(84 days)
The CELL-DYN® 3200 System is a fully automated hematology analyzer intended for in-vitro diagnostic use in the clinical laboratory.
The CELL-DYN® 3200 System has three main modules: the Analyzer, which identifies, mixes, and presents specimens for processing, aspirates, dilutes and analyzes each whole blood specimen. The Data Module, which controls all system processing; and the Display Station, which provides the primary operator interface with the system and generates reports automatically or on demand.
Here is an analysis of the provided information, structured to address your specific questions:
The provided documents do not contain a table of acceptance criteria or reported device performance in a format that directly maps to typical AI/software performance metrics (e.g., sensitivity, specificity, AUC). Instead, the studies described are focused on demonstrating "substantial equivalence" of the CELL-DYN® 3200 System to existing predicate devices for various hematological parameters. This typically involves accuracy, precision, and linearity data compared against manufacturer's specifications and predicate device performance.
Given the nature of the device (a hematology analyzer) and the era of the submission (1997), the evaluation methodology is characteristic of medical devices that provide quantitative measurements rather than AI-driven diagnostic interpretations. Therefore, many of your questions regarding AI-specific assessments (e.g., MRMC studies, human-in-the-loop, training set details) are not directly applicable or answerable from the provided text.
Here's what can be extracted and inferred:
1. Table of Acceptance Criteria and Reported Device Performance
As mentioned, a direct table of acceptance criteria and performance metrics (like sensitivity/specificity) is not present. Instead, the "acceptance criteria" are implied by the requirement to demonstrate substantial equivalence to predicate devices and adherence to "manufacturer's specifications" for accuracy, precision, and linearity for each parameter.
The following table summarizes the parameters for which substantial equivalence was demonstrated and the predicate devices used for comparison, which implicitly define the performance standard (acceptance criteria).
| Hematological Parameter | Predicate Device(s) for Substantial Equivalence | Reported Device Performance |
|---|---|---|
| White Blood Cell (WBC) | Abbott CELL-DYN® 3500R System, Technicon H 3 RTC | Data supported substantial equivalence for accuracy, precision, and linearity. |
| Red Blood Cell (RBC) | Abbott CELL-DYN® 3500R System, Technicon H 3 RTC | Data supported substantial equivalence for accuracy, precision, and linearity. |
| Hemoglobin (HGB) | Abbott CELL-DYN® 3500R System, Technicon H 3 RTC | Data supported substantial equivalence for accuracy, precision, and linearity. |
| Platelet (PLT) | Abbott CELL-DYN® 3500R System, Technicon H 3 RTC | Data supported substantial equivalence for accuracy, precision, and linearity. |
| WBC Differential | Abbott CELL-DYN® 3500R System, Technicon H 3 RTC | Data supported substantial equivalence for accuracy, precision, and linearity. |
| Mean Corpuscular Hemoglobin (MCH) | Abbott CELL-DYN® 3500R System, Technicon H 3 RTC | Data supported substantial equivalence for accuracy, precision, and linearity. |
| Mean Corpuscular Volume (MCV) | CELL-DYN® 4000 System, manual microhematocrit method (NCCLS H7-A2), Technicon H 3 RTC | Data supported substantial equivalence for accuracy, precision, and linearity. |
| Red Cell Distribution Width (RDW) | CELL-DYN® 4000 System, Technicon H 3 RTC | Data supported substantial equivalence for accuracy, precision, and linearity. |
| Mean Platelet Volume (MPV) | CELL-DYN® 4000 System, Technicon H 3 RTC | Data supported substantial equivalence for accuracy, precision, and linearity. |
| All Hemogram parameters | Technicon H 3 RTC | Data supported substantial equivalence for accuracy, precision, and linearity. |
2. Sample Size Used for the Test Set and Data Provenance
- Sample Size: The specific sample size used for the "test set" (i.e., the samples used in the equivalency studies) is not explicitly stated in the provided documents.
- Data Provenance: The documents do not specify the country of origin of the data or whether the studies were retrospective or prospective.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications
- Number of Experts: For WBC differential parameters, the NCCLS H20-A, "Reference Leukocyte Differential Count (Proportional) and Evaluation of Instrumental Methods," Approved Voluntary Standard 1992 was used "to arbitrate differences in the differential." This standard outlines the methodology for establishing a reference leukocyte differential count, which typically involves multiple trained technologists reviewing blood smears. The exact number of experts/technologists used in this specific study and their qualifications are not explicitly stated in the provided text.
- Qualifications of Experts: Not specified beyond the reference to the NCCLS H20-A standard.
4. Adjudication Method
For the white cell differential parameters, the NCCLS H20-A standard was used for arbitration. This standard details a process for establishing a reference differential count, which usually involves a consensus or weighted average from multiple expert reviews of stained blood smears (e.g., 500-cell or 200-cell manual counts). The specific "adjudication method" (e.g., 2+1, 3+1) within that standard, as applied to this study, is not explicitly detailed in the provided text. For other parameters, comparison was made against predicate devices or a manual method (microhematocrit for MCV), implying the predicate device's or manual method's result served as the "ground truth" or reference for comparison, rather than an independent expert adjudication.
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done, and Effect Size
- No, an MRMC comparative effectiveness study was not done. This type of study is typically associated with evaluating reader performance (e.g., radiologists, pathologists) with and without AI assistance, which is not applicable to a hematology analyzer that provides quantitative measurements. The studies focused on quantitative agreement with predicate devices and reference methods.
6. If a Standalone (Algorithm Only Without Human-in-the-Loop Performance) Was Done
- Yes, the studies are effectively standalone performance evaluations. The CELL-DYN® 3200 System is an automated analyzer designed to provide results directly. Its performance was evaluated by comparing its outputs (WBC, RBC, HGB, etc.) against established methods (predicate devices, manual microhematocrit, NCCLS standard for differentials). There is no "human-in-the-loop" component described in this context for generating the primary analytical results. The "human-in-the-loop" would be the clinical laboratory professional operating the device and interpreting its output, but the performance testing here is of the device's measurement capabilities themselves.
7. The Type of Ground Truth Used
The ground truth used varied by parameter:
- WBC, RBC, HGB, PLT, MCH, RDW, MPV, all hemogram parameters: The results from the predicate devices (Abbott CELL-DYN® 3500R System, CELL-DYN® 4000 System, Technicon H 3 RTC) served as the primary reference or "ground truth" for demonstrating substantial equivalence. This is a form of comparative measurement against an accepted standard.
- WBC Differential parameters: The NCCLS H20-A "Reference Leukocyte Differential Count" standard was used for arbitration, which implies an expert consensus or reference method defined by the standard.
- MCV: The manual microhematocrit method (NCCLS H7-A2) was used as a reference for comparison, in addition to the CELL-DYN® 4000 System. This is a reference standard method.
8. The Sample Size for the Training Set
- Not Applicable / Not Provided. Hematology analyzers of this type are not typically "trained" in the machine learning sense with a distinct training set. They are designed based on physical principles (flow cytometry, laser optical scatter, colorimetry) and algorithms are developed to process these signals. Any "training" or calibration would be part of the device's engineering and internal validation, not a separate, disclosed "training set" like in modern AI models. The 510(k) summary focuses on the performance of the final, developed device.
9. How the Ground Truth for the Training Set Was Established
- Not Applicable / Not Provided. As noted above, the concept of a "training set" with established ground truth as used in contemporary AI/ML is not relevant to this type of device and submission. The "ground truth" in development would likely refer to highly characterized reference materials and established analytical methods used during the engineering and validation phases of the device.
§ 864.5220 Automated differential cell counter.
(a)
Identification. An automated differential cell counter is a device used to identify one or more of the formed elements of the blood. The device may also have the capability to flag, count, or classify immature or abnormal hematopoietic cells of the blood, bone marrow, or other body fluids. These devices may combine an electronic particle counting method, optical method, or a flow cytometric method utilizing monoclonal CD (cluster designation) markers. The device includes accessory CD markers.(b)
Classification. Class II (special controls). The special control for this device is the FDA document entitled “Class II Special Controls Guidance Document: Premarket Notifications for Automated Differential Cell Counters for Immature or Abnormal Blood Cells; Final Guidance for Industry and FDA.”