(29 days)
ExacTrac Dynamic is intended to position patients at an accurately defined point within the treatment beam of a medical accelerator for stereotactic radiosurgery or radiotherapy procedures, to monitor the patient position and to provide a beam hold signal in case of a deviations in order to treat lesions, tumors and conditions anywhere in the body when radiation treatment is indicated.
ExacTrac Dynamic is a patient positioning device used in a radiotherapy environment as an addon system to standard linear accelerators. It uses patient planning and CT data to determine the patient's planned position and compares it via oblique x-ray images to the actual patient position. The calculated correction shift will then be transferred to the treatment machine to align the patient correctly at the machine's treatment position. During treatment is monitored with a surface camera and X-ray to ensure no misalignment due to patient movement.
ExacTrac as a medical device consists of Hardware and Software. Together they are ExacTrac Dynamic.
ExacTrac Dynamic 1.0.3 UDI DI: 04056481142506 ExacTrac Dynamic has no variants.
ExacTrac Dynamic shall be installed in a treatment room within a hospital radiation therapy department.
ExacTrac uses X-ray images acquired with two X-ray tubes and two amorphous silicon detectors to compare the current patient position with the previously planned patient position, based on CT volumetric scans.
The current patient position is monitored using a surface monitoring system.
If necessary, the patient position is corrected using 3rd party patient treatment tables with or without pitch and roll correction possibilities.
Stereotactic radiosurgery (SRS) and Stereotactic Body Radiation Therapy (SBRT) are a highly accurate form of the radiation therapy initially developed to treat small (benign or malignant) tumors and functional abnormalities of the brain.
During treatment a high dose of radiation is delivered within millimeter accuracy to the planned target volume (PTV) while minimizing the dose to the surrounding healthy tissue and organs at risk (OARs). For an accurate dose delivery a precise patient positioning is necessary and an eventual re-positioning of the patient in case of a patient movement during the treatment. The patient positioning system is based on two non-coplanar x-ray 2D image acquisition, that is accurately aligned to the volumetric planning CT (from the plan), the transformation matrix (of the 2D or 3D alignment) is sent to the treatment couch moving the patient to the planned iso-center of the treatment delivery system (Linac). During the treatment delivery the patient movement is monitored via a (3D) surface and a thermal information.
Additionally, an x-ray based position acquisition may be acquired and compared with the planned patient position. Quality assurance procedure are included to ensure a calibration of ExacTrac to itself and to the LINAC
The provided document does not contain the specifics of a study evaluating the device against acceptance criteria in the way typically required for machine learning or AI-driven medical devices. The K211939 submission primarily deals with a bug fix for an existing device (ExacTrac Dynamic 1.0) and focuses on regulatory compliance, electrical safety, EMC, and software verification/validation, rather than a clinical performance study with a test set, ground truth, and expert adjudication.
The "Performance Data" section lists "Essential Performance in detail" which could be interpreted as acceptance criteria, but no study is described to prove these criteria were met with specific quantitative results.
However, I can extract the "acceptance criteria" (from the "Essential Performance in detail" section) and describe what a study would need to look like based on the information provided, even if the detailed study results are not present in this document.
Here's a breakdown assuming the "Essential Performance in detail" are the acceptance criteria, and then explaining what data is missing:
Acceptance Criteria and Reported Device Performance
Based on the "Essential Performance in detail" section, the following can be inferred as acceptance criteria:
Table 1: Inferred Acceptance Criteria and Stated Performance
| Acceptance Criteria (Essential Performance) | Reported Device Performance (as stated in document) |
|---|---|
| i) Positioning: Providing a shift for the patient support system compensating the deviation between patient position and planned position, where the shift is calculated with an accuracy of 1mm based on stereo x-ray images. | The document states “calculated with an accuracy of 1mm based on stereo x-ray images” as part of the criteria itself, implying this is the target/achieved performance. No separate study data is presented to prove this 1mm accuracy for this specific criterion during a performance test. |
| ii) Positioning and Monitoring: Display of an X-ray / DRR overlay and indicating a corresponding deviation between patient position and planned position, where the deviation is calculated with an accuracy of 1mm based on stereo x-ray images. | Similar to (i), the document states “calculated with an accuracy of 1mm based on stereo x-ray images” as part of the criteria. No separate study data is presented to prove this 1mm accuracy for this specific criterion during a performance test. |
| iii) Monitoring: Indicating a deviation exceeding a predefined tolerance between patient position and a reference position based on surface tracking with an accuracy of 1mm (during irradiation). | Similar to (i) and (ii), the document states “with an accuracy of 1mm (during irradiation)” as part of the criteria. No separate study data is presented to prove this 1mm accuracy for this specific criterion during a performance test. |
| iv) Monitoring: Indicating a deviation exceeding a predefined tolerance between patient position and planned position based on stereo X-rays with an accuracy of 1mm (during irradiation). | Similar to (i), (ii), and (iii), the document states “with an accuracy of 1mm (during irradiation)” as part of the criteria. No separate study data is presented to prove this 1mm accuracy for this specific criterion during a performance test. |
| v) Providing a "Beam hold" signal to the LINAC within 3 sec, if there is a deviation between patient position and reference position exceeding tolerance limits as adjusted by the user. | The document states "within 3 sec" as part of the criteria, implying this is the target/achieved performance. No separate study data is presented to prove this 3-second response time for this specific criterion during a performance test. This functionality is also stated as "(optional)" in the Essential Performance summary. |
Important Note: The document presents the accuracy and timing requirements directly within the description of the "Essential Performance," rather than providing a separate table of acceptance criteria followed by a table of results from a performance study demonstrating that these criteria have been met. This suggests that the listed values are the expected or designed performance which the bug fix (K211939) did not alter, and that the original predicate device (K201276) would have been cleared based on demonstrating these values previously. The current submission focuses on demonstrating that the bug fix did not negatively impact these performances.
Details of the Study (Based on typical requirements, but largely missing from the provided document Specifics):
The provided document does not describe a detailed performance study with a test set, ground truth, or expert adjudication that quantitatively assesses the 1mm accuracy or 3-second beam hold for the ExacTrac Dynamic (K211939) as if it were a new device needing to prove these metrics. Instead, it refers to "Various verification and validation tests" and states that "The Subject device has passed all the necessary tests and thereby is considered safe and effective for its intended use." The submission is for a bug fix to an already cleared device, implying that the detailed performance validation would have been done for the predecessor (K201276).
Therefore, the following answers are based on what would typically be required for such a device to gain clearance or what could be inferred/assumed given the nature of the device, but the specific details are not explicitly stated in the provided text.
2. Sample Size and Data Provenance for the Test Set:
- Sample Size: Not specified in the provided document. For a precision device like this, the "test set" would likely involve phantoms or controlled patient movements, rather than a large clinical image dataset like for an AI diagnostic device. The "accuracy of 1mm" suggests a direct measurement against a known standard.
- Data Provenance: Not specified. Given the nature of a medical device manufacturer, testing would most likely be conducted in a controlled lab or clinical environment, possibly internally or by a contracted lab. The manufacturer (Brainlab AG) is based in Germany, so the primary testing might have occurred there, or potentially at US sites for clinical validation. The document implies these are technical performance specifications rather than clinical data.
3. Number of Experts and Qualifications for Ground Truth:
- Number of Experts: Not applicable/specified. For a device measuring physical displacement, ground truth would be established by precisely calibrated measurement tools (e.g., optical tracking systems, high-precision CMMs, or verified phantom positions) rather than human experts interpreting images.
- Qualifications of Experts: N/A, as ground truth would be physical measurements.
4. Adjudication Method for the Test Set:
- Method: Not applicable. Adjudication methods (like 2+1, 3+1) are used to resolve disagreements among human experts when establishing ground truth for subjective interpretations (e.g., classification of medical images). For a device performing precise physical positioning/monitoring, the "ground truth" is typically a direct physical measurement with a known, higher accuracy.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study:
- MRMC Study Done? No. This type of study assesses how AI assistance impacts human reader performance (e.g., radiologists interpreting images). ExacTrac Dynamic is a physical positioning and monitoring system, not an AI image interpretation tool. Its "AI" component would be in its tracking algorithms, not in assisting human diagnostic reading.
- Effect Size of Improvement with AI: Not applicable for this device type.
6. Standalone (Algorithm-Only) Performance:
- Standalone Performance Done? Yes, effectively. The device's stated "accuracy of 1mm" for positioning and monitoring (criteria i-iv) and "3-second beam hold" (criterion v) are standalone performance metrics of the algorithm and hardware working together. The device's function is to autonomously (or with user-configured tolerances) perform these actions. There isn't a human-in-the-loop directly influencing the "1mm accuracy" calculation or the "3-sec beam hold" trigger.
7. Type of Ground Truth Used:
- Type of Ground Truth: Likely physical measurements from highly accurate, independent reference systems or phantoms with precisely known geometries. For example:
- For positional accuracy (1mm), a calibrated phantom moved by a precision stage, with its true position simultaneously measured by an external optical tracking system with sub-millimeter accuracy, would serve as ground truth.
- For beam hold time, a timestamped signal from the LINAC triggered by a known deviation, compared to a timestamp from the ExacTrac system, would be used.
8. Sample Size for the Training Set:
- Sample Size: Not applicable/specified. This is not described as a deep learning/AI device trained on a large image dataset in the context of generating diagnostic insights. It's a precise medical device based on pre-programmed algorithms for image registration, surface tracking, and real-time computation of shifts and deviations. While internal algorithms might have been "trained" or optimized during development, it would not be in the typical sense of a "training set" for a sophisticated AI model like an image classifier. The "training" would be more about calibration and refinement of image processing and control loop parameters.
9. How Ground Truth for the Training Set Was Established:
- How Ground Truth Was Established: Not applicable/specified. As per point 8, the concept of a "training set" and associated ground truth is not detailed for this type of device in the provided document. The device's underlying principles are physics-based (X-ray geometry, surface topography) rather than statistical learning from a vast dataset of medical images.
§ 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.