Search Results

Dosimetry Check is a standalone software product intended to be used by an experienced radiological physicist for quality control purposes only. Dosimetry Check is intended to check the correctness of x-ray treatment plans delivered from high energy charged-particle radiation therapy treatment machines by using a measurement of the applied radiation fields that are planned to be or have been applied to a patient, and computing the dose to the patient from the measured radiation fields. This product is to be used as a quality control check for the treatment planning system and delivery system.

Dosimetry Check quality control software uses the radiation fields that are measured with media such as x-ray film, electronic portal imaging devices (EPID), diode or ion chamber arrays, or in the case of TomoTherapy, a fan line detector array, and provides a theoretical calculation. Dosimetry Check computes the dose and dose distribution using the patient specific CT or other image set or alternately a phantom that is likewise scanned. to calculate the reconstructed dose that is then compared to the plan dose. The results reported can include the computed percent difference at specific points as compared to the patient specific radiation treatment plan.

Dosimetry Check does not provide any conclusions regarding the comparisons and does not provide any criteria to be used for interpreting the results. The experienced radiological physicist can reevaluate his patient specific radiation treatment plan in accordance with his clinical judgment.

This product is not a treatment planning system and is not to be used as one. This product only checks the applied dose based on the measurement of each x-ray field applied to the patient and provided in an exported file, and a theoretical calculation. This product does not provide any quality assurance that the fields are in fact correctly applied to and correctly aligned with the patient anatomy as planned. In addition. the product may be used to display the above dose on other fused image sets which could provide additional supportive quality information to the user regarding the correctness of treatment.

System 2100 for which 510(k) K993530 was cleared by the FDA on December 15, 1999 that is a medical image display system serves as a foundation that provides basic image display functionality for Dosimetry Check.

Dosimetry Check is a software program that will compute the dose and dose distribution to the patient from a measurement of the radiation fields that are applied to the patient. The dose so computed serves as a means to verify the correctness of the radiation treatment and to serve as a final sanity check. The radiation fields are measured with media such as x-ray film or electronic devices that will measure over the area of the field, such as electronic portal imaging devices (EPID), or diode or ion chamber arrays.

To extend Dosmetry Check to support the TomoTherapy machine, the device uses the data measured by the fan beam radiation detector that is part of the TomoTherapy machine. The detectors capture the radiation intensity periodically at predetermined gantry angles and couch positions. known as control points, from the treatment plan. The detector only measures the intensity across the center of the radiation beam in the transverse plane. A prior measured profile in the perpendicular longitudinal direction is then applied to complete the radiation field map. The radiation field map is then applied as a stationary beam at the center gantry angle and couch position for the integration period (between two control points), from which the dose to the patient is computed. The patient dose is then summed up from all such radiation field maps.

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

1. Table of Acceptance Criteria and Reported Device Performance

The provided text does not explicitly state specific numerical acceptance criteria for the Dosimetry Check version 4 release 1. Instead, it relies on demonstrating equivalence to the predicate device and successful external validation. The performance is reported qualitatively.

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

• Sample Size for Test Set: Not explicitly quantifiable from the provided text in terms of individual patient cases or measurements. The external validation was performed at 2 sites in Europe and 3 sites in the US. The text mentions "All aspects of using Dosimetry Check for quality control for TomoTherapy were tested during the beta testing." This suggests a comprehensive evaluation rather than a fixed number of cases.
• Data Provenance (Retrospective/Prospective, Country of Origin): The external validation involved prospective "beta testing" at 2 sites in Europe and 3 sites in the US. It doesn't specify if the data used for internal testing (modular, regression, verification) was retrospective or prospective, or its origin.

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

The text does not explicitly state the number of experts used to establish ground truth or their specific qualifications (e.g., "radiologist with 10 years of experience"). It mentions that the device is "intended to be used by an experienced radiological physicist for quality control purposes only." It is implied that these "experienced radiological physicists" at the beta test sites would be the ones evaluating the device's output against their clinical judgment or established facility standards, effectively forming the "ground truth" for the device's utility in their context.

4. Adjudication Method for the Test Set

The document does not specify an adjudication method (e.g., 2+1, 3+1). The "beta testing" implies evaluation by the "intended users" (experienced radiological physicists), but the process of reaching consensus or resolving discrepancies is not described.

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. This device is a quality control software for dose calculation and comparison, not an AI-assisted diagnostic tool for human readers. It's intended to verify treatment plans, not to assist humans in interpreting images or making diagnoses in the typical MRMC study context. Therefore, there is no discussion of human reader improvement with or without AI assistance.

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

Yes, in essence. Dosimetry Check is described as "standalone software" and its primary function is to "compute the dose and dose distribution" and "compare to the plan dose." While it is used by an "experienced radiological physicist," the software itself performs the calculation and comparison autonomously. The physicist then interprets the results. This aligns with the concept of standalone algorithmic performance in its core function. The "Nonclinical Testing" (modular, regression, verification, installation, performance testing) likely evaluated the algorithm's standalone accuracy in these computations.

7. The Type of Ground Truth Used

The ground truth used for assessing the device appears to be the patient-specific radiation treatment plan (planned dose) against which the device calculates and compares a "reconstructed dose" from measured radiation fields. The text states: "...to calculate the reconstructed dose that is then compared to the plan dose." The overall evaluation of performance during external validation would be based on whether the device's output (the comparison) was deemed acceptable and useful by the "experienced radiological physicist" in their quality control workflow.

8. The Sample Size for the Training Set

The document does not explicitly state a sample size for a training set. Given the nature of this software (dose calculation and comparison based on physical principles), it's more likely developed using well-established physics models and possibly calibrated with phantom measurements, rather than being "trained" on a large dataset in the machine learning sense. The kernel for deconvolution in the predicate device (Dosimetry Check version 3) was derived from "phantom measurements," which might be analogous to a calibration or parameter tuning set, but not a typical "training set" for AI.

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

As there's no explicit "training set" mentioned in the context of machine learning, the question isn't directly applicable. However, if we consider calibration or model parameter derivation:

• For the predicate device, the "kernel" used in the deconvolution process was "derived prior from phantom measurements with the same imaging device or media." This suggests the ground truth for these derivations would be known physical properties of phantoms and direct measurements using established dosimetry techniques.
• For the modified device, the extension to TomoTherapy uses a "fan beam radiation detector" and "a prior measured profile in the perpendicular longitudinal direction is then applied to complete the radiation field map." The establishment of these "prior measured profiles" would similarly rely on physical measurements and known ground truth for radiation fields.

Ask a specific question about this device