(129 days)
The system is a diagnostic ultrasound imaging system used by qualified and trained healthcare professionals for ultrasound imaging, human body fluid flow analysis and puncture and biopsy guidance.
The clinical applications and exam types include: Fetal (including Obstetrics), Abdominal, Pediatric, Small Organ (also known as Small Parts), Adult Cephalic is also known as Adult trans-cranial), Neonatal Cephalic, Intra-operative, Intra-operative Neuro (also known as Neurosurgery). Trans-raginal. Musculo-skeletal (Conventional and Superficial), Cardiac Adult, and Peripheral Vessel (also known as Peripheral Vascular).
Modes of Operation:
- · 2D (B-Mode) (including Tissue Harmonic imaging)
- · M-Mode
- · Vector Flow Imaging (VFI)
- Strain Elastography
- · CW Doppler
- · Contrast Imaging
- · PWD Mode
- · CFM Mode (Includes Color Doppler and Amplitude (Power) Doppler)
The Ultrasound System 1300 is a multi-purpose mobile, software-controlled diagnostic ultrasound system with an on-screen display for thermal and mechanical indices related to potential bio-effect mechanisms.
The system consists of a mobile console (engine) that provides digital acquisition, processing and display capabilities. The user interface includes a glass touchpad, a 19" Clinical display monitor (CDM). In addition to this a variety of system accessories are available such as baskets, foot switch, printer start-up kit, and extra holders.
Here's a breakdown of the acceptance criteria and study details for the Prostate Volume Assist (PVA) AI feature, based on the provided text:
1. Table of Acceptance Criteria and Reported Device Performance
| Acceptance Criteria (Stated Goal) | Reported Device Performance |
|---|---|
| Segmentation algorithm performance (standard metric of similarity for testing) | "The test passed." (No specific metric value provided) |
| Initial caliper placement for 9018 transducer (compared to manual placement) | Expected deviation: 11% +/- 6% |
| Initial caliper placement for 9048 transducer (compared to manual placement) | Expected deviation: 7% +/- 15% |
| Workflow improvement (overall goal of PVA) | Verified by visual evaluations by clinical experts validation testing, comparing initial algorithm placement to manual placement. The purpose was to compare PVA volume calculation with clinical personnel's volume measurement. |
2. Sample Size and Data Provenance
- Test Set (Clinical Validation):
- Healthy Images: 975
- Diseased Images: 1461 (includes images from the Holland dataset specifically used for testing)
- Total Test Images: 2436
- Data Provenance: Czech Republic, Holland.
- Retrospective/Prospective: The text does not explicitly state retrospective or prospective, but as it's referred to as "data acquired in Amsterdam, Holland" and "data from curved and linear arrays," it implies retrospective use of existing data for a test set. The use of a "Holland dataset that are only used as Test images" further supports a retrospective approach.
3. Number of Experts and Qualifications for Ground Truth Establishment
- Number of Experts: Not explicitly stated as a single number, but referred to as "clinical personnel" who had "an average of about 15 years' experience in ultrasound." This suggests multiple experts.
- Qualifications: Clinical personnel with "an average of about 15 years' experience in ultrasound." (Specific roles like radiologist or sonographer are not mentioned, but the context implies experienced ultrasound practitioners).
4. Adjudication Method for the Test Set
- Adjudication Method: Not explicitly stated as a formal adjudication process like 2+1 or 3+1. However, the "human caliper setters scored the images individually," and the comparison was made between the "automatic initial placement" and "manual placement by clinical personnel." This indicates a direct comparison to individual expert measurements, rather than a consensus-driven ground truth process for each image.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
- Was an MRMC study done? A formal MRMC study, as typically understood for comparative effectiveness of AI vs. human readers, was not explicitly detailed. The study focused on comparing the AI's initial caliper placement and volume calculation against manual placement by clinical personnel, framed as a "workflow improvement."
- Effect Size of Human Reader Improvement: Not applicable, as the study design was not an MRMC to measure human reader improvement with AI assistance. The AI's performance (initial caliper placement deviation) was measured against human manual placement.
6. Standalone (Algorithm Only) Performance Study
- Was a standalone study done? Yes, the "segmentation part of the algorithm is tested using a standard metric of similarity," which represents a standalone performance test. The numerical result for this specific metric is not provided beyond "The test passed."
- The comparison of "automatic initial placement" to "manual placement by clinical personnel" also evaluates the standalone performance of the AI's caliper placement, even though it's compared against human input.
7. Type of Ground Truth Used
- Ground Truth Type: Expert consensus for the training data (implied, see point 9) and individual expert manual measurements for the test data ("The validator used 2 distance measurements (using 4 calipers) on the transverse image and one distance measurement (using 2 calipers) on the sagittal image."). The purpose was to compare the prostate volume calculation of PVA with volume measurement performed by clinical personnel. So the ground truth for volume measurement is based on clinical personnel's manual calculations.
8. Sample Size for the Training Set
- Healthy Images: 505
- Diseased Images: 13447
- Synthesized Data: 4104
- Total Training Images: 18056
9. How the Ground Truth for the Training Set was Established
- The document implies that the ground truth for the training set was established through expert annotation/labeling, as it refers to the test volumes being "completely distinct from the ones used during training process" and mentions that the algorithm uses both "segmentation and caliper placement." The process of generating synthesized data also suggests a controlled method for creating labeled examples, but the specifics of human involvement in training data ground truth are not detailed beyond inferring that it must have been expert-derived to train an algorithm providing "initial caliper placement."
§ 892.1550 Ultrasonic pulsed doppler imaging system.
(a)
Identification. An ultrasonic pulsed doppler imaging system is a device that combines the features of continuous wave doppler-effect technology with pulsed-echo effect technology and is intended to determine stationary body tissue characteristics, such as depth or location of tissue interfaces or dynamic tissue characteristics such as velocity of blood or tissue motion. This generic type of device may include signal analysis and display equipment, patient and equipment supports, component parts, and accessories.(b)
Classification. Class II.