(30 days)
Mobius3D software is used for quality assurance, treatment plan verification, and patient alignment and anatomy analysis in radiation therapy. It calculates radiation dose three-dimensionally in a representation of a patient or a phantom. The calculation is based on read-in treatment plans that are initially calculated by a treatment planning system and may additionally be based on external measurements of radiation fields from other sources such as linac delivery log data. Patient alignment and anatomy analysis is based on read-in treatment planning images (such as computed tomography) and read-in daily treatment images (such as registered cone beam computed tomography).
Mobius3D is not a treatment planning system. It is only to be used by trained radiation oncology personnel as a quality assurance tool.
Mobius3D (v. 2.2) is a software product used within a radiation therapy clinic for quality assurance and treatment plan verification. It is important to note that while Mobius3D operates in the field of radiation therapy, it is neither a radiation delivery device (e.g. a linear accelerator), nor is it a treatment planning system (TPS). Mobius3D cannot design or transmit instructions to a delivery device, nor does it control any other medical device. Mobius3D is an analysis tool meant solely for quality assurance (QA) purposes when used by trained medical professionals. Being a software-only QA tool, Mobius3D never comes into contact with patients.
Mobius3D performs dose calculation verifications for radiation treatment plans by doing an independent calculation of radiation dose. Radiation dose is initially calculated by a treatment planning system (TPS), which is a software tool that develops a detailed set of instructions (i.e. a plan) for another system (e.g. a linear accelerator) to deliver radiation to a patient. The dose calculation performed by Mobius3D uses a proprietary collapsed cone convolution superposition (CCCS) algorithm.
Mobius3D also performs dose delivery quality assurance for radiation treatment plans by using the measured data recorded in a linear accelerator's delivery log files to calculate a delivered dose. This is presented to the end user in a software component of Mobius3D called MobiusFX. The MobiusFX component is available to users through licensing as an add-on to the core Mobius3D software features.
Mobius3D performs quality assurance of a patient's alignment and anatomy analysis. This analysis is based on comparison of Cone Beam Computed Tomography (CBCT) images taken immediately before treatment to the images used for treatment planning, which are typically acquired using standard Computed Tomography (CT). This analysis is presented to the end user in an add-on software module within Mobius3D called CBCT Checks.
This document is a 510(k) premarket notification for the Mobius3D v2.2 software, which is used for quality assurance in radiation therapy. The submission focuses on demonstrating substantial equivalence to a predicate device (Mobius3D v2.0.0).
Based on the provided text, here's an analysis of the acceptance criteria and study information:
1. A table of acceptance criteria and the reported device performance:
The document briefly mentions "software verification and validation" and states, "The non-clinical data support the safety of the device and the software verification and validation demonstrate that subject device should perform as intended in the specified use conditions." However, a specific table outlining acceptance criteria and reported device performance metrics is NOT provided in the given text. This information would typically be in a separate section of the 510(k) submission, not in the provided summary.
The document implicitly refers to performance related to "dose calculation verifications" and "dose delivery quality assurance" and "patient alignment and anatomy analysis," but no quantitative performance metrics or acceptance thresholds are given.
2. Sample size used for the test set and the data provenance:
The document states, "No animal studies or clinical tests have been included with this pre-market submission." This indicates that the safety and effectiveness determination is based on non-clinical data, likely software verification and validation activities. Since no clinical tests were performed, there is no "test set" in the traditional sense of patient data. Therefore, questions about sample size and data provenance (country of origin, retrospective/prospective) are not applicable to this specific submission as described.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
Since no clinical or human-in-the-loop tests were performed with a test set requiring ground truth, this information is not applicable and not provided in the document. The "ground truth" for a software QA tool like Mobius3D would likely be based on established physics principles, dose calculation standards, and comparisons to known accurate systems or phantoms.
4. Adjudication method for the test set:
Not applicable, as no external test set requiring human adjudication was used for this submission.
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 study was done. The document explicitly states, "No animal studies or clinical tests have been included with this pre-market submission." Mobius3D is described as an analysis tool for quality assurance, not a diagnostic AI system intended to assist human readers in image interpretation. Therefore, a MRMC study investigating human reader improvement with AI assistance is not relevant to the scope of this device as described.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
While the document doesn't explicitly use the term "standalone performance study," the entire submission suggests an algorithm-only (or software-only) performance evaluation as the basis for substantial equivalence. The "software verification and validation" (V&V) would have evaluated the Mobius3D algorithm's accuracy in dose calculations and analysis functions against expected outputs, likely using simulated data, phantom data, or comparisons to established (predicate) systems.
The changes mentioned, such as "Dose Calculation Reimplemented in C++," would have necessitated thorough verification of the reimplemented algorithm's accuracy as a standalone component.
7. The type of ground truth used:
Given the nature of the device (a quality assurance tool for radiation dose calculation and patient alignment analysis) and the absence of clinical studies, the "ground truth" for the software's performance would likely be established through:
- Physics-based models and established dosimetry principles: For dose calculation verification, the ground truth would be the theoretically correct dose distribution for specific radiation fields and patient geometries.
- Phantom measurements and comparisons: Physical phantoms with known properties would be irradiated, and the measured dose (using ion chambers or film) would serve as a ground truth for comparison with the software's calculations.
- Comparison to predicate device performance: The submission directly compares Mobius3D v2.2 to Mobius3D v2.0.0, implying that the performance of the predicate serves as a benchmark for substantial equivalence.
- Known image registration accuracy standards: For patient alignment and anatomy analysis, the ground truth would relate to the accuracy of image registration algorithms against known transformations or expertly aligned images.
The document does not explicitly state the type of ground truth used, but it can be inferred from the device's function.
8. The sample size for the training set:
This is not applicable and not provided. Mobius3D, as described, performs dose calculations and analysis using a proprietary collapsed cone convolution superposition (CCCS) algorithm and image processing techniques. It is not described as a machine learning/AI model that requires a "training set" in the context of supervised learning from a dataset of clinical cases (e.g., for image classification or segmentation). The software's capabilities are based on deterministic algorithms and physics models, not trained on a large dataset of patient outcomes or expert annotations.
9. How the ground truth for the training set was established:
Not applicable, as there is no "training set" in the context of supervised machine learning for this device.
§ 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.