Related Experiment Video
Updated: Dec 17, 2025
Flow Cytometric Characterization of Murine B Cell Development
Published on: January 22, 2021
A Clinically Applicable Approach to the Classification of B-Cell Non-Hodgkin Lymphomas with Flow Cytometry and
Valentina Gaidano1,2, Valerio Tenace3, Nathalie Santoro4
1Department of Clinical and Biological Sciences, University of Turin, 10043 Orbassano, Italy.
Insights
Machine learning models accurately classify B-cell Non-Hodgkin Lymphomas (B-NHL) using flow cytometry (FC) immunophenotype data. This approach enhances diagnostic capabilities for B-NHL, improving classification accuracy and identifying key predictive markers.
Area of Science:
- Hematology
- Computational Biology
- Immunology
Background:
- Immunophenotyping is crucial for classifying B-cell Non-Hodgkin Lymphomas (B-NHL).
- Flow cytometry (FC) offers advantages over immunohistochemistry but lacks extensive validation and practical experience.
- Current FC use in B-NHL is often limited to specific involvement sites.
Purpose of the Study:
- To develop and validate machine learning models for classifying B-NHL using FC immunophenotype data.
- To assess the performance of predictive systems in distinguishing common B-NHL subtypes.
- To identify key markers and expression levels critical for accurate B-NHL classification.
Main Methods:
- Application of machine learning algorithms to a database of 1465 B-NHL samples.
- Development of four artificial predictive systems for B-NHL classification into nine clinico-pathological entities.
- Evaluation of model performance using metrics such as overall accuracy, mean sensitivity, and mean specificity.
Main Results:
- The best predictive model achieved an overall accuracy of 92.68%, mean sensitivity of 88.54%, and mean specificity of 98.77%.
- MIB1 and Bcl2 demonstrated strong discriminatory power, significantly enhancing model performance when integrated.
- The study highlighted the utility of non-canonical markers and suggested a nuanced interpretation of FC marker expression levels.
Conclusions:
- Machine learning models can effectively classify B-NHL using FC immunophenotype data, offering a robust diagnostic tool.
- Specific markers like MIB1 and Bcl2 are highly valuable for B-NHL classification, alongside potentially useful non-canonical markers.
- FC marker interpretation should consider expression levels rather than strict positive/negative thresholds for improved B-NHL subtyping.
Abstract:
The immunophenotype is a key element to classify B-cell Non-Hodgkin Lymphomas (B-NHL); while it is routinely obtained through immunohistochemistry, the use of flow cytometry (FC) could bear several advantages. However, few FC laboratories can rely on a long-standing practical experience, and the literature in support is still limited; as a result, the use of FC is generally restricted to the analysis of lymphomas with bone marrow or peripheral blood involvement. In this work, we applied machine learning to our database of 1465 B-NHL samples from different sources, building four artificial predictive systems which could classify B-NHL in up to nine of the most common clinico-pathological entities. Our best model shows an overall accuracy of 92.68%, a mean sensitivity of 88.54% and a mean specificity of 98.77%. Beyond the clinical applicability, our models demonstrate (i) the strong discriminatory power of MIB1 and Bcl2, whose integration in the predictive model significantly increased the performance of the algorithm; (ii) the potential usefulness of some non-canonical markers in categorizing B-NHL; and (iii) that FC markers should not be described as strictly positive or negative according to fixed thresholds, but they rather correlate with different B-NHL depending on their level of expression.
