(75 days)
Intranasal or sinuses
Acute and chronic sinusitis, endoscopic dacryocystorhinostomy, optic nerve and orbital decompression, the removal of polyps, the biopsy and removal of tumors, and the repair of CSF leaks, pituitary disorder, and encephalocele
The device consists of a wheeled cabinet enclosure with a 20-inch color monitor mounted on the top. Mounted within the cabinet is a computer and a spacial tracking device. The electromagnetic, six-degree-of-freedom tracking device is linked to the computer, which provides the monitor with a display of the patient's CT image data and superimposed crosshairs, indicating the position of the tip of the surgical instrument used with the device. The device is controlled via software.
Here's a breakdown of the acceptance criteria and study information based on the provided text:
Note: The document is a 510(k) summary from 1996, so the depth of information regarding rigorous clinical trial methodology (especially regarding modern AI/machine learning studies) is limited. The device described uses image-guided surgery technology, not an AI algorithm in the contemporary sense.
1. Table of Acceptance Criteria and Reported Device Performance
| Acceptance Criteria (Stated or Implied) | Reported Device Performance |
|---|---|
| Nonclinical Testing | |
| a. Device Accuracy (Laboratory) | Mean device accuracy of 0.79 mm (compares to 1.74 mm for predicate device). |
| b. Electromagnetic Field Distortion Detection | Device detected field distortion under normal use conditions before induced error became larger than 1.0 mm. |
| c. Reproducibility of Replaceable Headset Location | Headset shown to be replaceable such that the overall average effect upon device accuracy was less than 0.74 mm. |
| d. Reproducibility of Replaceable Pointing Instruments Location | Removal and replacement of a pointing instrument resulted in a change of less than ±0.914 mm (±3 σ). |
| e. Electromagnetic Compatibility (EMC) | Satisfactorily passed: Emissions (EN55011/CISPR 11, RE101 of MIL-STD-461C) and Immunity (IEC 801-2, 801-3, 801-4, 801-5, RS101, CS114 of MIL-STD-461C). |
| f. Battery Backup (implied by 10 msec dropout requirement) | Device has a battery backup (due to satisfying 10 msec dropout). |
| Clinical Testing | |
| a. Device Mean Accuracy (Clinical) and Confidence Interval | Mean accuracy of 2.28 mm with a 95% confidence interval of the mean of 0.78 mm. (Compares to 1.8 mm to 4.8 mm for predicate device with CI of 1.1 mm to 1.6 mm). |
| General Safety | |
| a. Enclosure Risk Current | < 100 uA (ANSI/AAMI ES1-1993). |
| b. Patient Risk Current | < 10 uA (ANSI/AAMI ES1-1993). |
| c. Magnetic Field Intensity | 0.12 Gauss at 4 cm. (14 kHz). |
2. Sample Size Used for the Test Set and Data Provenance
- Sample Size for Test Set: Not explicitly stated as a number of patients or cases. The document mentions a "multicenter study conducted at four clinical sites." For the laboratory testing, general "laboratory testing was conducted" with no sample size specified.
- Data Provenance: The multicenter study was conducted at "four clinical sites." The country of origin is not specified, but the applicant (Visualization Technology, Inc.) is based in Boston, MA, USA, suggesting a US-based study. The data is reported as originating from prospective clinical testing (a multicenter study) and laboratory nonclinical testing.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
This type of information is not provided in the document. The device is a surgical guidance system, and device accuracy in real-time is the primary metric, not diagnostic accuracy requiring expert interpretation of results. Therefore, traditional "ground truth" establishment by experts like radiologists isn't directly applicable in the same way it would be for a diagnostic AI device.
4. Adjudication Method for the Test Set
This information is not provided. Given the nature of a surgical guidance system's accuracy measurement, an adjudication method in the context of expert consensus (like for diagnostic AI) is not directly relevant. Accuracy measurements are quantitative.
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done, and Effect Size
No, a Multi-Reader Multi-Case (MRMC) comparative effectiveness study was not done in the context of human readers improving with AI assistance. This document describes an image-guided surgery system, not an AI diagnostic tool. The comparison made is between the device's accuracy and that of a predicate device, not between human performance with and without AI.
6. If a Standalone (i.e. algorithm only without human-in-the-loop performance) Was Done
Yes, performance was evaluated in a standalone manner. The device's accuracy (both in laboratory and clinical settings) was measured directly. While the device assists a surgeon, the measurement of its accuracy itself is a standalone assessment of the device's capability to correctly determine location. The "human-in-the-loop" aspect relates to the surgeon using the device, but the accuracy figures are a direct assessment of the system's output.
7. The Type of Ground Truth Used
The ground truth used for assessing accuracy would be a precisely measured true physical location in the laboratory setting, and assumed to be the actual anatomical position in the clinical setting (relative to the CT scan data). This is a physical, objective "ground truth" rather than an interpretive one from experts or pathology.
8. The Sample Size for the Training Set
This information is not applicable/not provided. The InstaTrak device described is not an AI/machine learning device that requires a "training set" in the modern sense. It's an electromagnetic tracking system that uses CT image data. The "software feature recognition algorithm" mentioned for autoheadset registration might imply some level of learning, but no training set size or methodology is described.
9. How the Ground Truth for the Training Set Was Established
This information is not applicable/not provided for the same reasons as point 8.
{0}------------------------------------------------
K96 033 0
APR - 8 1996
14. Summary of the Safet / and Effectiveness Information
IN PREMARKET NOTIFICATION SUBMISSION
General Information
| Classification: | Class II21 CFR 892.1750 | Computed Tomography X-ray System |
|---|---|---|
| Common Name: | Interactive CT Image-Guided SurgicalSystem | |
| Device Trade Name: | InstaTrak® | |
| Intended Uses: | Anatomical Region: | Intranasal or sinuses |
| Diagnoses: | Acute and chronic sinusitis, endoscopicdacryocystorhinostomy, optic nerve andorbital decompression, the removal ofpolyps, the biopsy and removal of tumors,and the repair of CSF leaks, pituitarydisorder, and encephalocele | |
| Description: | The device consists of a wheeled cabinetenclosure with a 20-inch color monitormounted on the top. Mounted within thecabinet is a computer and a spacial trackingdevice. The electromagnetic, six-degree-of-freedom tracking device is linked to thecomputer, which provides the monitor with adisplay of the patient's CT image data andsuperimposed crosshairs, indicating theposition of the tip of the surgical instrumentused with the device. The device iscontrolled via software. | |
| Establishment Nameand Address: | Visualization Technology, Inc.656 Beacon StreetBoston, MA 02215 | |
| EstablishmentRegistration Number: | (Planned) | |
| Performance Standard: | None established under Section 514 |
{1}------------------------------------------------
Safety and Effectiveness Information Supporting the Substantial Equivalence Determination
Safety Parameter Summary
Risk Current:
Enclosure < 100 uA; Patient < 10 uA; ANSI/AAMI ES1-1993
Magnetic Field Intensity:
0.12 Gauss at 4 cm. (14 kHz)
Technological Comparison to Predicate
The InstaTrak device uses an electromagnetic sensor to determine the location of the pointing instrument being used by the surgeon. The predicate device uses a six-jointed, six-degree-of-freedom mechanical articulated arm with an electrogoniometer, which functions as a 3-D digitizer.
The InstaTrak device uses an optional registration mode called autoheadset registration. which makes use of a removable headset that attaches to the patient via the external ear canals and the bridge of the nose. Registration is the process by which the position of the patient is correlated to the CT images. Marked points on the headset that appear on the CT image can be used by a software feature recognition algorithm to automatically locate, in the CT image data, the marked points on the headset. This information, plus the predetermined location of the marked headset points with respect to an electromagnetic transmitter on the headset, allows the entire registration process to be accomplished without the user having to take any action. The predicate device uses either fiducial markers for the registration process or fiducial markers plus unmarked positions on the skin surface of the face.
These technological differences do not effect the safety or effectiveness of the device since the ability of the two devices to accurately determine location is equivalent.
Nonclinical Testing Summary
Laboratory testing was conducted to determine the device accuracy and the performance of the electromagnetic field distortion mechanism. Testing was also performed to demonstrate the reproducibility of the location positioning of the replaceable headset and the replaceable pointing instruments.
A mean device accuracy of 0.79 mm was measured which compares to a value of 1.74 mm reported for the predicate device. Results showed that the device detected field distortion under normal use conditions before the induced error became larger than 1.0 mm. The headset was shown to be replaceable such that the overall average effect upon
{2}------------------------------------------------
the device accuracy was less than 0.74 mm. The removal and replacement of a pointing instrument result in a change of less than ±0.914 mm (±3 σ).
Electromagnetic Compatibility testing was conducted and satisfactorily passed. This included emissions in accordance with EN55011(CISPR 11) and RE101 of MIL-STD-461C. and immunity in accordance with IEC 801-2, 801-3, 801-4, 801-5, and RS101 and CS114, of MIL-STD-461C. Because the recommended 10 msec dropout was satisfied, the device has a battery backup.
Clinical Testing Summary
A multicenter study was conducted at four clinical sites. Results of this study indicate the mean accuracy of the device to be 2.28 mm with a 95% confidence interval of the mean of 0.78 mm. This compares to values of 1.8 mm to 4.8 mm for the mean accuracy of the predicate device while using various operating modalities and registration techniques. Under these conditions, the 95% confidence interval of the mean for the predicate device varied from 1.1 mm to 1.6 mm.
General Safety and Effectiveness Concerns
The device labeling contains instructions for use. It includes indications for use, cautions. contraindications, warnings, and planning guidance. This information assures safe and effective use of the device.
Substantial Equivalence
The InstaTrak is an image quided surgery device that uses electromagnetic sensing technology and a removable headset to assist during endoscopic surgery through the nasal passages. Its intended use is a subset of the Viewing Wand, manufactured by ISG Technologies. Inc. The technology is similar to the Viewing Wand, which uses an articulated mechanical arm and a patient head clamp. The InstaTrak also includes an aspirating function equivalent to the Barnes Suction Tube, a Class I exempt device. The InstaTrak device is substantially equivalent to a combination of these two devices.
§ 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).