(246 days)
The multifunctional ultrasound scanner is used to collect, display and analyze ultrasound images during ultrasound imaging procedures in combination with supported echographic probes.
Main application:
-
Cardiac
Districts: Cardiac Adult, Cardiac Pediatric
Invasive access: Transesophageal -
Vascular
Districts: Neonatal, Adult Cephalic, Vascular
Invasive access: Not applicable -
General Imaging
Districts: Abdominal, Breast, Musculoskeletal, Neonatal, Pediatric, Small Organs (Testicles), Thyroid, Urological
Invasive access: Intraoperative (Abdominal), Laparoscopic, Transrectal -
Women Health
Districts: OB/Fetal, Gynecology
Invasive access: Transrectal, Transvaginal
The primary modes of operation are: B-Mode, M-Mode, Tissue Enhancement Imaging (TEI), Multi View (MView), Doppler (both Pulsed Wave (PW) and Continuous Wave (CW)), Color Flow Mapping (CFM), Power Doppler, Tissue Velocity Mapping (TVM), Combined modes, Elastosonography, 3D/4D and CnTI.
The ultrasound scanner is suitable to be installed in professional healthcare facility environment and is designed for ultrasound practitioners.
6600 Ultrasound System is a general-purpose diagnostic ultrasound system, based on a mainframe platform that can be easily moved thanks to four swivelling wheels.
6600 Ultrasound System consists of a control panel assembly with LCD monitor and a console with the device electronics and connectors, housed in an ergonomic cart designed to be both highly mobile and adjustable for a range of users and operating conditions.
6600 Ultrasound System use the physical properties of the ultrasound (i.e. sound waves with frequency above 20 kHz and that are not audible to the human ear) for the visualization of deep structures of the body by recording the reflections or echoes of ultrasonic pulses directed into the tissues and of the Doppler effect, i.e. the frequency-shifted ultrasound reflections produced by moving targets (usually red blood cells) in the bloodstream, to determine both direction and velocity of blood flow in the target organs.
The primary modes of operation are: B-Mode, M-Mode, Tissue Enhancement Imaging (TEI), Multi View (MView), Doppler (both PW and CW), Color Flow Mapping (CFM), Power Doppler, Tissue Velocity Mapping (TVM), Combined modes. 6600 Ultrasound System also manages Elastosonography, 3D/4D and CnTI.
Several types of probes are used to cover different needs in terms of geometrical shape and frequency range.
6600 Ultrasound System can drive Phased Array, Convex Array, Linear Array, Doppler probes and Volumetric probes (Bi-Scan probes). The control panel is equipped with a pull out Qwerty alphanumeric keyboard that allows data entry. The touchscreen has an emulation of the Qwerty keyboard that allows data entry and has additional controls and mode-depending keys, integrated in the touchscreen.
6600 Ultrasound System is equipped with wireless capability.
6600 Ultrasound System will be available on the market in two models with the following commercial names: MyLabA50, MyLabA70. The difference between MyLabA50 and MyLabA70 models is only in the licenses configuration.
6600 Ultrasound System, defined herein, introduces new features and accessories listed below:
-
AutoOB: AutoOB (Automatic Obstetric Biometric Measurement) is a tool based on A.I. algorithms that supports the clinician in performing the Obstetric Biometric Measurements during an Obstetric ultrasound examination.
-
AutoCM: AutoCM (Automatic Cardiac Measurement), is a tool based on A.I. algorithm that supports the clinician in performing the Cardiac Measurements during a Cardiac ultrasound examination.
-
XStrain RV: XStrain allows clinicians to quantify endocardial velocities of contraction and relaxation and local deformation of the heart (Strain/Strain rate). XStrain RV is an advanced processing package for the Right Ventricle analysis.
-
New probes: C 1-8E, L 3-15E and P 1-5E, available for MyLabA50 model.
-
New probes: C 1-8A and P 1-5A available for MyLabA70 model.
6600 Ultrasound System employs the same fundamental technological characteristics as its predicate device cleared via K230179.
The document describes the FDA 510(k) clearance for the Esaote 6600 Ultrasound System, specifically highlighting the AI-powered features: AutoOB (Automatic Obstetric Biometric Measurement) and AutoCM (Automatic Cardiac Measurement). The study performed aimed to demonstrate the statistical equivalence between the AI-powered automatic measurements and manual measurements.
Here's a breakdown of the acceptance criteria and the study details for the AI functionalities:
AutoOB (Automatic Obstetric Biometric Measurements) Feature AI-powered
Acceptance Criteria and Reported Device Performance
For Scan Plane Classification Algorithm:
| Acceptance Criteria (Success Rate) | Reported Performance (Success Rate) |
|---|---|
| Head TT/TV plane: > 90% | In line with criteria |
| Head TCD plane: > 90% | In line with criteria |
| Abdomen plane: > 90% | In line with criteria |
| Bones: > 90% | In line with criteria |
| Sagittal CRL: > 90% | In line with criteria |
| Sagittal NT: > 85% | In line with criteria |
For Automatic Measurement Algorithm:
| Acceptance Criteria (Success Rate) | Reported Performance (Success Rate) |
|---|---|
| Head Circumference (HC): >= 90% | In line with criteria |
| Biparietal Diameter (BPD): >= 90% | In line with criteria |
| Abdominal Circumference (AC): >= 90% | In line with criteria |
| Femur Length (FL): >= 75% | In line with criteria |
| Crown Rump Length (CRL): >= 75% | In line with criteria |
| Transverse cerebellar diameter (TCD): >= 90% | In line with criteria |
| Humerus Length (HL): >= 90% | In line with criteria |
| Ulna Length (UL): >= 90% | In line with criteria |
| Tibia Length (TL): >= 90% | In line with criteria |
| Statistical Equivalence: | |
| Automatic and manual measures are statistically equivalent (not rejecting the null hypothesis) with 95% confidence level. | In line with criteria |
Note: The document states "All test results are in line with the acceptance criteria" for both algorithms, indicating that the reported performance met the acceptance criteria.
Study Details for AutoOB
-
Sample sizes used for the test set and the data provenance:
- Scan plane classification algorithm: 265 images (test dataset)
- Automatic measure algorithm: 521 images (test dataset)
- Data Provenance: Based on female, pregnant, Caucasian patients. The document does not explicitly state the country of origin but implies data was collected by Esaote (an Italian company) and its predicate device users. The information available suggests it's retrospective data, as images were "saved during the examinations."
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- Number of Experts: Two experts established the ground truth for both training and test datasets.
- Qualifications: Clinicians specialized in Radiology with 30 and 24 years of experience in Ob-fetal ultrasound imaging.
-
Adjudication method for the test set:
- Method: Consensus reading. Each expert contributed to the annotation, then reviewed the annotations of the other. A consensus reading was done whereby the two radiologists discussed if they agreed on or not.
-
If a multi-reader multi-case (MRMC) comparative effectiveness study was done: No, an MRMC study comparing human readers with and without AI assistance was not reported for AutoOB. The study focused on the equivalence between manual and AI measurements, rather than human reader performance improvement.
-
If a standalone (i.e., algorithm only without human-in-the-loop performance) was done: Yes, the study evaluates the performance of the algorithm in classifying scan planes and performing measurements automatically, comparing them to expert-derived ground truth. This is a standalone evaluation of the algorithm's accuracy.
-
The type of ground truth used: Expert consensus by a panel of two radiologists.
-
The sample size for the training set:
- Scan plane classification algorithm: 25597 images
- Automatic measure algorithm: 11698 images
-
How the ground truth for the training set was established: The consensus of the same expert panel (two radiologists with 30 and 24 years of experience) was used as ground truth for the training datasets, following the same adjudication method.
AutoCM (Automatic Cardiac Measurements) Feature AI-powered
Acceptance Criteria and Reported Device Performance
For Segmentation and Measurement Algorithm:
| Acceptance Criteria (Success Rate) | Reported Performance (Success Rate) |
|---|---|
| IVS: > 80% | In line with criteria |
| LVID: > 90% | In line with criteria |
| LVPW: > 70% | In line with criteria |
| Statistical Equivalence: | |
| Automatic and manual measures are statistically equivalent (not rejecting the null hypothesis) with 95% confidence level. | In line with criteria |
Note: The document states "All test results are in line with the acceptance criteria," indicating that the reported performance met the acceptance criteria.
Study Details for AutoCM
-
Sample sizes used for the test set and the data provenance:
- Test set sample size: 100 images
- Data Provenance: Based on both female and male Caucasian adult patients. The test dataset was "acquired and labelled in a different medical center" than where the training data experts established ground truth, implying prospective or at least independently collected retrospective data. The country of origin is not explicitly stated.
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- Number of Experts for Test Set Ground Truth: One expert.
- Qualifications for Test Set Ground Truth: Clinician specialized in Cardiology with 36 years of experience.
- (For training set ground truth, two cardiologists with 30 and 24 years of experience were used).
-
Adjudication method for the test set:
- Method: The test dataset was labelled by a single clinician. Therefore, no formal adjudication of multiple readers on the test set is reported for the AutoCM feature. (For the training set, a consensus reading of two cardiologists was performed).
-
If a multi-reader multi-case (MRMC) comparative effectiveness study was done: No, an MRMC study comparing human readers with and without AI assistance was not reported for AutoCM. The study focused on the equivalence between manual and AI measurements.
-
If a standalone (i.e., algorithm only without human-in-the-loop performance) was done: Yes, the study evaluates the performance of the algorithm in performing cardiac measurements automatically, comparing them to expert-derived ground truth. This is a standalone evaluation of the algorithm's accuracy.
-
The type of ground truth used: Expert ground truth. For the training set, it was expert consensus (two cardiologists). For the test set, it was established by a single expert.
-
The sample size for the training set: 2011 images
-
How the ground truth for the training set was established: The consensus of an expert panel consisting of two clinicians specialized in Cardiology with 30 and 24 years of experience was used. Each contributed to the annotation and then reviewed the annotations of the other, with a consensus reading to resolve disagreements.
§ 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.