(15 days)
The BioJet software is intended to be used by physicians in the clinic or hospital for 2D and 3D visualization of ultrasound images of the prostate gland. Additional software features include patient data management, multiplanar reconstruction, segmentation, image measurements, and 3-D registration.
The BioJet software is designed to display the 2-D live video received from commercially available ultrasound machines and use this 2-D video to reconstruct a 3-D ultrasound image stack. The system has been designed to work with the clinicians' existing ultrasound machine and TRUS (TransRectal UltraSound) probe, commercially available needles, needle guides or needle gun combination. Additional software features include patient data management, multi-planar reconstruction, image measurements and 3-D image registration. BioJet works with commercially available mechanical stepper and stablizer assemblies that holds the ultrasound probe and tracks the probe position while the physician performs a normal ultrasound imaging procedure of the subject prostate. The mechanical tracker is connected to a PC-based workstation containing a video image converter. Control of the ultrasound probe and ultrasound system is done manually by the physician, just as it would be in the absence of BioJet. However, by tracking the position and orientation of the ultrasound probe while capturing the video image, BioJet is able to reconstruct and display a 3-D rendered surface model of the prostate and to display the live image position within the prostate. Locations for biopsies, needles, markers, and other devices may be selected by the physician, displayed in the 3-D image stack and 3-D rendered surface model, and stored. Previously created 3-D models may be recalled and may be aligned or registered to the current live display of the prostate. The 3-D model used for co-registration may be based on another series of ultrasound images or DICOM images. Finally, the physician may attach a commercially available biopsy needle guide to the TRUS probe and use the probe and biopsy needle to perform biopsies. Whenever the ultrasound machine is turned on by the physician, the live 2-D ultrasound image is displayed on the screen of BioJet during the biopsy. As the TRUS probe with attached needle guide or needle grid is maneuvered by the position and orientation of the probe is tracked. BioJet is able to add, display and edit plans for biopsy sites as well as an estimate of the probe position and needle trajectory relative to the 3-D image series and rendered surface model of the prostate. BioJet offers the physician additional 3 -D information for assessing prostate abnormalities, planning and implementing biopsy procedures. The scientific concepts on which the BioJet Software is based is that organs such as the prostate can be visualized with a number of imaging modalities. All imaging modalities display the organs in a different way and deliver different information but they all display the same shape and size of the organ. Thus it is possible to co-register a a previously acquired 3-D model of a prostate to live ultrasound images acquired with a commercially available endo-rectal probe. The BioJet Software provides two and 3 dimensional image review, manipulation, and analysis tools to assist users in planning and performing image-guided interventional procedures such as biopsies and the placement of instruments and markers. Supported imaging modalities include DICOM 3 images and 2D live video images received from commercially available ultrasound (US) machines. DICOM images are received from various commercially available imaging systems via a memory stick or a CD ROM. Live video US images are collected from the video output stream manually or triggered by a commercially available tracking stepper, i.e. a calibrated spatial positioning device. Non DICOM video images are - in contrary to DICOM images - not calibrated and must be scaled within the program based on scaling marks in the image and the known image spacing if collected manually. All know images from US devices include such scaling marks. The contrast and brightness display of the images can be changed manually by the user. Additionally, the user can zoom and pan the images and also change the image coloration for better visibility. The images in the image stack can be interpolated by linear interpolation to allow an image display at 1 mm intervals. This device provides the capability to overlay annotations on 2D medical image displays. These annotations may represent the position of instruments including but not limited to biopsy needles, imaging probes or other tracking devices. Additionally, the user can manually draw contours of structures (prostate, urethra, seminal vesicals, bladder, etc.) and regions of interest (ROI) into the images. The 3-D image stack can be further processed to perform volume estimations based on the manually drawn organ contours or ROIs drawn by a physician. Patient information, notes, and images may be stored unchanged - for future retrieval. In-plane length and angle measurements are available. The angle data pertains to the biopsy guidelines, biopsy cores, markers and instruments used. Thus only volume, length, and angle measurements are conducted. All images independant of the image format (JPEG, BMP, DICOM, PNG) are stored and displayed unchanged. JPEG images may have been created with lossy compression. For that reason the user is informed that JPEG images may have been created using a lossy compression. Images may not be used for diagnostic purposes. Live 2-D ultrasound images are used for example during biopsy or placement of instruments. As the TRUS probe with the attached needle guide is maneuvered by the physician, the position and orientation of the probe is tracked manually or by an attached tracked device. The ultrasound image contains the available device trajectories that are manually match to the trajectories in the program. Thus the complete biopsy procedure is controlled by the user by observing the live uttrasound images and matching the respective trajectories. The user may load image display test patterns into the program at any time for quality assessment using one of the allowed imaging modalities. These test patterns can be used, for example, to check the quality of image re-slicing, grayscale or color depiction, measurement precision or screen resolution and display. All data and images are stored in the patient file for later retrieveal. Image are always stored as unchanged originals. The files are CRC checksum protected zip files. When a checksum is received or read, the device performs a CRC on the data and compares the resulting check value with the received one. If the check values do not match, then the block contains a data error. Otherwise, the data is assumed to be error-free. CRCs are specifically designed to protect against common types of errors on communication channels, where they provide assurance of the integrity of messages delivered. The zip format uses a 32-bit CRC algorithm and includes two copies of the directory structure of the respective file to provide greater protection against data loss.
Here's an analysis of the provided text regarding the BioJet medical device, focusing on acceptance criteria and study details:
1. Table of Acceptance Criteria and Reported Device Performance
The provided text does not explicitly state specific quantitative acceptance criteria for the BioJet device. Instead, it describes performance verification against general product and engineering specifications using test phantoms. The primary claim is "substantially equivalent" to a predicate device.
However, based on the narrative, we can infer some performance aspects that were tested and reported:
| Performance Aspect | Acceptance Criteria (Inferred) | Reported Device Performance |
|---|---|---|
| Biopsy Guideline Alignment | Precise alignment of biopsy guidelines in transverse and longitudinal sections with ultrasound images. | "Biopsy guidelines in transverse and longitudinal section are precisely aligned with the biopsy guidelines on the ultrasound images." |
| Longitudinal Contour Fit | Calculated longitudinal contours from transverse drawing fit onto actual longitudinal phantom contours. | "Demonstrated that the contours of the longitudinal display calculated from the transverse contour drawing fit onto the actual longitudinal contours of the phantom." |
| Volume Measurement Accuracy | Calculated volume within a specific error margin compared to commercially available ultrasound device. | "Volume tests proved that the calculated volume is within an error of less than 3% equal to the volume displayed by the shelf commercially available ultrasound device used." |
| Temporal Resolution (Graphics) | No detectable time delay in graphic functions. | "All graphics functions are accessible in real-time, i.e. the user cannot detect any time delay." |
| Temporal Resolution (Biopsy Needles) | Dependent on ultrasound machine's video frequency (standard NTSC). | "The temporal resolution the biopsy needles depends only on the video frequency of ultrasound machine used which is the standard NTSC video signal." |
2. Sample Size Used for the Test Set and Data Provenance
- Sample Size for Test Set: Not explicitly stated. The text mentions "Test phantoms incorporating simulated prostates were used" but does not quantify the number of phantoms or tests performed.
- Data Provenance: The study was conducted using "test phantoms incorporating simulated prostates." This suggests a controlled laboratory setting, not real patient data. The country of origin of the data is not specified beyond the company's location in Romania, but the testing was likely internal. The study is prospective in the sense that it involved experiments run specifically to test the device, rather than analyzing pre-existing clinical data.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
- Number of Experts: Not specified.
- Qualifications of Experts: Not specified. Given the use of test phantoms and the nature of the measurements (alignment, volume, real-time access), the ground truth would likely be established by the engineering and testing team responsible for the phantom setup and measurement, rather than clinical experts. For example, the "actual longitudinal contours of the phantom" would be part of the phantom's design specification/ground truth.
4. Adjudication Method for the Test Set
- Adjudication Method: Not applicable or not specified. With phantoms and objective measurements (like volume calculation or alignment against known phantom geometry), there isn't typically an adjudication process by multiple experts as there would be with subjective clinical interpretations. The measurements are likely compared against pre-defined, objective reference values from the phantoms.
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
- MRMC Study: No, an MRMC comparative effectiveness study was not conducted or reported in the provided text.
- The BioJet device is primarily a visualization and planning tool rather than an AI-driven diagnostic aid that would typically be evaluated with MRMC studies to assess human reader performance improvement. It assists physicians with "planning and implementing biopsy procedures" but states it "do not steer or in anyway control the positioning of the instruments used or of any treatment process what so ever."
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) was done
- Standalone Performance: Yes, the described performance tests appear to be standalone in nature, evaluating the software's inherent ability to perform tasks like image reconstruction, measurement, and alignment against known phantom properties. The volume calculations and alignment checks are objective measurements of the algorithm's output. While a "physician" is mentioned as controlling the ultrasound machine, the performance tests themselves focus on the software's accuracy in processing and displaying information.
7. The Type of Ground Truth Used (expert consensus, pathology, outcomes data, etc.)
- Type of Ground Truth: The ground truth for the performance tests was based on the known characteristics and measurements of the test phantoms ("simulated prostates"). For example, the actual volume of the simulated prostate in the phantom would be a known ground truth, and the physical alignment of biopsy guidelines on the phantom would serve as ground truth for alignment checks.
8. The Sample Size for the Training Set
- Sample Size for Training Set: Not applicable / Not specified. The document describes a "Medical Imaging Processing Software System" that uses "scientific concepts on which the BioJet Software is based is that organs such as the prostate can be visualized with a number of imaging modalities... all display the same shape and size of the organ." This suggests a rule-based or model-based image processing approach, rather than a machine learning model that requires a "training set." If any training data was used (e.g., for image recognition components), it is not mentioned.
9. How the Ground Truth for the Training Set Was Established
- Ground Truth for Training Set: Not applicable / Not specified, as no training set (in the machine learning sense) is mentioned.
§ 892.2050 Medical image management and processing system.
(a)
Identification. A medical image management and processing system is a device that provides one or more capabilities relating to the review and digital processing of medical images for the purposes of interpretation by a trained practitioner of disease detection, diagnosis, or patient management. The software components may provide advanced or complex image processing functions for image manipulation, enhancement, or quantification that are intended for use in the interpretation and analysis of medical images. Advanced image manipulation functions may include image segmentation, multimodality image registration, or 3D visualization. Complex quantitative functions may include semi-automated measurements or time-series measurements.(b)
Classification. Class II (special controls; voluntary standards—Digital Imaging and Communications in Medicine (DICOM) Std., Joint Photographic Experts Group (JPEG) Std., Society of Motion Picture and Television Engineers (SMPTE) Test Pattern).