Machine learning optimized multiparameter radar plots for <scp>B‐cell acute lymphoblastic leukemia</scp> minimal residual disease analysis
Bibliographic record
Abstract
BACKGROUND: Flow cytometry is widely used for B-ALL minimal residual disease (MRD) analysis given its speed, availability, and sensitivity; however, distinguishing B-lymphoblasts from regenerative B-cells is not always straightforward. Radar plots, which project multiple markers onto a single plot, have been applied to other MRD analyses. Here we aimed to develop optimized radar plots for B-ALL MRD analysis. METHODS: We compiled Children's Oncology Group (COG) flow data from 20 MRD-positive and 9 MRD-negative B-ALL cases (enriched for hematogones) to create labeled training and test data sets with equal numbers of B-lymphoblasts, hematogones, and mature B-cells. We used an automated approach to create hundreds of radar plots and ranked them based on the ability of support vector machine (SVM) models to separate blasts from normal B-cells in the training data set. Top-performing radar plots were compared with PCA, t-SNE, and UMAP plots, evaluated with the test data set, and integrated into clinical workflows. RESULTS: SVM area under the ROC curve (AUC) for COG tube 1/2 radar plots improved from 0.949/0.921 to 0.989/0.968 after optimization. Performance was superior to PCA plots and comparable to UMAP, but with better generalizability to new data. When integrated into an MRD workflow, optimized radar plots distinguished B-lymphoblasts from other CD19-positive populations. MRD quantified by radar plots and serial gating were strongly correlated. DISCUSSION: Radar plots were successfully optimized to discriminate between diverse B-lymphoblast populations and non-malignant CD19-positive populations in B-ALL MRD analysis. Our novel radar plot optimization strategy could be adapted to other MRD panels and clinical scenarios.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.007 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".