Search Results

The Infrascanner is indicated for the detection of traumatic supratentorial hematomas of greater than 3.5 mL in volume that are less than 2.5 cm from the brain surface, as an adjunctive device to the clinical evaluation in the acute hospital setting of patients 18 years old or greater with suspected traumatic supratentorial intracranial hematoma. The device is indicated to assess patients for CT scans but should not serve as a substitute for these scans. The Infrascanner is indicated for use by Physicians, or under the direction of a physician, who has been trained in the use of the device.

The device is a noninvasive device, which uses near-infrared spectroscopy ("NRS") to provide early information about the possible development of traumatic supratentorial intracranial hematomas in patients presenting to hospitals with head trauma. This technology involves comparing regional differences in absorbance of NIRS to hematoma evaluation is based on the principle that intracranial hemoglobin concentration will differ where a hematoma is present, compared to hemoglobin concentrations in normal intracranial regions. The system consists of a Class I NIR-based sensor is optically coupled to the patient's head through two disposable light guides in a "hairbrush" configuration. Examination with the Infrascanner is performed through placement of the sensor on designated areas of the head that represent the most common locations for traumatic hematoma. The examination is designed to be performed within two minutes.

Specifically, Model 2500 is the same device as the Infrascanner Model 2000 with following two categories of modifications:

• Scanner miniaturization
• System enhancements

The provided text is related to a 510(k) premarket notification for the InfraScanner Model 2500, a Near Infrared (NIR) Brain Hematoma Detector. However, it does not contain specific acceptance criteria tables, detailed study results (like sensitivity, specificity, AUC), or information on multi-reader multi-case (MRMC) studies, ground truth establishment methods for large datasets, or expert qualifications as typically seen in clinical validation studies for AI/CADe devices.

The document focuses on demonstrating substantial equivalence to a predicate device (InfraScanner Model 2000) based on technological characteristics and bench testing using a physical hematoma model, rather than a clinical study with human patients and expert ground truth.

Therefore, I cannot fulfill most of the request based on the provided text. The information is not present. I can only infer what kind of testing was done to support substantial equivalence.

Here's what can be extracted and what is missing:

1. Table of Acceptance Criteria and Reported Device Performance:

The document does not provide a table of precise acceptance criteria (e.g., minimum sensitivity/specificity thresholds) and specific performance metrics (like sensitivity, specificity, or AUC) from a clinical study on human patients. The "Performance Data" section discusses bench testing results, not clinical performance metrics against a defined ground truth in patients.

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

• Test Set Sample Size: Not applicable in the context of a human clinical test set. The testing described is bench testing using a physical phantom model, not a human patient test set. The "test set" here refers to configurations of the phantom.
  • Repeatability/Reproducibility: Involved "three different systems for each of the two models."
  • Agreement Test: "one system of each model."
• Data Provenance: Not applicable as it's bench testing with a phantom model, not human patient data.

3. Number of Experts Used to Establish Ground Truth and Qualifications:

• Not applicable. The ground truth for the bench testing was the physical configuration of the phantom model (e.g., presence, size, and depth of simulated hematoma). No human experts were involved in establishing "ground truth" for this type of test.

4. Adjudication Method for the Test Set:

• Not applicable. This was bench testing with a physical phantom; there was no need for expert adjudication.

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

• No. The document does not describe any MRMC study. The study presented here is a bench test comparing the new device (Model 2500) to a predicate device (Model 2000) using a physical phantom. There is no mention of human readers or AI assistance in this context.

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

• The device itself is a "Near Infrared (NIR) Brain Hematoma Detector" which provides an output. The "performance data" section describes the device's ability to "detect" simulated hematomas in a phantom. This could be considered akin to a standalone performance evaluation in a controlled environment, but it's not a clinical standalone study. The device's output is intended to be adjunctive to clinical evaluation.

7. Type of Ground Truth Used:

• Simulated Phantom Model: The ground truth for this study was established using a precisely controlled, mixed multi-layered solid and liquid optical head phantom mimicking human tissue. This phantom contained simulated hematomas of known size (5 mL, 50 mL) and depth (0 cm, 2 cm, 3 cm). The "ground truth" was the known physical state of the phantom.
  • Ovine whole blood was used to simulate hematomas.
  • The phantom was built with silicone, carbon black, and titanium dioxide to mimic optical and mechanical properties of tissue.

8. Sample Size for the Training Set:

• Not applicable. This document describes a traditional medical device (hardware with embedded firmware) and its physical performance comparison to a predicate device. There is no mention of an "AI algorithm" being trained on a dataset. The device uses near-infrared spectroscopy; it's not a machine learning model in the conventional sense requiring a training set for algorithm development described here.

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

• Not applicable, as there is no mention of an AI model with a training set.

Ask a specific question about this device