(53 days)
SoftVue™ is indicated for use as a B-mode ultrasonic imaging of a patient's breast when used with an automatic scanning curvilinear array transducer. The device is not intended to be used as a replacement for screening mammography.
SoftVue™ is indicated for use as a B-mode ultrasonic imaging system for imaging of a patient's breast using a curvilinear array transducer that completely surrounds the breast. The modification of SoftVue™ will allow the system to generate colorized and grayscale relative stiffness ultrasound images in addition to the grayscale B-mode images generated by the unmodified SoftVue™ system. This new feature will allow the user to be able to determine whether a region of interest is harder or softer than the surrounding tissue. No clinical diagnostic claims are being made.
SoftVue™ is comprised of the following subsystems: the Transducer, Table/Housing Assembly, Water Conditioning System, Computer Control System, Image Reconstruction System, Power System, and the Data Acquisition System.
Soft Vue™ has a built-in curvilinear transducer that is used to acquire ultrasound data. Data are acquired from a patient lying prone on the table with their breast submerged in an imaging chamber filled with warm (body temperature) water. The breast is positioned in the center of the transducer. A camera, located at the bottom of the imaging chamber provides a live video feed to the system operator to aid in positioning the patient's breast. Once the scan is initiated, the transducer collects data that are processed to produce a series of B-mode ultrasound image slices that can be stacked to yield a volumetric ultrasound image of the breast. Soft Vue™ also outputs colorized and grayscale relative stiffness image stacks that provide the radiologist with additional reference information.
The Water Control System is used to de-gas and warm the water that is used as the image acquisition medium. The Computer Control System controls all of the functionality of the other subsystems. The reconstruction engine processes the image data acquired by the transducer ring into B-mode ultrasound images.
The system includes a barcode reader and a touchscreen display (user interface). The touchscreen display allows the user to perform an imaging procedure. Patient information is entered into the system using either the QWERTY keyboard on the touchscreen display, using the barcode reader, or the user can import the patient information from a DICOM modality worklist on an externally networked RIS server. Device errors and warnings are displayed on the touchscreen display.
SoftVue™ outputs the images to an externally networked PACS server which allows the images to be stored until they are reviewed on a workstation.
The provided text does not contain detailed acceptance criteria for a specific device performance study beyond safety testing and general system requirements. It describes modifications to an existing device ("SoftVue") and a small-scale clinical study for "image characterization" of a new feature (color relative stiffness images). The study is not presented as a formal performance study demonstrating device efficacy against predefined acceptance criteria.
However, I can extract information related to the device's technical specifications and the qualitative observations from the clinical study, which could be interpreted as a form of "reported performance" against an implicit goal of providing "qualitative tissue stiffness information."
Here's an attempt to structure the information based on your request, highlighting what is implicitly or explicitly stated and noting where information is absent:
1. Table of Acceptance Criteria and Reported Device Performance
| Acceptance Criteria (Implied/Explicit) | Reported Device Performance |
|---|---|
| Safety: | |
| Max. Mechanical Index (MI) ≤ 1.9 | 0.431 (Pass) |
| Max. ISPTA.3 ≤ 94 mW/cm² | 15.1 mW/cm² (Pass) |
| Max. Soft Tissue Thermal Index (TIS) (No specific threshold given, but implicit pass for safety) | 0.0006 (Implicitly Pass, as no issues raised and general safety standards met) |
| Conformity to IEC 60601-1:2005, IEC 60601-1-2:2007, IEC 62304:2006, IEC 60601-2-37:2007 | All safety standards were met. |
| System Functionality (General): | |
| All software unit and system requirements met | All requirements were met and no new issues of safety or effectiveness were raised. |
| Subsystem and System design requirements met | All requirements were met and no new issues of safety or effectiveness were raised. |
| Human Factors/Usability requirements met | All requirements were met and no new issues of safety or effectiveness were raised. |
| Intended Use requirements met | All requirements were met and no new issues of safety or effectiveness were raised. |
| Image Characterization (Qualitative Tissue Stiffness Information): | |
| Phantom Study: Correlation of color stiffness images with known phantom inclusion stiffness | 7 phantom inclusions: 1 cancer and 3 fibroadenomas (known stiff) appeared red; 3 cysts (known soft) appeared blue. This showed complete concordance with known properties. |
| Clinical Study: Qualitative assessment of lesion stiffness in correlation with pathology (where known/biopsied) | 13 imaged masses (4 cancers, 4 fibroadenomas, 5 cysts). |
- All 4 cancers were characterized as "stiff" (red/green, redder than background).
- 2 fibroadenomas were "mixed," 1 was "stiff" (red), and 1 was "soft" (blue).
- 4 cysts were "soft" (blue/green, bluer than background), while 1 was "mixed." This demonstrates SoftVue's ability to qualitatively measure tissue stiffness. |
| Image Quality: Improve image quality of B-mode images (implicit acceptance of improved image quality) | "Updates to the pre-processing of the attenuation image result in better gain correction and equivalent contrast for the B-mode image."
"Updates to the sound speed image parameters results in better delay correction and equivalent contrast for the B-mode image." (These are described as improvements, implying they met the goal of improved quality.) |
2. Sample Size Used for the Test Set and Data Provenance
- Test Set Sample Size:
- Phantom Study: 1 anthropomorphic breast phantom with 7 inclusions.
- Clinical Study (Image Characterization): 10 patient exams, yielding a total of 13 imaged masses.
- Breast cup size ranged from B to DDD.
- Breast densities: 5 scattered, 2 heterogeneous, 3 dense (one extremely dense).
- Data Provenance:
- Phantom Study: Not explicitly stated, but typically labs or manufacturers produce phantoms.
- Clinical Study: Prospective. "Patient data were collected as part of the clinical study that Delphinus currently has in process under pre-IED # I120143." The modified SoftVue system was installed at the Alexander J Walt Breast Center at the Karmanos Cancer Institute in Detroit, Michigan, indicating U.S. origin.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of those Experts
- Phantom Study: The "ground truth" for the phantom was based on its known, predetermined characteristics (stiff/soft inclusions). No human experts were required to establish this ground truth, but the qualitative assessment against the known properties would have been done by the study team.
- Clinical Study:
- Number of Experts: One expert.
- Qualifications: "board certified radiologist, Dr. Peter Littrup."
- Role: Reviewed all patient images from the SoftVue system and guided the identification of lesions.
4. Adjudication Method for the Test Set
- Phantom Study: Not applicable; the ground truth was inherent to the phantom's design.
- Clinical Study: No formal adjudication method involving multiple experts is described for the clinical image analysis. The "ground truth" (lesion identification and likely pathology) was guided by "Previous biopsies and/or imaging." The evaluation of the SoftVue color stiffness images was performed by a single board-certified radiologist (Dr. Peter Littrup), matching the SoftVue appearance/assessment against the lesion pathology and characteristics derived from prior imaging/biopsy.
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 study was for image characterization of a new feature, not explicitly a comparative effectiveness study of human readers with/without AI assistance. The device is an imaging system, and the new feature provides "additional reference information for radiologists," but its impact on reader performance or an effect size is not reported.
6. If a Standalone (i.e. algorithm only without human-in-the-loop performance) was done
- Yes, in a qualitative sense for the new feature. The device now generates and outputs color and grayscale relative stiffness images. The phantom study, in particular, could be seen as a standalone evaluation of the algorithm's ability to represent stiffness, as it directly correlated the device's output (color) to the known physical properties of the phantom inclusions without human interpretation as a primary outcome. The clinical observations by the radiologist then confirmed that these images provided "qualitative tissue stiffness information."
7. The Type of Ground Truth Used
- Phantom Study: Known, predetermined physical characteristics (stiffness, size) of inclusions within an anthropomorphic breast phantom.
- Clinical Study: A combination of "previous biopsies and/or imaging" to identify and characterize lesions (appearance, size, location) and their pathology (cancer, fibroadenoma, cyst).
8. The Sample Size for the Training Set
- The document describes modifications to an existing device ("SoftVue") and the evaluation of a new feature for that device (color and grayscale relative stiffness images). It does not mention the use of a "training set" in the context of machine learning or AI algorithm development for these new features. The focus is on verifying the system's performance and characterizing the new image type. Therefore, there is no explicit mention of a training set sample size. The "image characterization" study (both phantom and clinical) serves more as a validation/demonstration of the new feature's output.
9. How the Ground Truth for the Training Set Was Established
- As no training set is mentioned for the development of the new feature related to stiffness imaging, this question is not applicable based on the provided text. The document refers to updates in pre-processing and sound speed image parameters to improve image quality, which are likely engineering/physics-based improvements rather than AI model training.
§ 892.1560 Ultrasonic pulsed echo imaging system.
(a)
Identification. An ultrasonic pulsed echo imaging system is a device intended to project a pulsed sound beam into body tissue to determine the depth or location of the tissue interfaces and to measure the duration of an acoustic pulse from the transmitter to the tissue interface and back to the receiver. This generic type of device may include signal analysis and display equipment, patient and equipment supports, component parts, and accessories.(b)
Classification. Class II (special controls). A biopsy needle guide kit intended for use with an ultrasonic pulsed echo imaging system only is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 892.9.