Anemia Risk Prediction Model for Osteosarcoma Patients Post-Chemotherapy Using Artificial Intelligence

Area of Science:

• Oncology
• Biostatistics
• Machine Learning

Background:

• Osteosarcoma is a primary bone malignancy often treated with chemotherapy.
• Chemotherapy can lead to various complications, including anemia, impacting patient outcomes.
• Predicting and managing post-chemotherapy anemia is crucial for effective osteosarcoma treatment.

Purpose of the Study:

• To develop and validate a machine learning model for predicting anemia in osteosarcoma patients following chemotherapy.
• To identify key clinical and laboratory risk factors associated with post-chemotherapy anemia.
• To create a tool for individualized diagnosis and treatment strategies.

Main Methods:

• Clinical data from 631 osteosarcoma patients were analyzed.
• Machine learning models including logistic regression, Random Forest (RF), Support Vector Machine (SVM), and LASSO were employed.
• A novel model was constructed by intersecting common risk factors identified by these algorithms.

Main Results:

• Twenty-five risk factors were associated with anemia post-chemotherapy (p < 0.05).
• Five common risk factors (Albumin, Calcium, Creatinine, D-dimer, ESR) were identified.
• The final predictive model demonstrated good performance with an AUC of 0.85, validated internally at 0.802.

Conclusions:

• A robust machine learning model utilizing clinical and laboratory data can predict anemia in osteosarcoma patients post-chemotherapy.
• The identified common risk factors provide insights into anemia development.
• The developed model and associated web application can support personalized patient management.