Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Apr;52(5):1739-1750.
doi: 10.1007/s00259-024-07024-x. Epub 2024 Dec 23.

Robust and interpretable deep learning system for prognostic stratification of extranodal natural killer/T-cell lymphoma

Chong Jiang #  1 , Zekun Jiang #  2   3 , Xinyu Zhang  4 , Linhao Qu  5 , Kexue Fu  5 , Yue Teng  6 , Ruihe Lai  6 , Rui Guo  7 , Chongyang Ding  8 , Kang Li  9   10 , Rong Tian  11
Affiliations

Robust and interpretable deep learning system for prognostic stratification of extranodal natural killer/T-cell lymphoma

Chong Jiang et al. Eur J Nucl Med Mol Imaging. 2025 Apr.

Abstract

Purpose: Extranodal natural killer/T-cell lymphoma (ENKTCL) is an hematologic malignancy with prognostic heterogeneity. We aimed to develop and validate DeepENKTCL, an interpretable deep learning prediction system for prognosis risk stratification in ENKTCL.

Methods: A total of 562 patients from four centers were divided into the training cohort, validation cohort and test cohort. DeepENKTCL combined a tumor segmentation model, a PET/CT fusion model, and prognostic prediction models. RadScore and TopoScore were constructed using radiomics and topological features derived from fused images, with SHapley Additive exPlanations (SHAP) analysis enhancing interpretability. The final prognostic models, termed FusionScore, were developed for predicting progression-free survival (PFS) and overall survival (OS). Performance was assessed using area under the receiver operator characteristic curve (AUC), time-dependent C-index, clinical decision curves (DCA), and Kaplan-Meier (KM) curves.

Results: The tumor segmentation model accurately delineated the tumor lesions. RadScore (AUC: 0.908 for PFS, 0.922 for OS in validation; 0.822 for PFS, 0.867 for OS in test) and TopoScore (AUC: 0.756 for PFS, 0.805 for OS in validation; 0.689 for PFS, 0.769 for OS in test) both exhibited potential prognostic capability. The time-dependent C-index (0.897 for PFS, 0.928 for OS in validation; 0.894 for PFS, 0.868 for OS in test) and DCA indicated that FusionScore offers significant prognostic performance and superior net clinical benefits compared to existing models. KM survival analysis showed that higher FusionScores correlated with poorer PFS and OS across all cohorts.

Conclusion: DeepENKTCL provided a robust and interpretable framework for ENKTCL prognosis, with the potential to improve patient outcomes and guide personalized treatment.

Keywords: Deep learning; Extranodal NK/T-cell lymphoma; Machine learning; PET/CT; Radiomics; Survival analysis; Topological data analysis.

PubMed Disclaimer

Conflict of interest statement

Declarations. Ethical approval: All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This study was approved by the institutional review board of West China Hospital, Sichuan University. Consent to participate: For the nature of retrospective study, written informed consent was waived. Conflict of interest: The authors declare that they have no conflict of interest.

Similar articles

See all similar articles

Cited by

References

    1. Costa RO, Pereira J, Lage LAPC, et al. Extranodal NK-/T-cell lymphoma, nasal type: what advances have been made in the last decade? Front Oncol. 2023;13:1175545. - PubMed - PMC
    1. Haverkos BM, Pan Z, Gru AA, et al. Extranodal NK/T Cell Lymphoma, nasal type (ENKTL-NT): an update on Epidemiology, Clinical Presentation, and natural history in north American and European cases. Curr Hematol Malig Rep. 2016;11(6):514–27. - PubMed - PMC
    1. Yan Z, Yao S, Wang Z, et al. Treatment of extranodal NK/T-cell lymphoma: from past to future. Front Immunol. 2023;14:1088685. - PubMed - PMC
    1. Jiang C, Zhang X, Jiang M, et al. Assessment of the prognostic capacity of pretreatment, interim, and post-therapy (18)F-FDG PET/CT in extranodal natural killer/T-cell lymphoma, nasal type. Ann Nucl Med. 2015;29(5):442–51. - PubMed
    1. Son SH, Kang SM, Jeong SY, et al. Prognostic Value of Volumetric parameters measured by pretreatment 18F FDG PET/CT in patients with cutaneous malignant melanoma. Clin Nucl Med. 2016;41(6):e266–73. - PubMed

LinkOut - more resources