Singapore

Models trained on data from one site often underperform at another due to distribution shift, which can impede reliable deployment for medical imaging tasks. We tackle the problem of multi-center healthcare AI, where models must generalize across hospitals with heterogeneous imaging protocols, scanner hardware, and patient demographics. We propose a principled framework based on optimal transport dataset distances (OTDD) to analyze and guide cross‑site transfer. While OT-based approaches have mostly been applied to benchmark, non-medical datasets, we demonstrate its utility in a multi-center healthcare setting. We (i) build OTDD distance matrices between centers from fixed image embeddings (ResNet‑50, ResNet‑18), (ii) cluster centers using OT distance to reveal structure, and (iii) test whether transfer performance (finetuning on a single source, testing on a target) aligns with OTDD. We find that continuous OTDD distances correlate negatively with target AUROC across source-target pairs indicating that centers close in OTDD yield higher AUROC upon finetuning. Our results support OTDD as a center‑level selector/ranker for cross‑site adaptation in multi‑center healthcare, particularly when evaluated with threshold‑free metrics like AUROC.

AAAI 2026

Multi-Center Domain Shift in Healthcare AI: Optimal Transport-Based Quantification and Adaptation

multi center healthcare ai

distribution shift

optimal transport

The purpose of the AAAI conference is to promote research in artificial intelligence (AI) and scientific exchange among AI researchers, practitioners, scientists, and engineers in affiliated disciplines. 

AAAI 2022

technical paper

We are pleased to announce the Fortieth AAAI Conference on Artificial Intelligence (AAAI-26), which will be held in Singapore EXPO from January 20 to January 27, 2026.

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines. AAAI-26 will feature technical paper presentations, special tracks, invited speakers, workshops, tutorials, poster sessions, senior member presentations, competitions, and exhibit programs, and a range of other activities to be announced.<br><br>

To access this event page, you need to log in with the **email address you registered with**. <br>Access credentials will be sent to your email from Underline -  subject line "Welcome to AAAI 2026". Please be sure to check your spam email folder if you do not see an email confirmation right away.

Please log in

To access this event page, you are required to register.
Please complete your registration to continue.

We recommend reading [**the registration information**](https://aaai.org/conference/aaai/aaai-26/registration/) first.

**Online Registration Form**: https://aaai.getregistered.net/conference-2026 

Registration Required

We are pleased to announce the Fortieth AAAI Conference on Artificial Intelligence (AAAI-26), which will be held in Singapore EXPO from January 20 to January 27, 2026.

Mental health challenges and cyberbullying are increasingly prevalent in digital spaces, necessitating scalable and interpretable detection systems. This paper introduces a unified multiclass classification framework for detecting ten distinct mental health and cyberbullying categories from social media data. We curate datasets from Twitter and Reddit, implementing a rigorous ‘split-then-balance’ pipeline to train on balanced data while evaluating on a realistic, held-out imbalanced test set. We conduct a comprehensive evaluation comparing traditional lexical models, hybrid approaches, and several end-to-end fine-tuned transformers. Our results demonstrate that end-to-end fine-tuning is critical for performance, with the domain-adapted MentalBERT emerging as the top model, achieving an accuracy of 0.92 and a Macro F1 score of 0.76, surpassing both its generic counterpart and a zero-shot LLM baseline. Grounded in a comprehensive ethical analysis, we frame the system as a human-in-the-loop screening aid, not a diagnostic tool. To support this, we introduce a hybrid SHAP-LLM explainability framework and present a prototype dashboard (“Social Media Screener”) designed to integrate model predictions and their explanations into a practical workflow for moderators. Our work provides a robust baseline, highlighting future needs for multi-label, clinically validated datasets at the critical intersection of online safety and computational mental health.

A Machine Learning Approach for Detection of Mental Health Conditions and Cyberbullying from Social Media

Accurate and reliable medical image classification is critical for clinical decision-making across diverse imaging modalities, including X-ray, CT, and MRI. Traditional convolutional neural networks often produce overconfident predictions, limiting their clinical trustworthiness. In this work, we propose an uncertainty-aware, attention-augmented neural network that integrates multi-scale SwirlAttention and FeedBackAttention modules with a Bayesian probabilistic classifier. This framework enables robust feature extraction, interpretable attention maps, and principled estimation of epistemic uncertainty. We evaluate our approach on four diverse datasets, including Diabetic Retinopathy, Kvasir, Skin Cancer, and fused multi-focal Oocyte images, covering a wide range of pathological and morphological variations. Extensive experiments demonstrate that our method outperforms state-of-the-art CNN and transformer-based baselines in terms of accuracy, calibration, and interpretability. Grad-CAM visualizations highlight clinically relevant regions, while uncertainty estimates provide actionable insights for ambiguous cases, making the framework suitable for reliable deployment in real-world clinical settings.

BUCAN: Bayesian Uncertainty-aware Classification with Attention Networks for Medical Images

Recent advances in large language models (LLMs) have shown strong reasoning capabilities through large-scale pretraining and post-training reinforcement learning, demonstrated by DeepSeek-R1. However, current post-training methods, such as Grouped Relative Policy Optimization (GRPO), mainly reward correctness, which is not aligned with the multi-dimensional objectives required in high-stakes fields such as medicine, where reasoning must also be faithful and comprehensive. We introduce Clinical-Objective Relative Policy Optimization (CRPO), a scalable, multi-objective, verifiable reinforcement learning method designed to align LLM post-training with clinical reasoning principles. CRPO integrates rule-based and verifiable reward signals that jointly optimize accuracy, faithfulness, and comprehensiveness without relying on human annotation. To demonstrate its effectiveness, we train Clinical-R1-3B, a 3B-parameter model for clinical reasoning. The experiments on three benchmarks demonstrate that our CRPO substantially improves reasoning on truthfulness and completeness over standard GRPO while maintaining comfortable accuracy enhancements. This framework provides a scalable pathway to align LLM reasoning with clinical objectives, enabling safer and more collaborative AI systems for healthcare while also highlighting the potential of multi-objective, verifiable RL methods in post-training scaling of LLMs for medical domains.

Clinical-R1: Empowering Large Language Models for Faithful and Comprehensive Reasoning with Clinical Objective Relative Policy Optimization

Federated learning enables collaborative model development across medical institutions without centralizing sensitive patient data, yet existing embedding-level generative approaches often degrade under non-IID clinical heterogeneity and offer limited formal protection against gradient leakage. We introduce FedHypeVAE, a differentially private, hypernetwork based conditional variational framework that generates client-specific decoders and priors from lightweight, trainable client codes. This bi-level formulation personalizes the generative process while ensuring privacy preserving parameter synthesis decoupled from raw medical images. Federated optimization with differential privacy and distributional alignment strategies improves stability and cross-site generalization. The proposed framework unifies personalization, privacy, and domain adaptability within the generative layer, offering a principled solution for privacy aware representation learning in multi institutional medical imaging.

FedHypeVAE: Federated Learning with Hypernetwork Generated Conditional VAEs for Differentially Private Embedding Sharing

Recent developments on chest radiographs has primarily focused on developing multi-disease frameworks that aim to diagnose a wide range of thoracic abnormalities from the Chest X-ray datasets. In contrast, this study specifically targets pneumothorax, a life-threatening condition commonly referred to as a collapsed lung, which requires timely detection and accurate clinical reporting. Existing automated report generation Vision-Language Models (VLMs) mainly rely on image-level features and often fail to fully leverage the rich structural information embedded in medical image segmentation. To address this limitation, we propose a distinct strategy to incorporate pneumothorax segmentation masks, which delineate affected regions and provide precise localization guidance to enhance the accuracy of medical image interpretation. Experimental results demonstrate that the proposed segmentation-guided approach integrates visual and textual understanding more effectively for pneumothorax diagnosis from chest radiographs. By employing segmentation masks as guidance, VLMs can accurately localize pathological regions while preserving anatomical context, thereby improving both interpretability and diagnostic precision. Quantitative evaluations across multiple metrics further confirm the effectiveness of the proposed methods in bridging the gap between image-level localization and report-level reasoning.

Segmentation-Guided Radiology Report Generation for Pneumothorax Detection in Chest X-Rays

Mammography is one of the primary imaging modalities for breast cancer screening and diagnosis, playing a pivotal role in early detection and mortality reduction. To alleviate the burden on radiologists interpreting mammographic data, artificial intelligence-based models have emerged as decision-making assistants, demonstrating promising results in several studies. However, a critical gap remains between AI capabilities and clinical integration. Current AI systems predominantly employ end-to-end classification approaches that bypass the structured, multi-step reasoning radiologists use in practice. Radiologists systematically detect abnormalities, characterize their features using standardized descriptors, correlate findings across imaging views, perform temporal comparisons, assess information sufficiency, and synthesize evidence into risk-stratified recommendations. In contrast, most AI models map directly from images to diagnoses without transparent intermediate reasoning, limiting their interpretability, clinical utility, and ability to generalize beyond training distributions. This paper examines the radiological chain-of-thought process in mammography interpretation, reviews current AI approaches and their limitations, and proposes a multi-stage reasoning framework that explicitly models each step of clinical decision-making. By decomposing the diagnostic task into sequential reasoning-based stages, this framework aims to create AI systems that not only predict outcomes but also reason transparently in alignment with clinical workflow. Crucially, we explain how contextual information such as breast density serves as a conditioning variable that dynamically optimizes subsequent reasoning stages, mirroring the adaptive decision-making process radiologists employ in clinical practice. We discuss the implications of this approach and identify critical dataset limitations that must be addressed to enable the development of truly reasoning-aware AI in breast imaging.

The Need to Move Beyond Explainability Toward Chain-of-Thought Reasoning: A Focus on AI for Mammography

Clinical ultrasound acquisition is highly operator-dependent, where rapid probe motion and brightness fluctuations often lead to reconstruction errors that reduce trust and clinical utility. We present UltrasODM, a dual-stream framework that assists sonographers during acquisition through calibrated per-frame uncertainty, saliency-based diagnostics, and actionable prompts. UltrasODM integrates (i) a contrastive ranking module that groups frames by motion similarity, (ii) an optical-flow stream fused with Dual Mamba temporal modules for robust 6-DoF pose estimation, and (iii) a Human-in-the-Loop (HITL) layer combining Bayesian uncertainty, clinician-calibrated thresholds, and saliency maps highlighting regions of low confidence. When uncertainty exceeds the threshold, the system issues unobtrusive alerts suggesting corrective actions such as re-scanning highlighted regions or slowing the sweep. Evaluated on a clinical freehand ultrasound dataset, UltrasODM reduces drift by 15.2%, distance error by 12.1%, and Hausdorff distance by 10.1% relative to UltrasOM, while producing per-frame uncertainty and saliency outputs. By emphasizing transparency and clinician feedback, UltrasODM improves reconstruction reliability and supports safer, more trustworthy clinical workflows.

UltrasODM: A Dual Stream Optical Flow Mamba Network for 3D Freehand Ultrasound Reconstruction

The rise in clinical AI has helped to define the diagnostic and decision-making process in the dimension of modern medicine. However, their implementation is constrained by multiple ethical, interpretability, and safety factors that hinder the effective utilization of these AI systems. Existing Human-in-the-Loop (HITL) systems rely heavily on externally triggered human oversight, which creates critical gaps in effective safety and accountability. This study introduces the AI-Instigated Human Oversight (AIHO) framework, an AI governance architecture that enables models to self-assess uncertainty or contextual failure and autonomously escalate decisions to human oversight through a four-layer detection and escalation mechanism. Each layer performs a distinct self-assessment: (Layer 1) predictive uncertainty quantification, (Layer 2) contextual validation and explainability, (Layer 3) ethical and proxy alignment monitoring, and (Layer 4) adaptive governance with human-in-command enforcement. These layers form part of a three-zone operational architecture: (1) Information Flow and AI Core Processing, (2) AIHO Oversight Core, and (3) Human Escalation Loops. These zones in conjunction create a continuous feedback loop which administers continuous self-evaluation, transforming ethical and technical anomalies into actionable human oversight triggers. AIHO establishes a dynamic pathway toward regulation-ready, self-aware clinical AI, aligning with the current international standards for trustworthy and accountable AI in medicine.

AI-Instigated Human Oversight: Rethinking Human-in-the-Loop Safety in Clinical AI

Recent advancements in deep learning have enabled expert-level performance in medical imaging for disease classification, but their black-box decision making processes limit trust in them and their wide-spread clinical deployment. While Explainable Artificial Intelligence (XAI) methods aim to bridge this gap, studies focus on 2D data or pre-processed research datasets that overlook the role of medical imaging pre-processing operations which is an essential component of real-world 3D medical imaging workflows. To address this limitation, we propose XIMED, a systematic and predictive model–centered framework for evaluating explainability under realistic medical pre-processing conditions for volumetric medical data. The framework integrates five volumetric pre-processing variants and ten post-hoc attribution methods, evaluated through three complementary criteria: Correctness, Contrastivity, and Completeness, which together evaluate explanation dependence on model input, internal structure, and output behavior. Across more than 300 experimental configurations, XIMED reveals that gradient-based methods, such as Integrated Gradients and Blur Integrated Gradients, provide the most consistent and model-aligned explanations, while noise-based approaches like SmoothGrad and VarGrad are less sensitive to model behavior. These findings emphasize the importance of clinically realistic evaluation pipelines for reliable explainability in 3D medical imaging.

XIME3D : A Systematic Framework for Evaluating Explainable AI in 3D Medical Imaging under CT Image Pre-processing Variations

Aortic stenosis (AS) remains a major cardiovascular challenge, where early ECG indicators often elude conventional diagnostic approaches. Despite the success of deep learning in AS detection, clinical translation is impeded by limited interpretability of model decisions. This study investigates how neural network architecture influences explainability and interpretability in AS classification. Three models—ResNet-18 (CNN-based), LSTM, and ViT-MAE—were trained on an open-access cardiomechanical signals dataset of 100 valvular heart disease patients. Using Integrated Gradients, all models achieved strong performance (ResNet-18: 97.23%, LSTM: 98.96%, ViT-MAE: 88.56%) but displayed distinct attention mechanisms: ResNet-18 utilized global ECG features, LSTM captured temporal QRS dependencies, and ViT-MAE emphasized T-wave and QT interval features. All models consistently attended to clinically meaningful regions aligned with known AS physiology. These results reveal that explainability is architecture-dependent and advocate for architectureaware interpretation strategies to foster transparency, reliability, and clinical trust in cardiovascular AI applications.

Downloads

Next from AAAI 2026

A Machine Learning Approach for Detection of Mental Health Conditions and Cyberbullying from Social Media

Similar lecture

Optimal Tensor Transport

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from AAAI 2026

A Machine Learning Approach for Detection of Mental Health Conditions and Cyberbullying from Social Media

Similar lecture

Optimal Tensor Transport

Downloads