Search Results

The X100 with the Full Field Bone Marrow Aspirate (BMA) Application is an automated cell locating device, intended for in vitro use only. The Full Field BMA application automatically locates and presents images of hematopoietic cells to trained operators for visual evaluation of Romanowsky stained bone marrow aspirate (BMA) smears. The Full Field BMA application assists trained operators to perform bone marrow smear quality assessment, blast cell, plasma cell, and M:E ratio estimation. A qualified operator must review, confirm or modify classification of each cell according to type, and verify results before finalizing and releasing the report.

The X100 with the Full Field Bone Marrow Aspirate (BMA) application presents images of Prussian Blue stained BMA smear.

X100HT with Full Field BMA Application:

The X100HT with the Full Field Bone Marrow Aspirate (Full Field BMA) Application is an automated cell locating device. intended for in vitro use only. The Full Field BMA application automatically locates and presents images of hematopoietic cells to trained operators for visual evaluation of Romanowsky stained bone marrow aspirate (BMA) smears. The Full Field BMA application assists trained operators to perform bone marrow smear quality assessment, blast cell, plasma cell, and M:E ratio estimation. A qualified operator must review, confirm or modify classification of each cell according to type, and verify results before finalizing and releasing the report.

The X100HT with the Full Field Bone Marrow Aspirate (BMA) application presents images of Prussian Blue stained BMA smear.

The X100 and X100HT instruments include a digital scanner and processing unit (computer) capable of acquiring high-resolution digital images from Romanowsky or Prussian Blue stained BMA slides with software application (Full Field Bone Marrow Aspirate (BMA) Application) for evaluation of acquired images. The X100 allows for three slides while X100HT comes with an additional component, a slide loader, a mechanical loading mechanism that holds up to 30 slides in three 10-slide cassettes and performs cover slipping and slide loading into the X100 scanner. The Full Field BMA application contains two scan modes - Romanowsky stain and Prussian Blue stain.

The X100 and X100HT devices with the Full Field Bone Marrow Aspirate (BMA) Application are automated cell-locating devices intended to assist trained operators in evaluating Romanowsky or Prussian Blue stained bone marrow aspirate smears. The application specifically aids in bone marrow smear quality assessment, blast cell estimation, plasma cell estimation, and Myeloid:Erythroid (M:E) ratio estimation. A qualified operator must review, confirm, or modify classifications and verify results before finalizing reports.

1. Table of Acceptance Criteria & Reported Device Performance

The acceptance criteria for the device performance are indicated by the successful completion of the studies and the reported sensitivities, specificities, and agreements, particularly for the key parameters of specimen quality, blast cells, plasma cells, and M:E ratio. The tables show the performance results from the clinical study comparing the device to manual microscopy and to de-identified medical data.

Clinical Study: Comparison of Candidate Device to Manual Microscopy

Clinical Study: Comparison of Candidate Device to De-identified Medical Data

2. Sample Sizes and Data Provenance

• Test Set Sample Size:

  • Clinical Study (Romanowsky Stain): 615 Romanowsky stained slides were used for the clinical study comparing the candidate device to manual microscopy.
  • Clinical Study (Prussian Blue Stain): 180 Prussian blue stained slides were used for the assessment of manually classified parameters. (Note: These are explicitly stated as separate from the 615 Romanowsky slides for DSS evaluation, but are part of the overall clinical study).
  • Reproducibility Study: 9 Romanowsky stained slides (3 per clinical group: R, M, N)
  • Repeatability Study: 12 Romanowsky stained slides (4 per clinical group: R, M, N)

• Data Provenance: The document does not explicitly state the country of origin of the data for the clinical studies. The samples used for reference range validation were "banked normal morphology BMA smears taken from U.S. individuals." The clinical study was conducted at three undisclosed sites. The studies were prospective in design, involving the evaluation of slides by the test method and a reference method with a washout period.

3. Number of Experts and Qualifications for Ground Truth

• Clinical Study (Test Set):
  • Manual Microscopy (Reference Method): The slides were evaluated by "two operators at each site." These operators were "qualified to review bone marrow aspirates." The document implies that their consensus or individual readings constituted the ground truth for comparison.
  • Reference Range Validation: Two investigators (from a total of four participating in the study) reviewed each sample, and the "means of the two differential counts were used in the validation of the results."
• Qualifications: "Qualified to review bone marrow aspirates." This suggests expertise in hematology/pathology, likely pathologists or laboratory scientists with extensive experience in bone marrow morphology, but specific certifications or years of experience are not provided.

4. Adjudication Method for the Test Set

• Clinical Study: The document states that the slides were "evaluated by two operators at each site." For the DSS-suggested results, operators were able to "review and confirm or reclassify" the device's classifications. For the manually classified parameters, the "images were reviewed and cell types were classified by the operators." There is no explicit mention of a formal adjudication process (e.g., 2+1, 3+1 consensus) if the initial two operators disagreed; it implies their combined evaluation or individual readings were used, potentially with an internal lab protocol for resolving discrepancies. The phrase "means of the two differential counts" for reference range validation suggests average.

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

• A formal MRMC comparative effectiveness study, directly quantifying the improvement of human readers with AI assistance versus without AI assistance, is not explicitly detailed in the provided text.
• The study design focused on comparing the device with operator review and modification to manual microscopy (the current standard). This implicitly involves human readers using the AI (DSS suggested results). The benefit is stated as "facilitating more effective BMA assessment and reducing potential user errors such as underestimating blast counts." However, a direct quantifiable measure of reader performance improvement (e.g., increased accuracy or reduced reading time for human readers) through AI assistance compared to unassisted reading is not provided in a comparative effectiveness manner from the reported studies.

6. Standalone (Algorithm Only) Performance

• The reported performance for the core parameters (Specimen quality, blast cells, plasma cells, and M:E ratio) from the repeatability and reproducibility studies were based on the "device DSS pre-classified results without operator reclassification." This indicates that a standalone performance assessment of the algorithm's initial suggestions was conducted for these analytical parameters.
• The clinical studies, however, state that the "DSS suggested results for the cell types/parameters were reviewed and confirmed or reclassified by the operators as necessary," meaning the reported clinical performance includes the human-in-the-loop component, not purely standalone algorithm performance.

7. Type of Ground Truth Used

• Clinical Study (Performance against manual microscopy): Expert Consensus/Manual Gold Standard. The ground truth was established by "manual microscopy," performed by "qualified operators" (two per site) reviewing each slide. This is essentially the current clinical gold standard for bone marrow evaluation.
• Clinical Study (Performance against de-identified medical data): Outcomes/Clinical Records Data. The comparison was made against "de-identified medical records data" retrieved from two of the three sites, representing real-world clinical findings.
• Precision Studies: No explicit separate ground truth was established; the device's own DSS pre-classified results were used to assess the consistency and agreement of its measurements.
• Reference Range Validation: "Published reference ranges found in literature" constituted the external ground truth, validated against BMA smears from U.S. individuals deemed "non-abnormal during routine lab evaluation."

8. Sample Size for the Training Set

• The document does not provide any information regarding the sample size or characteristics of the training set used to develop the device's algorithms. The focus is solely on the validation/test datasets.

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

• Since the document does not provide details on the training set, there is no information on how its ground truth was established.

Ask a specific question about this device