Artificial intelligence reveals the predictions of hematological indexes in children with acute leukemia

Zhangkai J Cheng1, Haiyang Li2,3, Mingtao Liu1

  • 1Department of Clinical Laboratory, National Center for Respiratory Medicine, National Clinical Research Center for Respiratory Disease, State Key Laboratory of Respiratory Disease, Guangzhou Institute of Respiratory Health, The First Affiliated Hospital of Guangzhou Medical University, Guangzhou, 510120, Guangdong, China.

BMC Cancer
|August 12, 2024
PubMed

Insights

This study developed an AI tool to predict childhood leukemia using blood biomarkers. The model accurately identifies acute lymphoblastic leukemia (ALL) and acute myeloid leukemia (AML), aiding early cancer detection in children.

Area of Science:

  • Pediatric Oncology
  • Hematology
  • Biochemistry

Background:

  • Childhood leukemia, including acute lymphoblastic leukemia (ALL) and acute myeloid leukemia (AML), is a significant pediatric cancer.
  • Early detection and treatment are crucial for improving survival rates in pediatric leukemia patients.

Purpose of the Study:

  • To develop an early and comprehensive predictor for childhood hematologic malignancies.
  • To analyze nutritional biomarkers, leukemia indicators, and granulocytes for predictive modeling.

Main Methods:

  • A machine learning algorithm, specifically a random forest model, was utilized.
  • Blood samples from 826 children with ALL, 255 with AML, and 200 healthy children were analyzed.
  • Ten blood indices were evaluated for predictive accuracy.

Main Results:

  • Significant differences in biochemical indicators were observed between leukemia patients and healthy children.
  • The random forest model achieved an Area Under the Curve (AUC) of 0.950 for predicting leukemia subtypes and 0.909 for AML.
  • Gender-based differences in certain indicators were noted, with boys showing higher levels.

Conclusions:

  • The study presents an efficient AI-driven diagnostic tool for the early screening of childhood blood cancers.
  • Artificial intelligence holds significant potential for enhancing modern pediatric healthcare diagnostics.
  • Biomarker analysis combined with machine learning offers a promising approach for early leukemia detection.