Research source
AI predicted ICU mortality risk well in blood cancer patients, especially for multiple myeloma.
Based on Machine learning-based ICU mortality prediction across hematologic malignancy subtypes: A comparative analysis using MIMIC-IV.
Published by Journal of Clinical OncologyJun 23, 2026
Peer-reviewed paperMedium evidenceHuman study
What researchers observed
Study developed AI models predicting ICU death risk in blood cancers, showing good accuracy, especially for multiple myeloma.
What this could mean
This could help researchers better measure and understand mortality risk differences among blood cancer types in ICU, if validated further. It may point toward improved subgroup models for patient risk assessment.
What researchers still need to learn
Researchers still need to learn whether these predictions improve patient outcomes or clinical decisions in ICU care.