Canonical text (the exact bytes fingerprinted)

TRUVACE RECORD VERSION
record: TRV-2026-1411
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-10-15T06:54:26.233797Z
status: published
lens: trace
sector: health
headline: Performance Evaluation of an Artificial Intelligence-Assisted Digital Morphology System for Bone Marrow Aspirate Examination
dek: Introduction: Morphological examination of bone marrow aspirates (BMA) remains essential for the diagnosis of hematologic diseases, but conventional microscopy is labor-intensive and subject to interobserver variability.
Methods: A total of 158 samples from patients were analyzed. Differential counts obtained by optical microscopy were compared with those of the CellaVision BMA Application used according to its intended workflow of automated preclassification followed by expert review. The two methods were compa…
gain_title: In 158 patient samples, automated preclassification plus expert review gave differential counts comparable with conventional microscopy for most nucleated cells, with 96.8% to 99.4% agreement at key decision thresholds.
problem_title: Agreement was weaker for basophils and plasma cells, and sensitivity at the 20% blast threshold was only 77.3%, so expert review remains necessary.
trace_subject: CellaVision BMA Application for bone marrow aspirate differential counts versus optical microscopy
gain_reading: In 158 patient samples, automated preclassification plus expert review gave differential counts comparable with conventional microscopy for most nucleated cells, with 96.8% to 99.4% agreement at key decision thresholds.
gain_evidence: produced differential counts comparable with conventional microscopy for most nucleated cells
problem_reading: Agreement was weaker for basophils and plasma cells, and sensitivity at the 20% blast threshold was only 77.3%, so expert review remains necessary.
problem_evidence: sensitivity at the 20% blast threshold was 77.3% | agreement was weaker for basophils and plasma cells
quick_read: Researchers compared differential cell counts from optical microscopy with those from the CellaVision BMA Application on 158 patient bone marrow aspirate samples. The software was used as intended, preclassifying cells before an expert reviewed them.

Counts correlated strongly for most cell types, which suggests this kind of tool could ease a labor-intensive diagnostic task. But weaker results for some cell types and a lower sensitivity at the 20% blast threshold show why human review is still needed, and a single-center series cannot settle wider performance.
limitation: The findings come from a single-center series and the system was assessed only within an expert-reviewed workflow, so expert review and further validation remain necessary.
tag: Dual reading
key_points: Pearson correlations ranged from 0.85 to 0.99 for all populations except basophils, which reached 0.68. | Passing-Bablok regression found no significant systematic or proportional bias for erythroblasts, granulocytic cells, eosinophils and monocytes. | Bland-Altman mean differences ranged from -1.4% to 1.2% across all populations. | Plasma cells and basophils showed significant deviations from linearity.
rundown: Bone marrow aspirate morphology remains essential for diagnosing hematologic diseases, yet conventional microscopy is labor-intensive and subject to interobserver variability, which motivated the evaluation of automated preclassification.

Methods included Pearson correlation, Passing-Bablok regression and Bland-Altman analysis, applied overall and within the blast and plasma cell intervals that bracket diagnostic thresholds. A small systematic difference appeared for blasts and a small proportional difference for lymphocytes.

At the clinically relevant blast and plasma cell thresholds, overall agreement between the two methods was high, although the 77.3% sensitivity at the 20% blast threshold marks a gap. The authors describe the results as a basis for further evaluation of this and similar systems.
sources:
- peer_reviewed | International Journal of Laboratory Hematology | https://doi.org/10.1111/ijlh.70263 | 2026-10-14
prev: 0000000000000000000000000000000000000000000000000000000000000000

sha256

9822a5e93ce803c1cbecebe29f4c0f49ca4895168e1bddb86155f05aea886ccf

previous

0000000000000000000000000000000000000000000000000000000000000000