Multi-modal Fusion Model and Deep Learning for Predicting Treatment Response in NKTCL

Multi-modal Fusion Model and Deep Learning for Predicting Treatment Response in NK/T-Cell Lymphoma

Status

Not yet recruiting

Phases

Unknown

Study type

Observational

Source

ClinicalTrials.gov

Registry ID

NCT07409168

Enrollment

100

Registered

2026-02-13

Start date

2026-08-15

Completion date

2027-12-31

Last updated

2026-04-28

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

Natural Killer/T-cell Lymphoma

Brief summary

This is a multicenter prospective study to develop and validate a multimodal, deep learning-based model for predicting treatment response in patients with extranodal natural killer/T-cell lymphoma (NKTCL) receiving first-line asparaginase-based therapy.

Interventions

None listed

Study design

Observational model

COHORT

Time perspective

PROSPECTIVE

Eligibility

Sex/Gender

ALL

Age

18 Years to No maximum

Healthy volunteers

Inclusion criteria

* 1\. Age ≥ 18 years. * 2\. Pathologically confirmed extranodal natural killer/T-cell lymphoma (NKTCL) according to the World Health Organization (WHO) classification. * 3\. Patients who are planned to receive first-line asparaginase-based chemotherapy or chemoradiotherapy. * 4\. Patients who have either contrast-enhanced MRI of the nasopharynx obtained as part of routine clinical care or pretreatment whole-slide images (WSI) of tumor tissue from hematoxylin and eosin (H\&E)-stained sections available for analysis. * 5\. Ability to understand the study and provide written informed consent (ICF).

Exclusion criteria

* 1\. History of other malignant tumors. * 2\. Patients with psychiatric disorders or those unable to provide informed consent.

Design outcomes

Primary

Measure	Time frame	Description
Predictive accuracy of first-line treatment response (CR vs non-CR) according to Lugano 2014 criteria	From baseline to disease response and follow-up assessments, up to 3 years.	The primary outcome is the predictive performance of the multimodal deep learning model for first-line treatment response in patients with extranodal natural killer/T-cell lymphoma (NKTCL). Treatment response is assessed according to the Lugano 2014 criteria. Model performance will be evaluated by receiver operating characteristic (ROC) analysis and quantified using the area under the curve (AUC), accuracy, sensitivity, specificity, positive predictive value, and negative predictive value by comparing model predictions with observed clinical response.

Contacts

CONTACTQingqing Cai, MD. PhD.

caiqq@sysucc.org.cn0208734282

Outcome results