating system and off-the-shelf computer server/hardware.

This document, K213787, is an FDA 510(k) premarket notification for the Elekta Monaco RTP System, Release 6.1. It asserts substantial equivalence to a predicate device, Monaco RTP System (K202789). The document focuses on the non-clinical performance testing of the device, particularly for enhancements in proton therapy functionality.

Here's an analysis of the acceptance criteria and the study that proves the device meets them, based on the provided text:

1. A table of acceptance criteria and the reported device performance

The document does not provide a specific table with numerical acceptance criteria and corresponding reported device performance values in a format like "Target Value (X%) vs. Observed Value (Y%)". Instead, it describes a more qualitative approach to verifying the enhancements.

However, based on the "SUMMARY OF PERFORMACE TESTING (NON-CLINICAL)" section, we can infer the following:

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 does not explicitly state a "sample size" in terms of number of patient cases for the non-clinical testing. The tests described are computational comparisons (Monaco vs. Geant4, Monaco vs. manual summation) for various energies, materials, spot arrangements, and beam arrangements. These are not patient-specific data sets but rather simulated or theoretical scenarios designed to evaluate the computational accuracy of the new features.

There is no mention of "country of origin" or whether it was "retrospective or prospective" as these terms typically apply to studies involving patient data, which this non-clinical testing does not appear to use.

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)

For the non-clinical tests described:

• LET calculation accuracy: The ground truth for monoenergetic spots was established by comparison to Geant4 (a Monte Carlo simulation toolkit) and "other published results". This implies a reliance on established physics models and potentially peer-reviewed literature rather than human expert interpretation of images.
• LET distribution accuracy: Ground truth for complex arrangements was established by "expected values as obtained through manual summation of individual spots according to the design equations." This suggests a mathematical derivation as the ground truth.
• Clinical workflows: "Formal validation... by competent and professionally qualified personnel." No specific number or detailed qualifications (e.g., "radiologist with 10 years of experience") are provided for these personnel.

4. Adjudication method (e.g. 2+1, 3+1, none) for the test set

Given that the non-clinical tests involve comparisons to established algorithms (Geant4), published results, and mathematical derivations, there is no mention or indication of an adjudication method like "2+1" or "3+1", which are typically used for establishing consensus among human interpreters. The comparisons are to objective, established computational or theoretical benchmarks.

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, a multi-reader, multi-case (MRMC) comparative effectiveness study was not done. The document explicitly states: "No animal or clinical tests were performed to establish substantial equivalence with the predicate device." Therefore, there is no information on how much human readers improve with or without AI assistance, as the changes are to the dose calculation algorithms themselves (new proton functionalities like robust optimization, robust evaluation, and LET calculation) rather than an AI-assisted diagnostic or contouring tool that directly impacts human reader performance.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done

Yes, the described performance testing is primarily a standalone (algorithm only) evaluation. The comparisons are:

• Monaco vs. Geant4 (algorithm vs. algorithm/physics model)
• Monaco vs. manual summation based on design equations (algorithm vs. mathematical derivation)

The "clinical workflow validation" does involve "competent and professionally qualified personnel" interacting with the system, but the core performance evaluation of the new proton features (LET, robustness) is algorithmic.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)

The ground truth used for the non-clinical testing appears to be a combination of:

• Established physics models/simulations: Geant4 for monoenergetic LET calculations.
• Mathematical derivations/design equations: For complex LET distributions.
• "Published results": For qualitative comparison of monoenergetic LET.

There is no mention of expert consensus (for image interpretation), pathology, or outcomes data being used as ground truth for these specific non-clinical tests.

8. The sample size for the training set

The document does not mention a "training set" or "training data". This is because the Monaco RTP System (as described in this submission) uses deterministic algorithms for dose calculation (e.g., Monte Carlo algorithm) rather than machine learning or AI models that require a separate training phase. The "development" mentioned refers to software engineering and algorithm implementation, not machine learning model training.

9. How the ground truth for the training set was established

Since no training set is mentioned or implied for the deterministic algorithms described, this question is not applicable based on the provided document.

Ask a specific question about this device

AATMA™ is a medical image processing library intended to produce derived data sets for use as input into radiation therapy treatment planning systems or other intermediate pre-treatment-planning applications. AATMA™ does not provide a user interface and is designed to be accessed through its application programming interface (API) by other devices. The data sets created by AATMA™ must be reviewed and validated by a qualified healthcare professional prior to clinical use.

AATMA™ is an optional accessory to treatment planning systems and intermediate pre-treatment planning applications. The auto-segmentation algorithm in AATMA™ is based on machine-learning convolutional neural networks and includes pre-trained models that will be used to automatically segment image sets. The algorithm itself functions as a computational engine and does not store any input data, output data, or logs. The available models have been pre-trained on specific datasets that exhibit similar characteristics (e.g., body site and imaging modality).

As a medical image processing library, AATMA™ is designed to produce derived datasets in standard formats (e.g., DICOM) that can be utilized by other applications. AATMA™ does not have a user interface and, as such, calling applications must execute the auto-segmentation algorithms via AATMA™'s application programming interface (API).

AATMA™ must be used in conjunction with appropriate software to review and edit results generated automatically by the auto-segmentation algorithm. A pre-treatment planning system or treatment planning system must be used to facilitate the review and edit of contours generated by the auto-segmentation algorithm within AATMA™.

The provided text describes the 510(k) premarket notification for Elekta Solutions AB's Advanced Algorithms for Treatment Applications (AATMA™). This device is a medical image processing library designed to produce derived datasets for radiation therapy treatment planning systems or intermediate pre-treatment planning applications, primarily through auto-segmentation using machine learning convolutional neural networks.

Here's an analysis of the acceptance criteria and the study that proves the device meets them, based solely on the provided text:

Acceptance Criteria and Reported Device Performance

The provided text implicitly defines acceptance criteria through the successful attainment of a stated DICE coefficient for model performance.

Study Proving Device Meets Acceptance Criteria (Non-Clinical Performance Testing):

The document details non-clinical performance testing for two specific models: Head & Neck and Male Pelvis.

1. Sample sizes used for the test set and the data provenance:

   • Head & Neck Model:
     • Verification Set: 6 unique patient 3D CT image sets.
     • Validation Set: 13 unique 3D CT image sets.
     • Data Provenance: The training data (from which these test sets are distinct but of similar characteristics) came "from a variety of institutions and equipment." The document does not specify the country of origin or whether the data was retrospective or prospective, but the nature of the training implies existing, likely retrospective, clinical data.
   • Male Pelvis Model:
     • Verification Set: 5 unique patient CT image sets.
     • Validation Set: 20 unique 3D CT image sets.
     • Data Provenance: The training data (from which these test sets are distinct) came "from a global variety of institutions and equipment from patients undergoing RT." Again, the document does not specify the exact countries or whether it was retrospective/prospective, but implies existing clinical data.

2. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:

   • The text states that the verification sets for both models had "expert contours." However, it does not specify the number of experts or their qualifications (e.g., "radiologist with 10 years of experience").

3. Adjudication method (e.g., 2+1, 3+1, none) for the test set:

   • The document mentions "expert contours" were used for the verification sets. It does not specify an adjudication method used if multiple experts were involved (e.g., 2+1, 3+1). If only one expert reviewed each, then no adjudication would be necessary.

4. 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, an MRMC comparative effectiveness study was not done. The document explicitly states: "No animal or clinical tests were performed to establish substantial equivalence with the predicate device." The study focused on the algorithm's performance against expert contours, not on human reader improvement with AI assistance.

5. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:

   • Yes, a standalone (algorithm only) performance assessment was done. The described testing ("average DICE coefficient over all structures was determined") measures the algorithm's output (auto-segmented contours) directly against the established ground truth (expert contours), without human intervention in the loop during the performance measurement itself. The device is designed as an API-only computational engine.

6. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):

   • The ground truth for the test (verification) sets was established using "expert contours." It is not specified if this was a single expert per case or expert consensus.

7. The sample size for the training set:

   • Head & Neck Model: Trained on 66 unique clinical patient 3D CT image sets.
   • Male Pelvis Model: Trained on 205 unique patient 3D CT image sets.

8. How the ground truth for the training set was established:

   • The document states the models were "pre-trained on specific datasets." It does not explicitly describe how the ground truth within these training datasets was established. It is implied that these datasets contained "expert contours" (similar to the verification data), but this is not explicitly stated for the training data itself.

Ask a specific question about this device

Elekta Unity using Magnetic Resonance Imaging is indicated for radiation therapy treatments and stereotactic radiation treatments of malignant and benign diseases anywhere in the body as determined by a licensed medical practitioner in accordance with a defined treatment plan.

Elekta Unity is intended for use with compatible Treatment Planning and Oncology Information Systems. The Elekta Unity 1.5T MR scanner is a magnetic resonance imaging system that produces cross-sectional images in any orientation of the internal structure of the whole body before, during, and after the radiotherapy treatment.

When interpreted by a trained physician magnetic resonance images acquired before, during, and after the radiotherapy treatment yield information that can be useful in diagnosis and may assist therapy planning, patient positioning and treatment delivery related to radiation oncology.

Elekta Unity is a multifunctional digital linear accelerator designed to assist licensed medical practitioners in the delivery of ionizing radiation to defined volumes (e.g. malignant and benign tumors). Elekta Unity is capable of both intensity modulated radiation therapy (IMRT) and image guided radiation therapy (IGRT).

The Elekta Unity 1.5T MRI scanner is a magnetic resonance imaging sub-system that produces cross-sectional images in any orientation of the internal structure of the whole body before, during, and after the radiotherapy treatment.

When interpreted by a trained physician the images acquired before, during, and after the radiotherapy treatment yield information that can be useful in diagnosis and may assist therapy planning, patient positioning and treatment delivery related to radiation oncology.

In addition to the MRI sequences cleared with the predicate device, the subject device Elekta Unity configuration has the ability to generate images using following techniques, before, during or after treatment:

Introduce 3D Vane
3D Vane XD is a free breathing acquisition method that can be used to compensate for respiratory motion and peristalsis in 3D/FFE and 3D/TFE body imaging.

Introduce CS-Sense - compressed images
Compressed SENSE is an acceleration technique that is less sensitive to noise allowing increased resolution and/or coverage without a scan time penalty.

Introduce Breath Hold (BH) .
Breath-hold is a technique where an image is acquired when a patient holds their breath in a defined phase of the breathing cycle.

The Elekta Unity device mentioned in the document is a medical charged-particle radiation therapy system. However, the provided document does not contain information about specific acceptance criteria or a dedicated study proving device performance against such criteria for AI/ML components.

The document primarily focuses on demonstrating substantial equivalence of the overall device (Elekta Unity with new imaging options) to a predicate device (Elekta Unity K192482). It lists performance testing for the overall system based on design verification, risk management, software verification, and conformance to recognized consensus standards.

Here's a breakdown of why the requested information cannot be fully provided based on the given text, and what information is available:

Missing Information (Specific to AI/ML acceptance criteria and performance study):

• No explicit acceptance criteria for an AI/ML component: The document introduces new imaging options (3D Vane XD, CS-Sense, Breath Hold) but does not frame these as AI/ML applications requiring specific performance metrics like sensitivity, specificity, or AUC against a ground truth. They are described as "additional imaging options" or "acceleration techniques."
• No dedicated study proving AI/ML device performance: There is no study described that evaluates the performance of any AI/ML algorithm within the Elekta Unity against a defined ground truth, nor are there details about sample size, expert readers, or adjudication methods for such a study.
• No MRMC comparative effectiveness study for AI assistance: The document does not describe any study where human readers' performance with and without AI assistance was compared.
• No standalone algorithm-only performance study: No study detailing the performance of an algorithm without human involvement is mentioned.
• No details on ground truth for AI/ML training or testing: Since no specific AI/ML component is detailed with its own performance study, there's no information about the ground truth used for training or testing such components.
• No sample size for training sets of AI/ML components: This information is not present.

What is available regarding overall device testing and compliance:

The document states broader performance testing was conducted for the Elekta Unity system, but these are general engineering and safety tests rather than AI/ML specific performance evaluations.

• Design verification and performance testing: Carried out in accordance with FDA's Quality System Regulation (21 CFR §820.30), ISO 13485, ISO 14971, and IEC 62304.
• Software verification testing: Conducted and documented in accordance with FDA guidance for devices that pose a major level of concern (Class C per IEC 62304).
• Basic safety and essential performance: Satisfied through conformance with device-specific recognized consensus standards (listed in a table).

However, none of this directly answers the AI/ML-specific questions in your prompt. The "new imaging options" are presented as new functionalities of the MR system rather than intelligent algorithms for interpretation or decision support.

Therefore, a table of acceptance criteria and reported device performance for an AI/ML component cannot be created from the provided text. The document focuses on demonstrating that the device as a whole with its new MR imaging sequences maintains safety and effectiveness comparable to its predicate, largely through non-clinical engineering and software testing and adherence to general medical device standards.

Ask a specific question about this device

The Monaco system is used to make treatment plans for patients with prescriptions for external beam radiation therapy. The system calculates dose for photon, proton, and electron treatment plans and displays, on-screen and in hard-copy, two- or threedimensional radiation dose distributions inside patients for given treatment plan setups.

The Monaco product line is intended for use in radiation treatment planning. It uses generally accepted methods for:

• · contouring
• · image manipulation
• · simulation
• · image fusion
• · plan optimization
• QA and plan review

Monaco is a radiation treatment planning system that first received FDA clearance in 2007 (K071938). The modified system received clearance in 2009. when Volumetric Modulated Arc Therapy (VMAT) planning capability was added (K091179), again when Dynamic Conformal Arc planning was added (K110730), and electron planning, support for stereotactic cones, and SUV calculation were added (K132971). Specialty image creation was added in 2015 (K151233), and adaptive planning and dose calculation in the presence of a magnetic field (e.g., MR-Linac) was added in 2018 (K183037). A 510(k) was filed in 2017 for the addition of carbon ion planning. The 510(k) was withdrawn because there was no hardware cleared for the US market capable of delivering carbon ion plans. Monaco's carbon ion planning functionality remains licensed off and inaccessible to US users.

The Monaco system accepts patient imaging data and "source" dosimetry data from a linear accelerator. The system then permits the user to display and define (contour) the target volume to be treated and critical structures which must not receive above a certain level of radiation on these diagnostic images.

Based on the prescribed dose, the user, a Dosimetrist or Medical Physicist, can create multiple treatment scenarios involving the number, position(s) and energy of radiation beams and the use of a beam modifier (MLC, block, etc.) between the source of radiation and the patient to shape the beam. The Monaco system then produces a display of radiation dose distribution within the patient, indicating doses to the target volume and surrounding structures. The "best" plan satisfying the clinican prescription is then selected, one that maximizes dose to the target volume while minimizing dose to surrounding healthy volumes.

Monaco 6.00 supports Proton Pencil Beam Scanning (Proton PBS) planning for IBA Proteus®ONE and Proteus®PLUS delivery systems (Ion Beam Applications S.A.).

The provided text is a 510(k) summary for the Elekta Monaco RTP System. It describes the device, its intended use, and a comparison to predicate devices, but it explicitly states that "Clinical trials were not performed as part of the development of this product. Clinical testing on patients is not advantageous in demonstrating substantial equivalence or safety and effectiveness of the device since testing can be performed such that no human subjects are exposed to risk. Validation testing involved simulated clinical workflows using actual patient data, such as patient images. Pre-defined pass/fail criteria were also equivalent to that of the previous version of Monaco."

Therefore, I cannot provide a detailed answer to your request regarding acceptance criteria and a study proving the device meets them in the way you've outlined, as there was no clinical study (MRMC, standalone, etc.) involving human readers or a traditional test set/ground truth establishment as would be done for an AI-based diagnostic device.

The study referenced is a non-clinical verification study that focuses on software functionality, safety, and effectiveness compared to an existing predicate device, primarily through regression testing and verification of new functionalities. The acceptance criteria are "pre-defined pass/fail criteria" that were equivalent to those used for the previous version of Monaco.

However, based on the information provided, I can infer and summarize what was done:

1. A table of acceptance criteria and the reported device performance:

The document broadly states that "Pre-defined pass/fail criteria were also equivalent to that of the previous version of Monaco." and "Conformity to the same pass/fail criteria as the predicate version of Monaco indicated that Monaco 6.00 was substantially equivalent in safety and effectiveness."

While specific numeric acceptance criteria and performance metrics are not detailed in this summary, the overall acceptance criterion was Substantial Equivalence to the predicate devices. The performance reported is that "Monaco 6.00 was deemed safe and effective for its intended use" based on the non-clinical testing.

2. Sample size used for the test set and the data provenance:

• Sample Size for Test Set: The document mentions "Over 600 test procedures were executed" and "Validation testing involved simulated clinical workflows using actual patient data, such as patient images." The exact number of patient datasets or specific test cases within those 600 procedures is not specified.
• Data Provenance: "actual patient data, such as patient images." No specific country of origin is mentioned, but "simulated clinical workflows" suggests internally generated or existing de-identified data. The testing was retrospective in the sense that it used pre-existing patient data for simulation, not prospective patient enrollment.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:

• This wasn't a ground truth establishment study for a diagnostic AI. The "ground truth" for this system's validation was primarily the expected output of the algorithms (e.g., dose calculations, plan optimizations) as per documented specifications and comparisons to known good results from the predicate device.
• The document states that "Once completed, plans are reviewed and approved by qualified clinicians and may be subject to quality assurance practices before treatment actually takes place." This implies that the system is used by "Dosimetrist or Medical Physicist" and reviewed by "qualified clinicians" in a clinical setting, but these roles were not part of a formal "ground truth" establishment for the validation study itself.

4. Adjudication method (e.g., 2+1, 3+1, none) for the test set:

• Not applicable. This was a software verification and validation against specified requirements and predicate performance, not a clinical adjudication of diagnostic findings.

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 performed. The document explicitly states: "Clinical trials were not performed as part of the development of this product." and "Clinical testing on patients is not advantageous in demonstrating substantial equivalence or safety and effectiveness of the device since testing can be performed such that no human subjects are exposed to risk."

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:

• The testing described is essentially a form of "standalone" algorithm verification in a simulated environment, focused on the software's functional correctness for tasks like dose calculation and plan optimization, rather than a diagnostic AI's performance. The product itself is a "Radiation Treatment Planning System," which is inherently a human-in-the-loop device, where the software outputs are reviewed and approved by clinicians before implementation. The verification ensured the software's outputs were correct according to its specifications and predicate performance.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):

• The "ground truth" for this validation was primarily derived from:
  • Validated algorithms/physical models: For dose calculation, the "ground truth" is based on established physics principles and validated dose calculation algorithms (Monte Carlo, Collapsed Cone, Pencil Beam).
  • Predicate device performance: Functional equivalence and similar calculation results to the previously cleared Monaco RTP System (K190178) and RayStation 8.1 (K190387).
  • Pre-defined pass/fail criteria: These would be based on engineering specifications, clinical requirements for accuracy in dose distribution, and comparison to known good results for test cases.
  • "Simulated clinical workflows using actual patient data": This implies that for these simulated workflows, the expected correct outcome (e.g., the accurate dose distribution for a given patient anatomy and treatment plan) served as the reference.

8. The sample size for the training set:

• This device is not an AI/ML model that undergoes a "training" phase in the typical sense (i.e., learning from annotated data). It's a software system built on established algorithms for radiation treatment planning. Therefore, there is no "training set" as would be applicable to a deep learning model.

9. How the ground truth for the training set was established:

• Refer to point 8. Not applicable for this type of device.

Ask a specific question about this device