LANTERN-XGB: An Interpretable Multi-Modal Machine Learning for Improving Clinical Decision-Making in Lung Cancer

Dalfovo, Davide; Sassorossi, Carolina; De Paolis, Elisa; Campanella, Annalisa; Nachira, Dania; Ciavarella, Leonardo Petracca; Boldrini, Luca; Troost, Esther G. C.; Ádány, Róza; Farré, Núria; Öztürk, Ece; Minucci, Angelo; Trisolini, Rocco; Bria, Emilio; Löck, Steffen; Margaritora, Stefano; Lococo, Filippo

doi:10.3390/ijms27073128

This is an early access version, the complete PDF, HTML, and XML versions will be available soon.

Open AccessArticle

LANTERN-XGB: An Interpretable Multi-Modal Machine Learning for Improving Clinical Decision-Making in Lung Cancer

by

Davide Dalfovo

^1,2,

Carolina Sassorossi

^3,4,*

,

Elisa De Paolis

^5,6,

Annalisa Campanella

^3,4,

Dania Nachira

^3,4

,

Leonardo Petracca Ciavarella

^3,4

,

Luca Boldrini

⁷

,

Esther G. C. Troost

^1,2,8,

Róza Ádány

⁹

,

Núria Farré

¹⁰,

Ece Öztürk

¹¹

,

Angelo Minucci

^5,6

,

Rocco Trisolini

¹²

,

Emilio Bria

¹³,

Steffen Löck

^1,2,8,†

,

Stefano Margaritora

^3,4 and

Filippo Lococo

^3,4,*,†

¹

OncoRay—National Center for Radiation Research in Oncology, Faculty of Medicine, University Hospital Carl Gustav Carus, TUD Dresden University of Technology, Helmholtz-Zentrum Dresden-Rossendorf, 01067 Dresden, Germany

²

Department of Radiotherapy and Radiation Oncology, Faculty of Medicine, University Hospital Carl Gustav Carus, TUD Dresden University of Technology, 01067 Dresden, Germany

³

Unit of Thoracic Surgery, Catholic University of the Sacred Heart, 00168 Rome, Italy

⁴

Thoracic Surgery Unit, A. Gemelli University Hospital Foundation IRCCS, 00168 Rome, Italy

⁵

Departmental Unit of Molecular and Genomic Diagnostics, Genomics Core Facility, Gemelli Science and Technology Park (G-STeP), A. Gemelli University Hospital Foundation IRCCS, 00168 Rome, Italy

⁶

Clinical Chemistry, Biochemistry and Molecular Biology Operations (UOC), A. Gemelli University Hospital Foundation IRCCS, 00168 Rome, Italy

⁷

Advanced Radiotherapy Center, A. Gemelli University Hospital Foundation IRCCS, 00168 Rome, Italy

⁸

Institute of Radiooncology—OncoRay, Helmholtz-Zentrum Dresden-Rossendorf, 01067 Dresden, Germany

⁹

ELKH-DE Public Health Research Group, Department of Public Health and Epidemiology, Faculty of Medicine, University of Debrecen, 4001 Debrecen, Hungary

¹⁰

Institut de Recerca de L’Hospital de la Santa Creu i Sant Pau (IR-HSCSP), 08001 Barcelona, Spain

¹¹

School of Medicine, Turkey and Koç University Research Center for Translational Medicine (KUTTAM), Koç University, Istanbul 34450, Turkey

¹²

Interventional Pulmonology Unit, A. Gemelli University Hospital Foundation IRCCS, 00168 Rome, Italy

¹³

Medical Oncology, A. Gemelli University Hospital Foundation IRCCS, Largo a. Gemelli 8, 00168 Rome, Italy

Show full affiliation list

Hide full affiliation list

^*

Authors to whom correspondence should be addressed.

^†

These authors contributed equally to this work.

Int. J. Mol. Sci. 2026, 27(7), 3128; https://doi.org/10.3390/ijms27073128

Submission received: 18 February 2026 / Revised: 17 March 2026 / Accepted: 24 March 2026 / Published: 30 March 2026

(This article belongs to the Special Issue Omics Science and Research in Human Health and Disease)

Download Versions Notes

Abstract

Non-small cell lung cancer (NSCLC) remains the leading cause of cancer-related mortality globally. While multi-modal artificial intelligence (AI) models offer significant predictive potential, their translation into routine clinical practice is delayed by the “black box” nature of complex algorithms and the fragmentation of heterogeneous data. We present LANTERN-XGB, a hierarchical machine learning workflow designed to bridge this gap by generating interpretable “digital human avatars” for precision oncology. The methodology employs a multi-stage scalable tree boosting system (XGBoost) architecture utilizing shapley additive explanations (SHAP) for rigorous hierarchical feature selection, missing value management, and patient-specific decision support. The workflow was developed and benchmarked using a retrospective cohort of 437 patients with clinical N0 NSCLC, followed by validation on a prospective dataset (n = 100) and an independent external dataset (n = 100). The pipeline integrates diverse data modalities to predict occult lymph node metastasis (OLM). LANTERN-XGB identified a robust consensus signature driven by non-linear interactions among CT textural fragmentation, PET metabolic heterogeneity, tumor density distribution, and systemic clinical modulators. Exploratory transcriptomic pathway analysis (GSVA) revealed that high-risk predictions strongly correlate with systemic molecular dysregulation, such as the enrichment of immune-inflammatory signaling and metabolic stress pathways. The model achieved robust discrimination in external validation (AUC ≈ 0.77), performing comparably to state-of-the-art nomogram benchmarks. Crucially, the LANTERN-XGB framework demonstrated superior utility in handling diagnostic ambiguity; local force plots allowed for the correct reclassification of “borderline” prediction by visualizing feature interactions that standard linear models fail to capture. LANTERN-XGB provides a validated, open-source framework that successfully balances predictive power with clinical transparency. By empowering clinicians to visualize and verify the logic behind AI predictions, this workflow offers a pragmatic path for integrating reliable multi-modal avatars into daily medical decision-making.

Keywords: multi-modal integration; artificial intelligence; precision oncology; lung cancer; radiogenomics

Share and Cite

MDPI and ACS Style

Dalfovo, D.; Sassorossi, C.; De Paolis, E.; Campanella, A.; Nachira, D.; Ciavarella, L.P.; Boldrini, L.; Troost, E.G.C.; Ádány, R.; Farré, N.; et al. LANTERN-XGB: An Interpretable Multi-Modal Machine Learning for Improving Clinical Decision-Making in Lung Cancer. Int. J. Mol. Sci. 2026, 27, 3128. https://doi.org/10.3390/ijms27073128

AMA Style

Dalfovo D, Sassorossi C, De Paolis E, Campanella A, Nachira D, Ciavarella LP, Boldrini L, Troost EGC, Ádány R, Farré N, et al. LANTERN-XGB: An Interpretable Multi-Modal Machine Learning for Improving Clinical Decision-Making in Lung Cancer. International Journal of Molecular Sciences. 2026; 27(7):3128. https://doi.org/10.3390/ijms27073128

Chicago/Turabian Style

Dalfovo, Davide, Carolina Sassorossi, Elisa De Paolis, Annalisa Campanella, Dania Nachira, Leonardo Petracca Ciavarella, Luca Boldrini, Esther G. C. Troost, Róza Ádány, Núria Farré, and et al. 2026. "LANTERN-XGB: An Interpretable Multi-Modal Machine Learning for Improving Clinical Decision-Making in Lung Cancer" International Journal of Molecular Sciences 27, no. 7: 3128. https://doi.org/10.3390/ijms27073128

APA Style

Dalfovo, D., Sassorossi, C., De Paolis, E., Campanella, A., Nachira, D., Ciavarella, L. P., Boldrini, L., Troost, E. G. C., Ádány, R., Farré, N., Öztürk, E., Minucci, A., Trisolini, R., Bria, E., Löck, S., Margaritora, S., & Lococo, F. (2026). LANTERN-XGB: An Interpretable Multi-Modal Machine Learning for Improving Clinical Decision-Making in Lung Cancer. International Journal of Molecular Sciences, 27(7), 3128. https://doi.org/10.3390/ijms27073128

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Article metric data becomes available approximately 24 hours after publication online.

Article Menu

LANTERN-XGB: An Interpretable Multi-Modal Machine Learning for Improving Clinical Decision-Making in Lung Cancer

Abstract

Share and Cite

Article Metrics

Further Information

Guidelines

MDPI Initiatives

Follow MDPI