Christopher Kelly

Shawn Xu, Lin Yang, Christopher Kelly et al.

In this work, we present an approach, which we call Embeddings for Language/Image-aligned X-Rays, or ELIXR, that leverages a language-aligned image encoder combined or grafted onto a fixed LLM, PaLM 2, to perform a broad range of chest X-ray tasks. We train this lightweight adapter architecture using images paired with corresponding free-text radiology reports from the MIMIC-CXR dataset. ELIXR achieved state-of-the-art performance on zero-shot chest X-ray (CXR) classification (mean AUC of 0.850 across 13 findings), data-efficient CXR classification (mean AUCs of 0.893 and 0.898 across five findings (atelectasis, cardiomegaly, consolidation, pleural effusion, and pulmonary edema) for 1% (~2,200 images) and 10% (~22,000 images) training data), and semantic search (0.76 normalized discounted cumulative gain (NDCG) across nineteen queries, including perfect retrieval on twelve of them). Compared to existing data-efficient methods including supervised contrastive learning (SupCon), ELIXR required two orders of magnitude less data to reach similar performance. ELIXR also showed promise on CXR vision-language tasks, demonstrating overall accuracies of 58.7% and 62.5% on visual question answering and report quality assurance tasks, respectively. These results suggest that ELIXR is a robust and versatile approach to CXR AI.

Beatriz Costa-Gomes, Pavel Tolmachev, Eloise Taysom et al.

We analyze over 500,000 de-identified health-related conversations with Microsoft Copilot from January 2026 to characterize what people ask conversational AI about health. We develop a hierarchical intent taxonomy of 12 primary categories using privacy-preserving LLM-based classification validated against expert human annotation, and apply LLM-driven topic-clustering for prevalent themes within each intent. Using this taxonomy, we characterize the intents and topics behind health queries, identify who these queries are about, and analyze how usage varies by device and time of day. Five findings stand out. First, nearly one in five conversations involve personal symptom assessment or condition discussion, and even the dominant general information category (40%) is concentrated on specific treatments and conditions, suggesting that this is a lower bound on personal health intent. Second, one in seven of these personal health queries concern someone other than the user, such as a child, a parent, a partner, suggesting that conversational AI can be a caregiving tool, not just a personal one. Third, personal queries about symptoms and emotional health queries increase markedly in the evening and nighttime hours, when traditional healthcare is most limited. Fourth, usage diverges sharply by device: mobile concentrates on personal health concerns, while desktop is dominated by professional and academic work. Fifth, a substantial share of queries focuses on navigating healthcare systems such as finding providers, and understanding insurance, highlighting friction in the delivery of existing healthcare. These patterns have direct implications for platform-specific design, safety considerations, and the responsible development of health AI.

Stanislav Nikolov, Sam Blackwell, Alexei Zverovitch et al.

Over half a million individuals are diagnosed with head and neck cancer each year worldwide. Radiotherapy is an important curative treatment for this disease, but it requires manual time consuming delineation of radio-sensitive organs at risk (OARs). This planning process can delay treatment, while also introducing inter-operator variability with resulting downstream radiation dose differences. While auto-segmentation algorithms offer a potentially time-saving solution, the challenges in defining, quantifying and achieving expert performance remain. Adopting a deep learning approach, we demonstrate a 3D U-Net architecture that achieves expert-level performance in delineating 21 distinct head and neck OARs commonly segmented in clinical practice. The model was trained on a dataset of 663 deidentified computed tomography (CT) scans acquired in routine clinical practice and with both segmentations taken from clinical practice and segmentations created by experienced radiographers as part of this research, all in accordance with consensus OAR definitions. We demonstrate the model's clinical applicability by assessing its performance on a test set of 21 CT scans from clinical practice, each with the 21 OARs segmented by two independent experts. We also introduce surface Dice similarity coefficient (surface DSC), a new metric for the comparison of organ delineation, to quantify deviation between OAR surface contours rather than volumes, better reflecting the clinical task of correcting errors in the automated organ segmentations. The model's generalisability is then demonstrated on two distinct open source datasets, reflecting different centres and countries to model training. With appropriate validation studies and regulatory approvals, this system could improve the efficiency, consistency, and safety of radiotherapy pathways.

Harsha Nori, Mayank Daswani, Christopher Kelly et al.

Artificial intelligence holds great promise for expanding access to expert medical knowledge and reasoning. However, most evaluations of language models rely on static vignettes and multiple-choice questions that fail to reflect the complexity and nuance of evidence-based medicine in real-world settings. In clinical practice, physicians iteratively formulate and revise diagnostic hypotheses, adapting each subsequent question and test to what they've just learned, and weigh the evolving evidence before committing to a final diagnosis. To emulate this iterative process, we introduce the Sequential Diagnosis Benchmark, which transforms 304 diagnostically challenging New England Journal of Medicine clinicopathological conference (NEJM-CPC) cases into stepwise diagnostic encounters. A physician or AI begins with a short case abstract and must iteratively request additional details from a gatekeeper model that reveals findings only when explicitly queried. Performance is assessed not just by diagnostic accuracy but also by the cost of physician visits and tests performed. We also present the MAI Diagnostic Orchestrator (MAI-DxO), a model-agnostic orchestrator that simulates a panel of physicians, proposes likely differential diagnoses and strategically selects high-value, cost-effective tests. When paired with OpenAI's o3 model, MAI-DxO achieves 80% diagnostic accuracy--four times higher than the 20% average of generalist physicians. MAI-DxO also reduces diagnostic costs by 20% compared to physicians, and 70% compared to off-the-shelf o3. When configured for maximum accuracy, MAI-DxO achieves 85.5% accuracy. These performance gains with MAI-DxO generalize across models from the OpenAI, Gemini, Claude, Grok, DeepSeek, and Llama families. We highlight how AI systems, when guided to think iteratively and act judiciously, can advance diagnostic precision and cost-effectiveness in clinical care.

Christopher Kelly, Angelica Chowdhury, Alexandra Campili et al.

This work establishes a foundational framework for standardizing AI evaluation RCTs (sometimes called human uplift studies). Drawing on established experimental practices from disciplines with established RCT traditions, including software engineering, economics, clinical and health sciences, and psychology, we adopt the (Shadish et al., 2002) four-validity framework and extend it with a fifth principle on transparency, repeatability, and verification adapted from the Transparency and Openness Promotion (TOP) Guidelines (Center for Open Science, 2025). We operationalize all five principles into 33 guidelines adapted for AI evaluation RCT contexts, expressed as requirements with rationales, implementation instructions, and evidence bases. We position the principles and guidelines as serving three key roles for AI evaluation RCTs: a design tool for planning studies, an evaluation rubric for assessing existing work, and a blueprint for standard setting as the field converges on norms. Our framework extends prior work by centering evaluation on human performance rather than model output alone, formalizing causal inference through RCT methodology for AI contexts, integrating heterogeneity analysis and practical significance assessment, implementing a graded transparency and repeatability framework, and addressing AI-specific challenges including model versioning, human-AI interaction dynamics, contamination and spillover effects, and equitable impact assessment.

Vijayalakshmi Saravanan, Perry Siehien, Shinjae Yoo et al.

Detecting abrupt changes in real-time data streams from scientific simulations presents a challenging task, demanding the deployment of accurate and efficient algorithms. Identifying change points in live data stream involves continuous scrutiny of incoming observations for deviations in their statistical characteristics, particularly in high-volume data scenarios. Maintaining a balance between sudden change detection and minimizing false alarms is vital. Many existing algorithms for this purpose rely on known probability distributions, limiting their feasibility. In this study, we introduce the Kernel-based Cumulative Sum (KCUSUM) algorithm, a non-parametric extension of the traditional Cumulative Sum (CUSUM) method, which has gained prominence for its efficacy in online change point detection under less restrictive conditions. KCUSUM splits itself by comparing incoming samples directly with reference samples and computes a statistic grounded in the Maximum Mean Discrepancy (MMD) non-parametric framework. This approach extends KCUSUM's pertinence to scenarios where only reference samples are available, such as atomic trajectories of proteins in vacuum, facilitating the detection of deviations from the reference sample without prior knowledge of the data's underlying distribution. Furthermore, by harnessing MMD's inherent random-walk structure, we can theoretically analyze KCUSUM's performance across various use cases, including metrics like expected delay and mean runtime to false alarms. Finally, we discuss real-world use cases from scientific simulations such as NWChem CODAR and protein folding data, demonstrating KCUSUM's practical effectiveness in online change point detection.

Lin Yang, Shawn Xu, Andrew Sellergren et al.

Many clinical tasks require an understanding of specialized data, such as medical images and genomics, which is not typically found in general-purpose large multimodal models. Building upon Gemini's multimodal models, we develop several models within the new Med-Gemini family that inherit core capabilities of Gemini and are optimized for medical use via fine-tuning with 2D and 3D radiology, histopathology, ophthalmology, dermatology and genomic data. Med-Gemini-2D sets a new standard for AI-based chest X-ray (CXR) report generation based on expert evaluation, exceeding previous best results across two separate datasets by an absolute margin of 1% and 12%, where 57% and 96% of AI reports on normal cases, and 43% and 65% on abnormal cases, are evaluated as "equivalent or better" than the original radiologists' reports. We demonstrate the first ever large multimodal model-based report generation for 3D computed tomography (CT) volumes using Med-Gemini-3D, with 53% of AI reports considered clinically acceptable, although additional research is needed to meet expert radiologist reporting quality. Beyond report generation, Med-Gemini-2D surpasses the previous best performance in CXR visual question answering (VQA) and performs well in CXR classification and radiology VQA, exceeding SoTA or baselines on 17 of 20 tasks. In histopathology, ophthalmology, and dermatology image classification, Med-Gemini-2D surpasses baselines across 18 out of 20 tasks and approaches task-specific model performance. Beyond imaging, Med-Gemini-Polygenic outperforms the standard linear polygenic risk score-based approach for disease risk prediction and generalizes to genetically correlated diseases for which it has never been trained. Although further development and evaluation are necessary in the safety-critical medical domain, our results highlight the potential of Med-Gemini across a wide range of medical tasks.

Christopher Kelly, Nikolaos Pitropakis, Sean McKeown et al.

A large majority of cheap Internet of Things (IoT) devices that arrive brand new, and are configured with out-of-the-box settings, are not being properly secured by the manufactures, and are vulnerable to existing malware lurking on the Internet. Among them is the Mirai botnet which has had its source code leaked to the world, allowing any malicious actor to configure and unleash it. A combination of software assets not being utilised safely and effectively are exposing consumers to a full compromise. We configured and attacked 4 different IoT devices using the Mirai libraries. Our experiments concluded that three out of the four devices were vulnerable to the Mirai malware and became infected when deployed using their default configuration. This demonstrates that the original security configurations are not sufficient to provide acceptable levels of protection for consumers, leaving their devices exposed and vulnerable. By analysing the Mirai libraries and its attack vectors, we were able to determine appropriate device configuration countermeasures to harden the devices against this botnet, which were successfully validated through experimentation.

Peter Boyle, Michael Chuvelev, Guido Cossu et al.

We discuss practical methods to ensure near wirespeed performance from clusters with either one or two Intel(R) Omni-Path host fabric interfaces (HFI) per node, and Intel(R) Xeon Phi(TM) 72xx (Knight's Landing) processors, and using the Linux operating system. The study evaluates the performance improvements achievable and the required programming approaches in two distinct example problems: firstly in Cartesian communicator halo exchange problems, appropriate for structured grid PDE solvers that arise in quantum chromodynamics simulations of particle physics, and secondly in gradient reduction appropriate to synchronous stochastic gradient descent for machine learning. As an example, we accelerate a published Baidu Research reduction code and obtain a factor of ten speedup over the original code using the techniques discussed in this paper. This displays how a factor of ten speedup in strongly scaled distributed machine learning could be achieved when synchronous stochastic gradient descent is massively parallelised with a fixed mini-batch size. We find a significant improvement in performance robustness when memory is obtained using carefully allocated 2MB "huge" virtual memory pages, implying that either non-standard allocation routines should be used for communication buffers. These can be accessed via a LD\_PRELOAD override in the manner suggested by libhugetlbfs. We make use of a the Intel(R) MPI 2019 library "Technology Preview" and underlying software to enable thread concurrency throughout the communication software stake via multiple PSM2 endpoints per process and use of multiple independent MPI communicators. When using a single MPI process per node, we find that this greatly accelerates delivered bandwidth in many core Intel(R) Xeon Phi processors.