Machine learning-integrated immunoinflammatory profiling facilitates diagnostic risk stratification for systemic lupus erythematosus

Purpose Systemic lupus erythematosus (SLE) often presents with significant phenotypic overlap with other autoimmune conditions, complicating accurate differential diagnosis. To address this clinical challenge, we developed a computational architecture integrating immunoinflammatory biomarkers to facilitate objective diagnostic risk stratification for SLE among symptomatic patients. Methods A retrospective observational framework was adopted, encompassing 1,950 individuals partitioned into 1,000 definitive lupus cases and 950 symptomatic non-SLE controls. To extract core distinguishing indicators, a bipartite screening methodology merging LASSO techniques with Boruta random forest evaluations was deployed. This isolated seven essential molecular targets: antinuclear antibodies (ANA), C-reactive protein (CRP), interferon-gamma inducible protein 10 (IP-10), tumor necrosis factor-alpha (TNF-α), interleukin-6 (IL-6), B-cell activating factor (BAFF), and anti-double-stranded DNA (anti-dsDNA). Among seven rigorously evaluated machine learning algorithms, logistic regression was selected as the primary diagnostic model due to its optimal balance of predictive robustness and interpretability in the internal validation cohort. This algorithm was translated into a graphical scoring index. Furthermore, to address potential incorporation bias, an autoantibody-independent cytokine-only panel was evaluated. Results During internal validation cohort verification, the integrated logistic regression model achieved 81.0% specificity, 83.0% sensitivity, 82.0% accuracy, and an area under the receiver operating characteristic curve (AUC) of 0.851, outperforming individual predictors ( P < 0.001) and complex algorithms prone to structural overfitting. Crucially, the autoantibody-independent panel (comprising BAFF, IL-6, TNF-α, IP-10, and CRP) demonstrated robust independent diagnostic efficacy, achieving an AUC of 0.820 in the internal validation cohort and substantially outperforming the conventional ANA and anti-dsDNA baseline (AUC = 0.765). Decision curve analysis confirmed that the integrated multifactorial tool consistently delivered greater net clinical benefit across extensive threshold probabilities compared to standard diagnostic baselines. Conclusions This investigation delivers a promising computational strategy to assist in distinguishing SLE from phenotypically overlapping disorders. The developed graphical tool serves as a potential decision-support framework for objective risk stratification. Furthermore, prospective multicenter external validation will further enhance its generalizability and clinical utility.

Authors

Institutions

Publication Details

Journal
Frontiers in Immunology
Published
2026-11-17
DOI
https://doi.org/10.3389/fimmu.2026.1997874
Primary Topic
Systemic Lupus Erythematosus Research
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Machine learning-integrated immunoinflammatory profiling facilitates diagnostic risk stratification for systemic lupus erythematosus

Binfeng Peng, Yanli Hu, Xiaobing Xie, Haiying Li et al.
2 links
Frontiers in Immunology
Systemic Lupus Erythematosus Research
article

Machine learning-integrated immunoinflammatory profiling facilitates diagnostic risk stratification for systemic lupus erythematosus

Binfeng Peng, Yanli Hu, Xiaobing Xie, Haiying Li, Jing Li, Can Zuo
article en

Abstract

Purpose Systemic lupus erythematosus (SLE) often presents with significant phenotypic overlap with other autoimmune conditions, complicating accurate differential diagnosis. To address this clinical challenge, we developed a computational architecture integrating immunoinflammatory biomarkers to facilitate objective diagnostic risk stratification for SLE among symptomatic patients. Methods A retrospective observational framework was adopted, encompassing 1,950 individuals partitioned into 1,000 definitive lupus cases and 950 symptomatic non-SLE controls. To extract core distinguishing indicators, a bipartite screening methodology merging LASSO techniques with Boruta random forest evaluations was deployed. This isolated seven essential molecular targets: antinuclear antibodies (ANA), C-reactive protein (CRP), interferon-gamma inducible protein 10 (IP-10), tumor necrosis factor-alpha (TNF-α), interleukin-6 (IL-6), B-cell activating factor (BAFF), and anti-double-stranded DNA (anti-dsDNA). Among seven rigorously evaluated machine learning algorithms, logistic regression was selected as the primary diagnostic model due to its optimal balance of predictive robustness and interpretability in the internal validation cohort. This algorithm was translated into a graphical scoring index. Furthermore, to address potential incorporation bias, an autoantibody-independent cytokine-only panel was evaluated. Results During internal validation cohort verification, the integrated logistic regression model achieved 81.0% specificity, 83.0% sensitivity, 82.0% accuracy, and an area under the receiver operating characteristic curve (AUC) of 0.851, outperforming individual predictors ( P < 0.001) and complex algorithms prone to structural overfitting. Crucially, the autoantibody-independent panel (comprising BAFF, IL-6, TNF-α, IP-10, and CRP) demonstrated robust independent diagnostic efficacy, achieving an AUC of 0.820 in the internal validation cohort and substantially outperforming the conventional ANA and anti-dsDNA baseline (AUC = 0.765). Decision curve analysis confirmed that the integrated multifactorial tool consistently delivered greater net clinical benefit across extensive threshold probabilities compared to standard diagnostic baselines. Conclusions This investigation delivers a promising computational strategy to assist in distinguishing SLE from phenotypically overlapping disorders. The developed graphical tool serves as a potential decision-support framework for objective risk stratification. Furthermore, prospective multicenter external validation will further enhance its generalizability and clinical utility.

Frontiers in ImmunologyVol. 17
Central South University (CN), Hunan University of Traditional Chinese Medicine (CN), First Affiliated Hospital of Hunan University of Traditional Chinese Medicine (CN), Xiangya Hospital Central South University (CN)
Good health and well-being
Systemic Lupus Erythematosus Research
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon