Machine Learning Predicts CD3+ Apheresis Yield
In a 2025 study, researchers applied three machine learning algorithms (random forest, logistic regression, XGBoost) to a homogeneous cohort of 98 DLBCL patients undergoing mononuclear cell apheresis to identify predictors of CD3+ yield. The logistic regression model achieved an AUC of 0.824 and identified four key predictors: CD3+ absolute count, NK cell percentage, total blood volume, and CD3+ percentage. NK percentage and CD3+ absolute count showed the strongest negative associations, informing CAR‑T manufacturing optimization.
Key Points
- 1Identify predictive features: CD3+ absolute count, NK percentage, total blood volume, CD3+ percentage.
- 2Demonstrate predictive performance: logistic regression achieves AUC 0.824, showing useful discrimination.
- 3Inform CAR‑T manufacturing: use predictors to optimize collection protocols and improve apheresis efficiency.
Scoring Rationale
Useful ML-validated predictors and AUC support practical improvements, limited by single-centre cohort size and retrospective design.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
