Clinical-Injection Transformer with Domain-Adapted MAE for Lupus Nephritis Prognosis Prediction

Yuewen Huang; Zhitao Ye; Guangnan Feng; Fudan Zheng; Xia Gao; Yutong Lu

eess.IVMar 4, 2026

Clinical-Injection Transformer with Domain-Adapted MAE for Lupus Nephritis Prognosis Prediction

Authors: Yuewen Huang, Zhitao Ye, Guangnan Feng, Fudan Zheng, Xia Gao, Yutong Lu

Organizations: Sun Yat-sen University · Department of Nephrology, Guangzhou Women and Children’s Medical Center, Guangzhou Medical University, Guangzhou, China

Abstract

Lupus nephritis (LN) is a severe complication of systemic lupus erythematosus that affects pediatric patients with significantly greater severity and worse renal outcomes compared to adults. Despite the urgent clinical need, predicting pediatric LN prognosis remains unexplored in computational pathology. Furthermore, the only existing histopathology-based approach for LN relies on multiple costly staining protocols and fails to integrate complementary clinical data. To address these gaps, we propose the first multimodal computational pathology framework for three-class treatment response prediction (complete remission, partial response, and no response) in pediatric LN, utilizing only routine PAS-stained biopsies and structured clinical data. Our framework introduces two key methodological innovations. First, a Clinical-Injection Transformer (CIT) embeds clinical features as condition tokens into patch-level self-attention, facilitating implicit and bidirectional cross-modal interactions within a unified attention space. Second, we design a decoupled representation-knowledge adaptation strategy using a domain-adapted Masked Autoencoder (MAE). This strategy explicitly separates self-supervised morphological feature learning from pathological knowledge extraction. Additionally, we introduce a multi-granularity morphological type injection mechanism to bridge distilled classification knowledge with downstream prognostic predictions at both the instance and patient levels. Evaluated on a cohort of 71 pediatric LN patients with KDIGO-standardized labels, our method achieves a three-class accuracy of 90.1% and an AUC of 89.4%, demonstrating its potential as a highly accurate and cost-effective prognostic tool.

Figures & tables

Figure 1: Overview of the proposed framework

Method	Acc(%)	W-F1(%)	M-F1(%)	AUC(%)
ABMIL [ 17 ] (image-only)	68.6 $\pm$ 1.8	62.1 $\pm$ 1.0	38.9 $\pm$ 2.0	45.2 $\pm$ 3.3
TransMIL [ 22 ] (image-only)	67.6 $\pm$ 2.1	63.5 $\pm$ 2.1	43.5 $\pm$ 2.1	60.6 $\pm$ 0.6
CLAM [ 23 ] (image-only)	65.8 $\pm$ 0.8	65.4 $\pm$ 1.4	51.3 $\pm$ 2.5	61.0 $\pm$ 1.5
Clinical MLP (0m+3m)	80.7 $\pm$ 4.3	80.4 $\pm$ 3.3	70.2 $\pm$ 4.9	81.9 $\pm$ 3.2
Late Fusion (0m+3m)	83.1 $\pm$ 2.3	82.5 $\pm$ 2.7	74.6 $\pm$ 3.4	85.4 $\pm$ 2.1
CIT (Ours, 0m)	86.4 $\pm$ 2.6	85.0 $\pm$ 2.6	77.0 $\pm$ 4.3	88.0 $\pm$ 1.9

Table 1: Three-class prognosis prediction (CR/PR/NR) on

n

=71 patients. Results are 5-fold cross-validation mean

\pm

std. Best in bold .

Figure 2: Mean MIL attention weights by glomerulus morphological type for each outcome class (3-seed aggregate, 0m+3m $\rightarrow$ 12m). PR and NR patients show elevated attention to pathological types (mesangial proliferative, sclerotic), while CR patients exhibit more uniform attention distribution.

(a) Fusion Architecture (0m+3m $\rightarrow$ 12m, $n$ =71)
Method	Acc(%)	W-F1(%)	M-F1(%)	AUC(%)
Cross-Attention [ 8 ]	83.1 $\pm$ 3.9	83.6 $\pm$ 3.7	76.7 $\pm$ 5.8	87.1 $\pm$ 5.5
Late Fusion	83.1 $\pm$ 2.3	82.5 $\pm$ 2.7	74.6 $\pm$ 3.4	85.4 $\pm$ 2.1
CIT (Ours)	90.1 $\pm$ 1.1	89.2 $\pm$ 1.0	83.9 $\pm$ 1.3	89.4 $\pm$ 2.9
(b) Feature Extractor (0m+3m $\rightarrow$ 12m, $n$ =71)
Feature Source	Acc(%)	W-F1(%)	M-F1(%)	AUC(%)

Table 2: Ablation studies. (a) Fusion architecture, (b) feature extractor, and (c) morphological type injection, all on 0m+3m

\rightarrow

12m (

n

=71). 5-fold cross-validation, mean

\pm

std over 3 seeds.

Explore similar work

Mar 16, 2026

Learning from Limited and Incomplete Data: A Multimodal Framework for Predicting Pathological Response in NSCLC

Major pathological response (pR) following neoadjuvant therapy is a clinically meaningful endpoint in non-small cell lung cancer, strongly associated with improved survival. However, accurate preoperative prediction of pR remains challenging, particularly in real-world clinical settings characterized by limited data availability and incomplete clinical profiles. In this study, we propose a multimodal deep learning framework designed to address these constraints by integrating foundation model-based CT feature extraction with a missing-aware architecture for clinical variables. This approach enables robust learning from small cohorts while explicitly modeling missing clinical information, without relying on conventional imputation strategies. A weighted fusion mechanism is employed to leverage the complementary contributions of imaging and clinical modalities, yielding a multimodal model that consistently outperforms both unimodal imaging and clinical baselines. These findings underscore the added value of integrating heterogeneous data sources and highlight the potential of multimodal, missing-aware systems to support pR prediction under realistic clinical conditions.

Alice Natalina Caragliano, Giulia Farina, Fatih Aksu +10

Unit of Artificial Intelligence and Computer Systems, Department of Engineering, Università Campus Bio-Medico di Roma, Rome, Italy · Fondazione Policlinico Universitario Campus Bio-Medico, Rome, Italy · Department of Medicine and Surgery, Università Campus Bio-Medico di Roma, Rome, Italy +1

Clinical Outcome Prediction Clinical Prediction Multimodal Learning Learning with Missing Data

Ask PDF

Mar 3, 2026

GloPath: An Entity-Centric Foundation Model for Glomerular Lesion Assessment and Clinicopathological Insights

Glomerular pathology is central to the diagnosis and prognosis of renal diseases, yet the heterogeneity of glomerular morphology and fine-grained lesion patterns remain challenging for current AI approaches. We present GloPath, an entity-centric foundation model trained on over one million glomeruli extracted from 14,049 renal biopsy specimens using multi-scale and multi-view self-supervised learning. GloPath addresses two major challenges in nephropathology: glomerular lesion assessment and clinicopathological insights discovery. For lesion assessment, GloPath was benchmarked across three independent cohorts on 52 tasks, including lesion recognition, grading, few-shot classification, and cross-modality diagnosis-outperforming state-of-the-art methods in 42 tasks (80.8%). In the large-scale real-world study, it achieved an ROC-AUC of 91.51% for lesion recognition, demonstrating strong robustness in routine clinical settings. For clinicopathological insights, GloPath systematically revealed statistically significant associations between glomerular morphological parameters and clinical indicators across 224 morphology-clinical variable pairs, demonstrating its capacity to connect tissue-level pathology with patient-level outcomes. Together, these results position GloPath as a scalable and interpretable platform for glomerular lesion assessment and clinicopathological discovery, representing a step toward clinically translatable AI in renal pathology.

Qiming He, Jing Li, Tian Guan +26

Institute of Biopharmaceutical and Health Engineering, Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen, China · Interdisciplinary Institute for Medical Engineering, Fuzhou University, Fuzhou, China · Medical Optical Technology R&D Center, Research Institute of Tsinghua, Pearl River Delta, Guangzhou, China +6

Pathology Foundation Models Histopathology Image Classification

Ask PDF

Jun 10, 2026

AGE-MIL: Anchor-Guided Evidence Learning for Patient-Level Prediction

Existing computational pathology methods predominantly operate within whole-slide image (WSI)-level multiple instance learning (MIL) paradigms, while patient-level modeling remains underexplored. In routine pathological practice, however, pathologists derive diagnostic and prognostic conclusions by integrating evidence across multiple WSIs rather than relying on any single slide. This discrepancy creates a fundamental misalignment when patient-level supervision is directly imposed on conventional MIL frameworks, often leading to unstable optimization and degraded predictive reliability. To address this issue, we propose Anchor-Guided Evidence MIL (AGE-MIL), a weakly supervised framework for patient-level prediction. AGE-MIL constructs a patient-level anchor from slide representations to capture global pathological context and guide the retrieval and integration of diagnostically relevant local patches, enabling robust patient-level modeling. Patient-level risk is further modeled as an evidence accumulation process, promoting stable optimization under weak supervision. AGE-MIL is evaluated on six clinically relevant patient-level prediction tasks from two independent cohorts. Experimental results show that the proposed framework consistently outperforms eight state-of-the-art MIL methods. Code is available at https://github.com/wodeniua/AGE-MIL.

Jiawei Niu, Jian Chen, Di Zhang +8

School of Computer Science and Technology, Xi’an Jiaotong University, Xi’an, China · Department of Oncology, University of Cambridge, Cambridge, UK · Xiangya School of Medicine, Central South University, Changsha, Hunan, China

Computational Pathology Weakly Supervised Learning Clinical Prediction Multiple Instance Learning Whole-Slide Image Analysis

Ask PDF