Unsupervised Global CBC/RUO/CPD Phenotyping Identifies Haematological Clusters Enriched for Thrombocytopenia Severity and Mechanisms
6 affiliations
- 1School of Dentistry and Medical Sciences, Faculty of Science and Health Charles Sturt University Wagga Wagga New South Wales Australia
- 2Haematology Laboratory Liverpool Hospital Liverpool New South Wales Australia
- 3Sydney Centres for Thrombosis and Haemostasis, Institute of Clinical Pathology and Medical Research (ICPMR) and Research and Education Network (REN) Westmead Hospital Westmead New South Wales Australia
- 4AI and Cyber Futures Centre Charles Sturt University Bathurst New South Wales Australia
- 5School of Computing, Mathematics and Engineering Charles Sturt University Bathurst New South Wales Australia
- 6Department of Haematology, Liverpool Hospital, Faculty of Medicine & Health University of New South Kensington New South Wales Australia
Abstract
Introduction: Thrombocytopenia has heterogeneous mechanisms, including peripheral destruction, consumption, and reduced marrow production. This study developed an unsupervised global complete blood count/research‐use‐only/cell population data (CBC/RUO/CPD) phenotyping model and evaluated post hoc whether thrombocytopenia severity and clinically adjudicated mechanisms were enriched within the resulting clusters.
Methods: Retrospective haematology analyser data from 254 289 unique specimens were analysed after cleaning 537 279 extracted rows. Exploratory missingness filtering retained 123 features and 173 851 complete‐case specimens. The final model used a refined 41‐feature CBC_EXTENDED complete‐case dataset containing 241 591 specimens. Features were standardised, reduced using principal component analysis to 20 components explaining 92.94% of cumulative variance, and clustered using k‐means. Thrombocytopenia thresholds and electronic health record (EHR) adjudicated primary mechanism labels were evaluated post hoc.
Results: The final global k‐means model (k = 4) generated four broad haematological clusters comprising 14%–37% of specimens. Internal separation was modest but interpretable, with sampled silhouette 0.10, Davies–Bouldin 2.21, and Calinski–Harabasz 6536.92 in the fixed 60 000 specimen model comparison analysis. Cluster 3 showed progressive post hoc enrichment with thrombocytopenia severity, accounting for approximately 46%, 60%, and 71% of specimens with platelet counts < 150, < 100, and < 50 × 109/L, respectively. In 225 clinically annotated thrombocytopenic specimens, k‐means k = 4 showed limited but non‐random concordance with adjudicated mechanisms (purity 0.59; adjusted Rand index 0.14; normalised mutual information 0.18).
Conclusion: Unsupervised global CBC/RUO/CPD clustering identified broad laboratory phenotypes with post hoc enrichment for thrombocytopenia severity and adjudicated mechanisms. These findings support underutilised analyser‐derived parameters for laboratory phenotyping and hypothesis generation, not standalone diagnosis.
Related
- Unsupervised machine learning identifies two distinct posterior tibial slope clusters largely independent of Coronal Plane Alignment of the Knee in osteoarthritic knees
- Data-driven phenotyping of infective endocarditis using unsupervised machine learning: toward a personalized approach.
- Cluster-Based Phenotyping in an HFpEF Population Identifies a Biomarker-High, Atrial-Dysfunction Phenotype with Impaired Quality of Life, Reduced Exercise Capacity and Higher Healthcare Utilization
- Metabolic Phenotyping Identifies Distinct Urinary Anti-Adhesion Phenotypes and Differential Responses to Cranberry Juice Intake in Healthy Women
- Beyond Synovitis: Molecular and Neuroimmune Mechanisms Informing Target-Joint Attribution and Pain Phenotyping for Arthroplasty in Inflammatory Arthritis.
BibTeX
@article{altememi2026unsupervised,
title = {{Unsupervised Global CBC/RUO/CPD Phenotyping Identifies Haematological Clusters Enriched for Thrombocytopenia Severity and Mechanisms}},
author = {Mohammad A. Altememi and Emmanuel J. Favaloro and Md Zahidul Islam and Sara Ng and Danny Hsu and Abishek B. Santhakumar},
journal = {International Journal of Laboratory Hematology},
year = {2026},
month = aug,
doi = {10.1111/ijlh.70217},
url = {https://doi.org/10.1111/ijlh.70217},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Unsupervised Global CBC/RUO/CPD Phenotyping Identifies Haematological Clusters Enriched for Thrombocytopenia Severity and Mechanisms
AU - Mohammad A. Altememi
AU - Emmanuel J. Favaloro
AU - Md Zahidul Islam
AU - Sara Ng
AU - Danny Hsu
AU - Abishek B. Santhakumar
PY - 2026
DA - 2026/08/11
JO - International Journal of Laboratory Hematology
DO - 10.1111/ijlh.70217
UR - https://doi.org/10.1111/ijlh.70217
AB - Introduction: Thrombocytopenia has heterogeneous mechanisms, including peripheral destruction, consumption, and reduced marrow production. This study developed an unsupervised global complete blood count/research‐use‐only/cell population data (CBC/RUO/CPD) phenotyping model and evaluated post hoc whether thrombocytopenia severity and clinically adjudicated mechanisms were enriched within the resulting clusters. Methods: Retrospective haematology analyser data from 254 289 unique specimens were analysed after cleaning 537 279 extracted rows. Exploratory missingness filtering retained 123 features and 173 851 complete‐case specimens. The final model used a refined 41‐feature CBC_EXTENDED complete‐case dataset containing 241 591 specimens. Features were standardised, reduced using principal component analysis to 20 components explaining 92.94% of cumulative variance, and clustered using k‐means. Thrombocytopenia thresholds and electronic health record (EHR) adjudicated primary mechanism labels were evaluated post hoc. Results: The final global k‐means model (k = 4) generated four broad haematological clusters comprising 14%–37% of specimens. Internal separation was modest but interpretable, with sampled silhouette 0.10, Davies–Bouldin 2.21, and Calinski–Harabasz 6536.92 in the fixed 60 000 specimen model comparison analysis. Cluster 3 showed progressive post hoc enrichment with thrombocytopenia severity, accounting for approximately 46%, 60%, and 71% of specimens with platelet counts < 150, < 100, and < 50 × 109/L, respectively. In 225 clinically annotated thrombocytopenic specimens, k‐means k = 4 showed limited but non‐random concordance with adjudicated mechanisms (purity 0.59; adjusted Rand index 0.14; normalised mutual information 0.18). Conclusion: Unsupervised global CBC/RUO/CPD clustering identified broad laboratory phenotypes with post hoc enrichment for thrombocytopenia severity and adjudicated mechanisms. These findings support underutilised analyser‐derived parameters for laboratory phenotyping and hypothesis generation, not standalone diagnosis.
SN - 1751-5521
SN - 1751-553X
ER -