(140 days)
The Astroid Planning App is an interactive end user application for proton treatment planning for the intended use and primary purpose of enabling radiotherapy professionals to efficiently design and analyze proton radiotherapy treatment plans. This Astroid Planning App leverages the existing .decimal Astroid Dosimetry App [FDA 510(k) K150547), which is a library of treatment planning functions accessed through the Thinknode® cloud services framework, for device creation, dose calculation, optimization, and all other dosimetry and processing calculations. Since the Astroid Dosimetry App is responsible for performing the calculations, the scope of this Astroid Planning App is to be the user interface for end users to input treatment planning data and review the results. Typical indications for the treatment of persons with cancer, over a wide range of potential disease locations. In the most common use case of the software, users will import patient data from existing imaging software programs, manage physician prescription and intent information, develop a proton treatment plan, and analyze the plan to determine how well it meets the physician's goals. Since the critical treatment planning functions are handled outside this software application, by a software of known quality and pedigree, the primary and most frequently used functions of this software are the record keeping service (for patient data storage), user interface controls, and visualization tools.
Furthermore, since the accuracy of information computed and displayed by an application such as this is very important to the proper treatment of patients, it is critical that users have the appropriate educational and clinical experience backgrounds to adequately understand and use the product. Additionally, since each radiotherapy treatment machine produces a unique beam of radiation, there is much responsibility on the end users to adequately commission and test this software over the full range of expected treatment conditions before the system is utilized for patient treatment.
The primary purpose of the Astroid Planning App device is for facilitating the planning and analysis of proton radiation therapy treatments. The Astroid Planning App device is an interactive end user application in which the user interacts with the interface to perform proton treatment planning tasks. The data constructed in the Astroid Planning App device will be used as inputs to the Astroid Dosimetry App device [FDA 510(k) K150547], which is the foundational proton dosimetry calculation library that contains all algorithms and calculation processing for the proton treatment planning. The Astroid Planning App device composes and otherwise constructs the calculation requests required for the development of the proton treatment plan, leveraging the functions externally located in the Astroid Dosimetry App to then perform the requested calculations.
The Astroid Planning App is an interactive end-user application for proton treatment planning. It leverages the existing .decimal Astroid Dosimetry App (K150547) for device creation, dose calculation, and optimization. The Planning App's scope is to provide a user interface for inputting treatment planning data and reviewing results, while the critical treatment planning functions are handled by the Dosimetry App.
1. Table of acceptance criteria and the reported device performance:
The provided text does not contain a specific table of quantitative acceptance criteria. However, it states that the device was evaluated against predicate devices and through validation tests. The overall reported performance is that the Astroid Planning App "performed as well as the predicate devices and that the Astroid Planning App is deemed safe and effective for clinical use when properly commissioned for a proton treatment machine."
2. Sample size used for the test set and the data provenance:
The document states: "Validation tests comparing results of proton dose calculations, with the inclusion of all applicable treatment delivery devices, to experimental and analytical datasets were performed." It does not specify the exact sample size or the provenance of the experimental and analytical datasets used for comparison. It does not indicate if the data was retrospective or prospective or its country of origin.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
This information is not explicitly provided in the document.
4. Adjudication method for the test set:
The document does not describe a formal adjudication method (e.g., 2+1, 3+1). It states that "Clinically oriented validation test cases were written and executed by .decimal personnel and hospital-based testing partners." This suggests internal validation and potentially collaboration with clinical partners, but not a formal adjudication process for establishing ground truth as might be expected in a clinical study.
5. If a multi-reader multi-case (MRMC) comparative effectiveness study was done, and if so, what was the effect size of how much human readers improve with AI vs without AI assistance:
A multi-reader multi-case study comparing human readers with and without AI assistance was not explicitly mentioned. The testing focused on the performance of the Astroid Planning App itself, comparing it to predicate devices. The device serves as a planning tool, not a diagnostic aid that would typically involve human reader performance changes.
6. If a standalone (i.e., algorithm only without human-in-the-loop performance) was done:
The document states that "the primary and most frequently used functions of this software are the record keeping service (for patient data storage), user interface controls, and visualization tools." The critical treatment planning functions and calculations are handled by the separate Astroid Dosimetry App. Therefore, the Astroid Planning App, as described, is inherently a "human-in-the-loop" interface for the Astroid Dosimetry App, and a standalone algorithm-only performance assessment of the Planning App itself (without considering the Dosimetry App) would not be applicable to its stated function. However, the underlying Astroid Dosimetry App (K150547) would have undergone its own standalone performance evaluation. This document focuses on the Planning App's role as an interface.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
The ground truth for the dose calculation comparisons was based on "experimental and analytical datasets." This implies physical measurements and/or established theoretical models for proton dose calculations. For other functions, the "performance as well as the predicate devices" suggests the predicate devices' established accuracy served as a benchmark for comparison.
8. The sample size for the training set:
Information regarding a specific training set size is not provided. As this is a planning application, rather than a machine learning model that would typically have a distinct training phase, the concept of a "training set" might not directly apply in the same way. The development and verification process likely involved extensive iterative testing and refinement.
9. How the ground truth for the training set was established:
Since a distinct training set (in the machine learning sense) is not described, the method for establishing its ground truth is not mentioned. The validation process, however, involved comparison to "experimental and analytical datasets," which would serve as the reference for accuracy.
Summary of the Study:
The study was a non-clinical testing effort focused on demonstrating that the Astroid Planning App performs comparably to predicate devices and is safe and effective when properly commissioned.
- Tests performed:
- Validation tests comparing proton dose calculations to experimental and analytical datasets.
- Verification and validation tests for all other externally available functions.
- Usability testing, including analysis of all system displays and user options.
- Plan quality studies and full end-to-end testing of the entire planning process as compared to the predicate devices.
- Personnel involved: .decimal personnel and hospital-based testing partners.
- Comparison: The device was compared to predicate devices (Eclipse Treatment Planning System K172163 and RayStation 4.0.2 K140187) in terms of technology, intended uses, and end-user profiles.
- Conclusion: The tests showed that the Astroid Planning App performed as well as the predicate devices.
§ 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.