Multi-center validation of a machine learning model for early detection of monoclonal immunoglobulin-related disorders using routine laboratory data - Takeaways - MDSpire

Multi-center validation of a machine learning model for early detection of monoclonal immunoglobulin-related disorders using routine laboratory data

  • By

  • Yujiao Hu

  • Xiaoyan Hao

  • Xiaoyan Li

  • Hailong Tang

  • Ruixi Liu

  • Jianrui Yang

  • Weihua Zhang

  • Juan Wang

  • Xinran Liu

  • Yitong Zhou

  • Ying Zhang

  • Jun Zhang

  • Yanjun Diao

  • Lei Zhou

  • June 16, 2026

  • 0 min

Share

  • 1

    Monoclonal immunoglobulin-related disorders often go underrecognized, leading to delayed diagnosis and poor clinical outcomes.

  • 2

    Eight machine learning models were developed and compared using clinical and laboratory data to facilitate early MIg diagnosis.

  • 3

    The LightGBM model outperformed conventional methods, achieving 87.0% sensitivity and 89.6% specificity in identifying MIg-related cases.

  • 4

    Key variables for the model included urinary cast count, prothrombin time, and age, among others, enhancing its predictive capability.

  • 5

    This ML model may improve early identification of MIg disorders, potentially reducing diagnostic delays and improving patient outcomes.

Original Source(s)

Multi-center validation of a machine learning model for early detection of monoclonal immunoglobulin-related disorders using routine laboratory data

  • by Yujiao Hu, Xiaoyan Hao, Xiaoyan Li, Hailong Tang, Ruixi Liu, Jianrui Yang, Weihua Zhang, Juan Wang, Xinran Liu, Yitong Zhou, Ying Zhang, Jun Zhang, Yanjun Diao, Lei Zhou

  • June 16, 2026

Related Content

Waldenström macroglobulinemia-associated renal AL amyloidosis: a case report and literature review

Predicting CAR-T outcomes in R/R DLBCL: a multicenter real-world study of a 5-index model

Autologous stem cell transplantation for pediatric solid tumors in a resource-limited setting: a single-center experience of 15 years