Automated bone marrow cell classification using ensemble learning: performance, generalization, and clinical interpretability - Takeaways - MDSpire

Automated bone marrow cell classification using ensemble learning: performance, generalization, and clinical interpretability

  • By

  • Shahid Mehmood

  • Muhammad Zubair

  • Sagheer Abbas

  • Rowida Mohammed Alharbi

  • Mai Alduailij

  • Muhammad Adnan Khan

  • Taher M. Ghazal

  • June 3, 2026

  • 0 min

Share

  • 1

    Automated classification of bone marrow cells is crucial for diagnosing hematological disorders but faces challenges like morphological overlap and class imbalance.

  • 2

    An ensemble-learning framework combining MobileNetV3 and ResNet18 was developed to enhance classification performance while maintaining low computational costs.

  • 3

    Boosting achieved the highest classification accuracy of 96%, with external validation confirming robust performance across various imaging conditions.

  • 4

    The ensemble model outperformed individual models in accuracy and interpretability, focusing on relevant morphological features for clinical plausibility.

  • 5

    The proposed framework supports AI-assisted hematological diagnostics, potentially reducing diagnostic time and interobserver variability.

Original Source(s)

Automated bone marrow cell classification using ensemble learning: performance, generalization, and clinical interpretability

  • by Shahid Mehmood, Muhammad Zubair, Sagheer Abbas, Rowida Mohammed Alharbi, Mai Alduailij, Muhammad Adnan Khan, Taher M. Ghazal

  • June 3, 2026

Related Content

Successful conversion therapy for presumed granulocyte colony-stimulating factor–producing hepatocellular carcinoma: a case report

Multimodal feature fusion model for breast mass malignant risk stratification

Preliminary construction and validation of a prognostic prediction model for cervical cancer based on tumor mechanics-related genes