(84 days)
For In Vitro Diagnostic Use
Intelligent Medical Imaging, Inc.'s MICRO21 with WBC Estimate is a laboratory instrument for locating, digitally storing and displaying white blood cells to aid the technologist in performing the WBC Estimate. Examination and determination of the results must be performed by qualified individuals.
The MICRO21 with WBC Estimate is a new intended use that follows the same process as the currently approved MICRO21 with White Blood Cell (WBC) Differential (Diff). The MICRO21 with WBC Diff, Ref. No. K925670/A, is an automated microscopic system that locates WBCs, stores digital images of the cells and displays the images in an organized manner to aid technologists in performing the WBC Diff procedure. The MICRO21 process is substantially equivalent to the manual microscopic process.
The MICRO21 with WBC Estimate is an automated microscopic procedure that calculates an estimate of WBCs/uL using the information collected during the MICRO2/ with WBC Differential. Upon completion of the review process by a technologist, an algorithm calculates the WBC Estimate by using the number of classified WBCs and the number for low-power (200x magnification) fields visited. An estimated range for the WBCs/uL is calculated and reported.
Here's an analysis of the provided text regarding the acceptance criteria and study for the MICRO21 with WBC Estimate device:
Acceptance Criteria and Study Analysis for MICRO21 with WBC Estimate
The primary objective of the studies was to demonstrate the substantial equivalence of the MICRO21 with WBC Estimate to a Manual WBC Estimate and to automated cell counter results for total WBC count.
1. Table of Acceptance Criteria and Reported Device Performance
The document does not explicitly state pre-defined acceptance criteria with specific thresholds (e.g., "accuracy must be >X%"). Instead, it describes a correlation study where the MICRO21's performance is compared against established methods. The "acceptance" is implied by demonstrating substantial equivalence, meaning the results are comparable and within clinically acceptable ranges.
| Acceptance Criteria (Implied) | Reported Device Performance |
|---|---|
| Accuracy: Equivalence of MICRO21 WBC Estimate with automated cell counter Total WBC Count. | Correlated (specific metrics not provided, but the conclusion states "confirm that the MICRO21 with WBC Estimate is substantially equivalent to a Manual WBC Estimate" and "correlated the WBC Estimate on the MICRO21 with the total White Blood Cell Count generated by an automated cell counter"). |
| Accuracy: Equivalence of MICRO21 WBC Estimate with Manual WBC Estimate. | Confirmed substantially equivalent. |
| Precision: Intra-instrument precision (variability when processing the same specimen multiple times). | Results were used to provide intra-instrument and within specimen precision (specific metrics like CV% not provided, but implied to be acceptable for equivalence). |
| Reporting Range: Ability to report WBC estimates within a specified clinical range. | Reportable range from 100 to ≥ 25,000 WBCs/μL, displayed as MICRO21 WBC Estimate x,xxx - y,yyy WBC/μL. |
2. Sample Size and Data Provenance
- Test Set Sample Size: 86 blood smear samples.
- Data Provenance: Not explicitly stated, but given the nature of a 510(k) submission for commercialization, it is highly likely a prospective study using clinical samples. The document mentions "various sites" for automated cell counters, suggesting a multi-center data collection effort, but specific countries are not mentioned.
3. Number of Experts and Qualifications for Ground Truth
- Number of Experts:
- For Manual WBC Estimate (Accuracy Study): Three technologists.
- For MICRO21 WBC Estimate (Precision and Classification Review): One technologist for the precision study, and three technologists for the "WBC Classification Review" section, which contributed to confirming "real cells" used in the MICRO21 calculation.
- Qualifications of Experts: Described as "technologists." Specific experience levels (e.g., years of experience, certification) are not provided in the document. The conclusion states "Examination and determination of the results must be performed by qualified individuals," but doesn't detail the qualifications of those who performed the study.
4. Adjudication Method for the Test Set
The document does not describe a formal adjudication method (like 2+1 or 3+1).
- For the Manual WBC Estimate, it states "Three technologists performed the manual reviews," implying they each performed their own estimate, but it doesn't specify how their results were consolidated or if consensus was required.
- For the MICRO21 Classification Review, "Three technologists performed the MICRO21 reviews" to verify image classification. Again, it doesn't mention an adjudication process if there were disagreements among them.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
An MRMC study, specifically comparing human readers with AI assistance versus without AI assistance, was not explicitly done or reported in this document. The MICRO21 is described as an "automated microscopic system that locates WBCs, stores digital images of the cells and displays the images in an organized manner to aid technologists." The "aid" here is the display and organization of images, and then the automated calculation based on the technologist's reviewed classifications. The study focuses on the accuracy and precision of the device's estimate compared to manual/automated methods, not on the improvement in human performance due to the aid.
Therefore, no effect size of how much human readers improve with AI vs without AI assistance is provided.
6. Standalone (Algorithm Only) Performance
A standalone performance study was performed for the core calculation of the WBC Estimate. The algorithm automatically calculates the WBC Estimate based on:
- The number of classified WBCs (which are verified by a technologist, but the calculation itself is algorithmic).
- The number of low-power fields visited.
- A Correction Factor.
The "Test Method 1, Accuracy" and "Test Method 2, Precision" evaluate the output of this algorithmic estimate against established methods. While human input (technologist review) confirms the "Real Cells" used in the formula, the final calculation and the resulting "MICRO21 WBC Estimate" range are generated by the device's algorithm.
7. Type of Ground Truth Used
The ground truth used was a combination of:
- Established device measurements: Total White Blood Cell Count generated by automated cell counters (e.g., TOA Sysmex NE 9000, Coulter STKS, Technicon H2). These automated counters are considered a gold standard or a highly reliable reference for total WBC count.
- Expert Consensus/Reference Standard: Manual WBC Estimate performed by three technologists. This serves as a clinical reference method for comparison.
8. Sample Size for the Training Set
The document does not explicitly state the sample size for a training set. The 86 specimens mentioned are for the "Accuracy" test method, which appears to be the primary validation set.
There is a "Correction Factor" (CF) that was "defined as the mean value of the 86 specimens tested." This suggests that the 86 specimens were used to determine or fine-tune this factor, which could be considered a form of calibration or training data for that specific parameter. However, it's not a typical "training set" in the context of modern machine learning where a completely separate hold-out or test set would be used.
9. How the Ground Truth for the Training Set Was Established
As noted above, a distinct "training set" is not described. However, the Correction Factor was derived from the mean value of the 86 specimens tested. For these 86 specimens, the ground truth involved:
- Total WBC Counts from automated cell counters (controlled and calibrated according to manufacturer's specifications).
- Manual WBC Estimates performed by three technologists.
These established ground truths for the 86 specimens were then used to define the Correction Factor, making it essentially a calibrated parameter for the algorithm.
§ 864.5260 Automated cell-locating device.
(a)
Identification. An automated cell-locating device is a device used to locate blood cells on a peripheral blood smear, allowing the operator to identify and classify each cell according to type. (Peripheral blood is blood circulating in one of the body's extremities, such as the arm.)(b)
Classification. Class II (performance standards).