Search Results

The VP-4440HD unit is used for endoscopic observation, diagnosis, treatment, and image recording. It is intended to process electronic signals transmitted from a video endoscope). This product may be used on all patients requiring endoscopic examination and when using a Fujinon/Fujifilm medical endoscope, light source, monitor, recorder and various peripheral devices. FICE is an adjunctive tool for gastromtestination which can be used to supplement FUJIFILM white light endoscopy. FICE is not intended to replace histopathological sampling as a means of diagnosis.

This system is intended to be used in coniunction with Fuiifilm endoscopes to provide illumination, visual display and data storage during endoscopic procedures.

The EPX-4440HD Digital Video Processor with FICE and light source consists of three components used in conjunction with one another:

The VP-4440HD Video Processor: The Processor relays the image from the endoscope to a video monitor. Projection can be either analog or digital at the user's preference. The Processor incorporates internal or external digital storage capacity. The Processor also controls the light projected to the body cavity. The Processor provides for optional structural enhancement at the user's option. Spectral and structural enhancements are achieved through proprietary software. The device is AC operated at a power setting of 120V/60MzJ0.8A. The Processor is housed in a steel-polycarbonate case measuring 390x105x460mm.

The XL-4450 Light Source: The Fujinon/Fujifilm endoscope employs fiber bundles to transmit light from the light source and subsequently to the body cavity. The Light Source employs a 300W Xenon lamp with a 75W emergency back-up Halogen lamp. Brightness control is performed by the user. The device is AC operated at a power setting of 120V/60MzJ3.3A. The Light Source is housed in a steel-polycarbonate case measuring 390x155x450mm.

The DK-4440E Keyboard: The Keyboard is used to enter pertinent procedural information (patient, physician, date, etc.) for display on the video monitor and digital/analog storage systems. The Keyboard is also used to control operational features of the VP-4440HD Processor. The Keyboard resembles a standard computer keyboard in size and shape.

The VP-4440HD Digital Video Processor with FICE and light source is the same as the cleared EPX-4440HD Digital Video Processor and Light source (K102466) device, with the exception of the addition of the imaging algorithm termed "FICE" (Flexible spectral-Imaging Color Enhancement). As with the predicate device, the subject processor receives an image from the endoscope via a connector cable. The subject processor is capable of bi-directional communication that allows importation and exportation of various patient and procedural data between EMR (Electronic Medical Record) systems.

FICE is an image processing feature that a licensed medical practitioner (user) can utilize, as an adiunctive tool, to supplement the white light endoscopic examination. FICE extracts spectral images of specific wavelength components from the original full spectral (white light) image through an image processing algorithm that, assigns a spectral wavelength component in the Red (R), Green (G) and Blue (B) spectrum and displays the enhanced color image. FICE may enhance color contrast to improve visibility of the selected structures, borders of areas of interests.

1. Table of Acceptance Criteria and Reported Device Performance

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

• Test Set (Clinical Study): Each subject underwent two (2) gastrointestinal endoscopic examinations during a single clinical procedure. The document doesn't explicitly state the total number of subjects or individual images in the test set.
• Data Provenance: Prospective clinical trial. The country of origin of the data is not specified.

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

• Number of Experts: Multiple blinded investigators. The exact number is not specified.
• Qualifications of Experts: Not specified.

4. Adjudication Method for the Test Set

• Adjudication Method: Images were analyzed independently by "multiple blinded investigators using a 4-point Likert scale." The document does not specify a formal adjudication method (e.g., 2+1, 3+1 consensus). It appears to rely on the "overall means between all readers."

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: Yes, a reader concurrence study was performed, which is a type of MRMC study comparing FICE (an AI/image processing feature) to optical filtering (NBI).
• Effect Size: The study aimed to establish "substantial equivalence of diagnostic visualization (e.g. image quality) to the predicate optical filtering method." The reported outcome is that "Overall means between all readers demonstrated at least comparable image quality of each FICE setting in one or more locations within the gastrointestinal tract." This indicates FICE performed similarly to or as well as NBI for image quality; a specific quantitative effect size of human reader improvement with AI assistance vs. without is not provided, as the study focused on comparability rather than improvement. The FICE feature itself is described as an "adjunctive tool," not necessarily a comprehensive AI diagnosis system.

6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done

• Based on the provided text, the FICE feature is described as an "adjunctive tool for gastrointestinal endoscopic examination which can be used to supplement FUJIFILM white light endoscopy." The clinical study focused on readers evaluating image quality generated by FICE. There is no indication of a standalone performance study of the FICE algorithm without human interpretation.

7. The Type of Ground Truth Used

• Type of Ground Truth: The clinical study compared the "overall image quality" observed by multiple blinded investigators using a Likert scale. This suggests a form of expert consensus/reader perception for image quality as the ground truth. It's important to note that the FICE feature is "not intended to replace histopathological sampling as a means of diagnosis," indicating that the "ground truth" for diagnosis in a clinical sense would still be pathology. However, for the purpose of this study measuring the effectiveness of the FICE image processing, reader evaluation of image quality served as the primary measure.

8. The Sample Size for the Training Set

• The document describes performance testing for an image processing feature (FICE) which is integrated into a video processor. It does not explicitly mention a "training set" in the context of machine learning model development. This suggests the FICE algorithm might be based on pre-defined spectral filters or rule-based image processing rather than a deep learning model that requires a dedicated training set. If there was a training phase for the algorithm, its sample size is not disclosed.

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

• As a dedicated "training set" is not mentioned and the FICE algorithm's nature isn't fully detailed as a deep learning model, the method for establishing ground truth for a training set is not applicable or not provided in this document.

Ask a specific question about this device

The VP-4440HD unit is used for endoscopic observation, diagnosis, treatment, and image recording. It is intended to process electronic signals transmitted from a video endoscope (a video camera in an endoscope). This product may be used on all patients requiring endoscopic examination and when using Fujinon medical Endoscope, light source, monitor, recorder and various peripheral devices.

The XL-4450 light source is used for endoscopic observation, diagnosis, treatment, and image recording. It is intended to provide illumination to an endoscope. The light source also functions as a pump to supply air through the endoscope while inside the body to assist in obtaining clear visualization to facilitate diagnostic examination. This product may be used on all patients requiring endoscopic examination and when using Fujinon medical Endoscope, video processor, monitor, recorder and various peripheral devices.

This system is intended to be used in conjunction with Fujinon endoscopes to provide illumination, visual display and data storage during endoscopic procedures.

The EPX-4440HD Digital Video Processor consists of three components used in conjunction with one another:

The VP-440HD Video Processor: The Processor relays the image from the endoscope to a video monitor. Projection can be either analog or digital at the user's preference. The Processor also incorporates internal or external digital storage capacity. The Processor also controls the light projected to the body cavity. The Processor also provides for optional structural enhancement at the user's option. Spectral and structural enhancements are achieved through proprietary software. The device is AC operated at a power setting of 120V/60MzJO.8A. The Processor is housed in a steel-polycarbonate case measuring 390X105X460mm.

The XL-4450 Light Source: The Fujinon endoscope employs fiber bundles to transmit light from the light source and subsequently to the body cavity. The Light Source employs a 300W Xenon lamp with a 75W emergency back-up Halogen lamp. Brightness control is performed by the user. The device is AC operated at a power setting of 120V/60MzJ3.3A. The Processor is housed in a steel-polycarbonate case measuring 390X155X485mm.

The DK-440E Keyboard: The Keyboard is used to enter pertinent procedural information (patient, physician, date, etc) for display on the video monitor and digital/analog storage systems. The Keyboard is also used to control operational features of the VP-4440HD Processor. The Keyboard resembles a standard computer keyboard in size and shape.

Here's an analysis of the provided text regarding the acceptance criteria and study for the Fujinon EPX-4440HD Digital Video Processor:

NOTE: The provided 510(k) summary focuses primarily on the safety and performance aspects of a modified device compared to a predicate device, rather than defining and proving specific acceptance criteria in the way one might for a novel AI/software-as-a-medical-device (SaMD) product. The "acceptance criteria" discussed are more about meeting established regulatory and engineering standards, and the "study" is functional testing with physician evaluators for general performance and safety.

1. Table of Acceptance Criteria and Reported Device Performance

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

• Sample Size for Test Set: Not explicitly stated. The document mentions "physician evaluators" without specifying how many cases, images, or evaluation instances were part of their assessment.
• Data Provenance: Not explicitly stated. The context suggests general functional testing, likely performed in a controlled environment by the manufacturer. It does not mention country of origin or whether clinical patient data was used. It's likely simulated or controlled test conditions given the nature of the evaluation.
• Retrospective/Prospective: Not explicitly stated, but given it's functional testing for a device modification, it would typically be a prospective evaluation of the new device's capabilities in a controlled setting.

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

• Number of Experts: Not explicitly stated. The document refers to "physician evaluators" (plural), but the exact number is not provided.
• Qualifications of Experts: Not explicitly stated beyond "physician evaluators." No specialty (e.g., gastroenterologist who performs endoscopy) or years of experience are listed.

4. Adjudication Method for the Test Set

• Adjudication Method: Not explicitly stated. The statement "In all cases they passed the required testing regimens" suggests a unanimous agreement or a pass/fail criterion that was met by all evaluators, rather than a formal adjudication process between conflicting interpretations.

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

• Was an MRMC study done? No, a traditional MRMC comparative effectiveness study as typically understood for AI algorithms (i.e., comparing human readers with and without AI assistance on diagnostic accuracy) was not performed. The study mentioned was a functional performance test for a hardware/software update, not an AI-assisted diagnostic tool.
• Effect size of human readers with vs. without AI assistance: Not applicable, as no such study was conducted.

6. Standalone (Algorithm Only Without Human-in-the-Loop Performance) Study

• Was a standalone study done? Yes, in a sense. The "software validation" and "electrical safety" testing would represent a standalone evaluation of the device's technical functionality. The "performance testing using physician evaluators" also tested the device itself, although with human interaction. However, this is not an AI algorithm performing a diagnostic task. The spectral and structural enhancements are "achieved through proprietary software," implying an algorithm, but its performance was evaluated by physicians using the system, not as a standalone diagnosis.

7. Type of Ground Truth Used

• Type of Ground Truth: For the "visual comparison tests conducted using physician evaluators," the ground truth was essentially the expert opinion/consensus of the physician evaluators regarding the successful functionality and acceptable image quality produced by the system. This is not pathology, outcomes data, or a pre-established clinical diagnosis.

8. Sample Size for the Training Set

• Sample Size for Training Set: Not applicable. This device is a video processor and light source, not a machine learning model that requires a "training set" in the conventional AI sense. While there is proprietary software for "spectral and structural enhancements," the document does not describe the development of this software in terms of machine learning training, but rather as a feature set of the device.

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

• How Ground Truth for Training Set Was Established: Not applicable, as there is no mention of a training set for a machine learning model. If the "proprietary software" involved some form of learned enhancement, the document does not provide details on how such a model would have been trained or its ground truth established.

Ask a specific question about this device