Structure of the Discretised Complete Blood Count Profile Space: A Reproducible Analysis of 700,000 Records

Background/Objectives: Qualitative complete blood count (CBC) profiles, formed by classifying each result relative to its reference interval, underlie rule-based interpretation and autoverification. The structure of the resulting profile space has not been characterised. Methods: We analysed 729,469 adult CBC records from 414,907 patients (14 analytes; Sysmex XN-10; January 2024–May 2025). The results were coded into five ordinal states relative to each record’s reference interval. We measured concentration, diversity, algebraic dependence and unobserved richness under three reference interval frameworks and at the patient level. Results: Under the laboratory’s limits, the records formed 40,376 distinct profiles. Seven profiles covered half of the records; the all-normal profile covered 42.2%. Removing four algebraically dependent analytes contracted the observed space by 63.0%; their discretised states were not fully determined by the other analytes. New profiles kept appearing, so the tail of the distribution remained open: with the framework held constant, 3.03% of the 2025 records carried a profile that was unseen in 2024 (predicted, 3.06%). Across the frameworks, the number of profiles varied 1.65-fold, but the concentrated head (a few profiles covering most records), the open tail and the contraction (52.3–63.0%) persisted. With one record per patient, the concentration was higher and the tail remained open. Conclusions: In this corpus, the reference framework changed the coordinates of the profile space (the number of profiles and the all-normal share) but not its structural properties. These properties suggest hypotheses for the design of rule-based interpretation systems.

Authors

Institutions

Publication Details

Journal
Diagnostics
Published
2026-11-18
DOI
https://doi.org/10.3390/diagnostics16223860
Primary Topic
Clinical Laboratory Practices and Quality Control
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Structure of the Discretised Complete Blood Count Profile Space: A Reproducible Analysis of 700,000 Records

Paweł Wąsowicz, Tomasz Anyszek, Jakub Swadźba
1 links
Diagnostics
Clinical Laboratory Practices and Quality Control
article

Structure of the Discretised Complete Blood Count Profile Space: A Reproducible Analysis of 700,000 Records

Paweł Wąsowicz, Tomasz Anyszek, Jakub Swadźba
article en

Abstract

Background/Objectives: Qualitative complete blood count (CBC) profiles, formed by classifying each result relative to its reference interval, underlie rule-based interpretation and autoverification. The structure of the resulting profile space has not been characterised. Methods: We analysed 729,469 adult CBC records from 414,907 patients (14 analytes; Sysmex XN-10; January 2024–May 2025). The results were coded into five ordinal states relative to each record’s reference interval. We measured concentration, diversity, algebraic dependence and unobserved richness under three reference interval frameworks and at the patient level. Results: Under the laboratory’s limits, the records formed 40,376 distinct profiles. Seven profiles covered half of the records; the all-normal profile covered 42.2%. Removing four algebraically dependent analytes contracted the observed space by 63.0%; their discretised states were not fully determined by the other analytes. New profiles kept appearing, so the tail of the distribution remained open: with the framework held constant, 3.03% of the 2025 records carried a profile that was unseen in 2024 (predicted, 3.06%). Across the frameworks, the number of profiles varied 1.65-fold, but the concentrated head (a few profiles covering most records), the open tail and the contraction (52.3–63.0%) persisted. With one record per patient, the concentration was higher and the tail remained open. Conclusions: In this corpus, the reference framework changed the coordinates of the profile space (the number of profiles and the all-normal share) but not its structural properties. These properties suggest hypotheses for the design of rule-based interpretation systems.

DiagnosticsVol. 16(22)
Andrzej Frycz Modrzewski Krakow University (PL)
Good health and well-being
Clinical Laboratory Practices and Quality Control
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon