Search Results

The intended use of the device is:

· quality assurance of patient specific treatment delivery prior to the treatment in IMRT (including VMAT) and 4DRT (e.g. respiratory gating and tumour tracking).

· quality assurance of the radiation delivery system.

The device consists out of matrices of semiconductors embedded in a phantom. These matrices are inserted into the radiation field of a medical linear accelerator. If radiation (from a radiotherapy treatment field) hits the semiconductors a signal is created and transferred to a computer where it is analysed and among others compared with the intended dose distribution.

This document is a 510(k) premarket notification for the Delta4 Phantom+ device, which is an improved version of the predicate device, Delta4. The submission focuses on demonstrating substantial equivalence to the predicate, rather than an independent clinical study to establish acceptance criteria and performance against those criteria as would be typical for a device with a new clinical function or significant safety/performance changes.

Therefore, the information you're requesting regarding acceptance criteria and performance studies in the context of an AI device is not fully available or directly applicable from this specific document. This document describes a medical device used for Quality Assurance of patient-specific radiation treatment delivery in radiotherapy. It is not an AI/ML powered device in the context of diagnostic or predictive tasks.

However, I can extract the closest analogous information available from the document regarding performance comparison against its predicate device.

Here's a breakdown of what can be extracted based on your request, with caveats where the information isn't directly present or relevant to AI device assessment:

1. Table of Acceptance Criteria and Reported Device Performance

Note: The document does not explicitly define specific numerical acceptance criteria for performance beyond stating that the measurements "had very good correlation" with the predicate device. This is typical for a 510(k) submission where substantial equivalence to a legally marketed predicate is being demonstrated for a physical quality assurance device, rather than a novel diagnostic AI algorithm.

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

• Sample Size for Test Set: Not explicitly stated in terms of a specific number of cases or measurements. The document mentions "numerous treatment plans."
• Data Provenance: Not specified (e.g., country of origin). The study involved "pre-treatment verification measurements." Given the manufacturer is Swedish, data may originate from European or other regions where the predicate device is used.
• Retrospective or Prospective: Not explicitly stated, but "pre-treatment verification measurements" suggest they were likely conducted as part of internal testing or possibly in a clinical setting but not necessarily as a formal prospective clinical trial for this submission.

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

• Not Applicable. For this type of physical quality assurance device, the "ground truth" and performance evaluation are typically based on physical measurement comparisons, not expert human interpretation for diagnostic tasks. The device itself is designed to measure physical radiation dose distributions. Performance is assessed against the expected dose distribution from the treatment plan and comparison with a known, validated device (the predicate).

4. Adjudication Method for the Test Set

• Not Applicable. As the "ground truth" is not established by human experts in a diagnostic context, there is no expert adjudication method like 2+1 or 3+1. The comparison is between the new device's measurements and those of the predicate device, against expected dose distributions.

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

• No. This document describes a physical measurement device for quality assurance in radiotherapy, not an AI diagnostic tool. Therefore, an MRMC study comparing human reader performance with and without AI assistance is not relevant or reported here.

6. Standalone (Algorithm Only Without Human-in-the-Loop Performance) Study

• Yes, analogous to standalone performance. The "performance" assessment described for the Delta4 Phantom+ is inherently standalone. The device measures dose distributions and compares them to intended distributions via its integrated software. The validation involved comparing these measurements against the predicate device. Human involvement would be in operating the device and interpreting its outputs, but the measurement and comparison itself is an algorithmic function of the device. The "very good correlation" indicates its standalone performance similarity to the predicate.

7. Type of Ground Truth Used

• The "ground truth" for evaluating this device's performance is multi-faceted:
  • Intended Dose Distribution: The theoretical dose distribution prescribed by the treatment planning system.
  • Predicate Device Measurements: Measurements obtained from the legally marketed predicate device (Delta4), which is assumed to be accurate and provide a valid reference.

8. Sample Size for the Training Set

• Not Applicable. This is a physical hardware and software device for measurements, not an AI/ML model that is "trained" on a dataset in the conventional sense. The device's algorithms are based on physics and signal processing principles.

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

• Not Applicable. See point 8. No "training set" or "ground truth for training set" in the context of an AI/ML model is relevant to this device. Its operational principles are founded on established physics and engineering.

Ask a specific question about this device