Search

Large language models (LLMs) have shown remarkable progress in encoding clinical knowledge and responding to complex medical queries with appropriate clinical reasoning. However, their applicability in subspecialist or complex medical settings remains underexplored. In this work, we probe the performance of AMIE, a research conversational diagnostic AI system, in the subspecialist domain of breast oncology care without specific fine-tuning to this challenging domain. To perform this evaluation, we curated a set of 50 synthetic breast cancer vignettes representing a range of treatment-naive and treatment-refractory cases and mirroring the key information available to a multidisciplinary tumor board for decision-making (openly released with this work). We developed a detailed clinical rubric for evaluating management plans, including axes such as the quality of case summarization, safety of the proposed care plan, and recommendations for chemotherapy, radiotherapy, surgery and hormonal therapy. To improve performance, we enhanced AMIE with the inference-time ability to perform web search retrieval to gather relevant and up-to-date clinical knowledge and refine its responses with a multi-stage self-critique pipeline. We compare response quality of AMIE with internal medicine trainees, oncology fellows, and general oncology attendings under both automated and specialist clinician evaluations. In our evaluations, AMIE outperformed trainees and fellows demonstrating the potential of the system in this challenging and important domain. We further demonstrate through qualitative examples, how systems such as AMIE might facilitate conversational interactions to assist clinicians in their decision making. However, AMIE's performance was overall inferior to attending oncologists suggesting that further research is needed prior to consideration of prospective uses.

The scarcity of subspecialist medical expertise, particularly in rare, complex and life-threatening diseases, poses a significant challenge for healthcare delivery. This issue is particularly acute in cardiology where timely, accurate management determines outcomes. We explored the potential of AMIE (Articulate Medical Intelligence Explorer), a large language model (LLM)-based experimental AI system optimized for diagnostic dialogue, to potentially augment and support clinical decision-making in this challenging context. We curated a real-world dataset of 204 complex cases from a subspecialist cardiology practice, including results for electrocardiograms, echocardiograms, cardiac MRI, genetic tests, and cardiopulmonary stress tests. We developed a ten-domain evaluation rubric used by subspecialists to evaluate the quality of diagnosis and clinical management plans produced by general cardiologists or AMIE, the latter enhanced with web-search and self-critique capabilities. AMIE was rated superior to general cardiologists for 5 of the 10 domains (with preference ranging from 9% to 20%), and equivalent for the rest. Access to AMIE's response improved cardiologists' overall response quality in 63.7% of cases while lowering quality in just 3.4%. Cardiologists' responses with access to AMIE were superior to cardiologist responses without access to AMIE for all 10 domains. Qualitative examinations suggest AMIE and general cardiologist could complement each other, with AMIE thorough and sensitive, while general cardiologist concise and specific. Overall, our results suggest that specialized medical LLMs have the potential to augment general cardiologists' capabilities by bridging gaps in subspecialty expertise, though further research and validation are essential for wide clinical utility.

MIMO (multiple input, multiple output) approaches are a recent trend in neural network architectures for video restoration problems, where each network evaluation produces multiple output frames. The video is split into non-overlapping stacks of frames that are processed independently, resulting in a very appealing trade-off between output quality and computational cost. In this work we focus on the low-latency setting by limiting the number of available future frames. We find that MIMO architectures suffer from problems that have received little attention so far, namely (1) the performance drops significantly due to the reduced temporal receptive field, particularly for frames at the borders of the stack, (2) there are strong temporal discontinuities at stack transitions which induce a step-wise motion artifact. We propose two simple solutions to alleviate these problems: recurrence across MIMO stacks to boost the output quality by implicitly increasing the temporal receptive field, and overlapping of the output stacks to smooth the temporal discontinuity at stack transitions. These modifications can be applied to any MIMO architecture. We test them on three state-of-the-art video denoising networks with different computational cost. The proposed contributions result in a new state-of-the-art for low-latency networks, both in terms of reconstruction error and temporal consistency. As an additional contribution, we introduce a new benchmark consisting of drone footage that highlights temporal consistency issues that are not apparent in the standard benchmarks., Comment: See the project web page to download the associated videos

In this work, we introduce Gemma 2, a new addition to the Gemma family of lightweight, state-of-the-art open models, ranging in scale from 2 billion to 27 billion parameters. In this new version, we apply several known technical modifications to the Transformer architecture, such as interleaving local-global attentions (Beltagy et al., 2020a) and group-query attention (Ainslie et al., 2023). We also train the 2B and 9B models with knowledge distillation (Hinton et al., 2015) instead of next token prediction. The resulting models deliver the best performance for their size, and even offer competitive alternatives to models that are 2-3 times bigger. We release all our models to the community.

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.

Excellence in a wide variety of medical applications poses considerable challenges for AI, requiring advanced reasoning, access to up-to-date medical knowledge and understanding of complex multimodal data. Gemini models, with strong general capabilities in multimodal and long-context reasoning, offer exciting possibilities in medicine. Building on these core strengths of Gemini, we introduce Med-Gemini, a family of highly capable multimodal models that are specialized in medicine with the ability to seamlessly use web search, and that can be efficiently tailored to novel modalities using custom encoders. We evaluate Med-Gemini on 14 medical benchmarks, establishing new state-of-the-art (SoTA) performance on 10 of them, and surpass the GPT-4 model family on every benchmark where a direct comparison is viable, often by a wide margin. On the popular MedQA (USMLE) benchmark, our best-performing Med-Gemini model achieves SoTA performance of 91.1% accuracy, using a novel uncertainty-guided search strategy. On 7 multimodal benchmarks including NEJM Image Challenges and MMMU (health & medicine), Med-Gemini improves over GPT-4V by an average relative margin of 44.5%. We demonstrate the effectiveness of Med-Gemini's long-context capabilities through SoTA performance on a needle-in-a-haystack retrieval task from long de-identified health records and medical video question answering, surpassing prior bespoke methods using only in-context learning. Finally, Med-Gemini's performance suggests real-world utility by surpassing human experts on tasks such as medical text summarization, alongside demonstrations of promising potential for multimodal medical dialogue, medical research and education. Taken together, our results offer compelling evidence for Med-Gemini's potential, although further rigorous evaluation will be crucial before real-world deployment in this safety-critical domain.

We present an experimental study of the magnetoelectric coupling (MEC) in the Fe-Ga/PMN-PT thin film multiferroic composite by means of x-ray magnetic circular dichroism (XMCD) and ferromagnetic resonance (FMR). Our measurements show evidence for a charge mediated coupling mechanism, suggested by the asymmetric magnetic remanence (M$_{rem}$) behaviour under opposite electric fields ($\pm$ E) and the asymmetric resonance field (H$_r$) in the FMR measurements. Also, the FMR measurements reveal a perpendicular magnetic anisotropy (PMA), that can be related to an interface charge effect. and it is tunable by E field. Ab initio calculations support the existence of a charge mediated coupling at the Fe-Ga/PMN-PT interface., Comment: 7 pages, 5 figures, Supplemental Material

The emerging field of quantum computing has shown it might change how we process information by using the unique principles of quantum mechanics. As researchers continue to push the boundaries of quantum technologies to unprecedented levels, distributed quantum computing raises as an obvious path to explore with the aim of boosting the computational power of current quantum systems. This paper presents a comprehensive survey of the current state of the art in the distributed quantum computing field, exploring its foundational principles, landscape of achievements, challenges, and promising directions for further research. From quantum communication protocols to entanglement-based distributed algorithms, each aspect contributes to the mosaic of distributed quantum computing, making it an attractive approach to address the limitations of classical computing. Our objective is to provide an exhaustive overview for experienced researchers and field newcomers.

Large language models (LLMs) hold promise to serve complex health information needs but also have the potential to introduce harm and exacerbate health disparities. Reliably evaluating equity-related model failures is a critical step toward developing systems that promote health equity. We present resources and methodologies for surfacing biases with potential to precipitate equity-related harms in long-form, LLM-generated answers to medical questions and conduct a large-scale empirical case study with the Med-PaLM 2 LLM. Our contributions include a multifactorial framework for human assessment of LLM-generated answers for biases, and EquityMedQA, a collection of seven datasets enriched for adversarial queries. Both our human assessment framework and dataset design process are grounded in an iterative participatory approach and review of Med-PaLM 2 answers. Through our empirical study, we find that our approach surfaces biases that may be missed via narrower evaluation approaches. Our experience underscores the importance of using diverse assessment methodologies and involving raters of varying backgrounds and expertise. While our approach is not sufficient to holistically assess whether the deployment of an AI system promotes equitable health outcomes, we hope that it can be leveraged and built upon towards a shared goal of LLMs that promote accessible and equitable healthcare.

This work introduces Gemma, a family of lightweight, state-of-the art open models built from the research and technology used to create Gemini models. Gemma models demonstrate strong performance across academic benchmarks for language understanding, reasoning, and safety. We release two sizes of models (2 billion and 7 billion parameters), and provide both pretrained and fine-tuned checkpoints. Gemma outperforms similarly sized open models on 11 out of 18 text-based tasks, and we present comprehensive evaluations of safety and responsibility aspects of the models, alongside a detailed description of model development. We believe the responsible release of LLMs is critical for improving the safety of frontier models, and for enabling the next wave of LLM innovations.

This article discusses the development of an IoT system for monitoring and controlling various devices and systems from different vendors. The authors considered key challenges in IoT projects, such as interoperability and integration, scalability, and data storage, processing, and visualization, during the design and deployment phases. In addition to these general challenges, the authors also delve into the specific integration challenges they encountered. Various devices and systems were integrated into the system and five real-world scenarios in a university campus environment are used to illustrate the challenges encountered. The scenarios involve monitoring various aspects of a university campus environment, including air quality, environmental parameters, energy efficiency, solar photovoltaic energy, and energy consumption. The authors analyzed data and CPU usage to ensure that the system could handle the large amount of data generated by the devices. The platform developed uses open source projects such as Home Assistant, InfluxDB, Grafana, and Node-RED. All developments have been published as open source in public repositories. In conclusion, this work highlights the potential and feasibility of IoT systems in various real-world applications, the importance of considering key challenges in IoT projects during the design and deployment phases, and the specific integration challenges that may be encountered., Comment: 21 pages, 16 figures, Published in Internet of Things

In the world of Internet of Things (IoT), obtaining the physical location of devices has always been a task of great interest for developing increasingly complex location-based services (LBS). That is why in recent years wireless communication standards have been incorporating new additions focused on providing localization mechanisms to technologies widely used in the IoT world, such as Wi-Fi or Bluetooth. In particular, the IEEE 802.11-2016 Wi-Fi standard introduced ranging estimation between two devices through the so-called fine time measurement (FTM) protocol, defined by the IEEE 802.11mc. FTM is not yet widespread in the IoT field, but commercial modules capable of offering this functionality at a reasonable price are starting to appear. In early 2021, the most widespread system on a chip (SOC) family among IoT devices, the ESP32-XX series, added support for this Wi-Fi standard, enabling, for the first time, the use of a standard designed for location-based systems. This article analyzes the performance of this FTM implementation by carrying out and studying several measurement campaigns in different indoor and outdoor scenarios. Additionally, this work proposes an alternative real-time implementation for distance estimation inside the ESP32 using an approach based on machine learning. Such an implementation is successfully validated in a scenario totally different than those considered for the training and test sets. Finally, both the measurement sets and the developed software are available to the scientific community., Comment: 14 pages, 21 figures, Published in IEEE Internet of Things Journal

In contemporary society, the pressing challenge of preserving user privacy clashes with the imperative for smart buildings to efficiently manage their resources, particularly in the context of occupancy monitoring for optimized energy utilization. This paper delves into the application of millimiter wave (mmWave) frequency modulated continuous wave (FMCW) radar technology for occupancy monitoring. mmWave FMCW radar, unlike conventional methods that often require the use of identifiable tags or involve image analysis, operates without the need for such identifiers, mitigating privacy concerns. However, challenges arise when attempting to cover extensive indoor spaces due to the limited range of individual mmWave FMCW radar devices. The present work proposes the use of a flexible software architecture to integrate the measurements of several mmWave FMCW radar devices, so that the whole behaves as a single sensor. To validate the proposal, an example of use in a real environment in an indoor space monitored with three mmWave FMCW radar devices is also presented. The example details the whole process, from the physical installation of the devices to the use of the different software modules that allow the integration of the data into a common internet of things (IoT) management platform such as Home Assistant. All the elements, from the measurements captured during the test to the different software implementations, are shared publicly with the scientific community., Comment: 13 pages, 9 figures, submitted to IEEE Internet of Things Journal

Today's needs for monitoring and control of different devices in organizations require an Internet of Things (IoT) platform that can integrate heterogeneous elements provided by multiple vendors and using different protocols, data formats and communication technologies. This article provides a comprehensive review of all the architectures, technologies, protocols and data formats most commonly used by existing IoT platforms. On this basis, a comparative analysis of the most widely used open source IoT platforms is presented. This exhaustive comparison is based on multiple characteristics that will be essential to select the platform that best suits the needs of each organization., Comment: 15 pages, 4 figures, Published in Internet of Things

At the heart of medicine lies the physician-patient dialogue, where skillful history-taking paves the way for accurate diagnosis, effective management, and enduring trust. Artificial Intelligence (AI) systems capable of diagnostic dialogue could increase accessibility, consistency, and quality of care. However, approximating clinicians' expertise is an outstanding grand challenge. Here, we introduce AMIE (Articulate Medical Intelligence Explorer), a Large Language Model (LLM) based AI system optimized for diagnostic dialogue. AMIE uses a novel self-play based simulated environment with automated feedback mechanisms for scaling learning across diverse disease conditions, specialties, and contexts. We designed a framework for evaluating clinically-meaningful axes of performance including history-taking, diagnostic accuracy, management reasoning, communication skills, and empathy. We compared AMIE's performance to that of primary care physicians (PCPs) in a randomized, double-blind crossover study of text-based consultations with validated patient actors in the style of an Objective Structured Clinical Examination (OSCE). The study included 149 case scenarios from clinical providers in Canada, the UK, and India, 20 PCPs for comparison with AMIE, and evaluations by specialist physicians and patient actors. AMIE demonstrated greater diagnostic accuracy and superior performance on 28 of 32 axes according to specialist physicians and 24 of 26 axes according to patient actors. Our research has several limitations and should be interpreted with appropriate caution. Clinicians were limited to unfamiliar synchronous text-chat which permits large-scale LLM-patient interactions but is not representative of usual clinical practice. While further research is required before AMIE could be translated to real-world settings, the results represent a milestone towards conversational diagnostic AI., Comment: 46 pages, 5 figures in main text, 19 figures in appendix

$Ab$ $initio$ calculations show the presence of a strong magnetoelectric interfacial coupling in CaMnO$_3$ ultra-thin film grown on a strained BaTiO$_3$ ferroelectric film. This heterostructure presents a polarization driven magnetic transition from a G-type to an A-type antiferromagnetic structure. Together with this magnetic transition we find a metallic behaviour at the interface between these two insulators, where the charge character of the carriers can be tuned from electrons to holes by switching the electric polarization direction of the FE film. Besides, the electron gas is spin-polarized while the hole-gas is not., Comment: 8 pages, 5 figures, supplementary

In 1981, Frisch and Morf [1] postulated the existence of complex singularities in solutions of Navier-Stokes equations. Present progress on this conjecture is hindered by the computational burden involved in simulations of the Euler equations or the Navier-Stokes equations at high Reynolds numbers. We investigate this conjecture in the case of fluid dynamics on log-lattices, where the computational burden is logarithmic concerning ordinary fluid simulations. We analyze properties of potential complex singularities in both 1D and 3D models for lattices of different spacings. Dominant complex singularities are tracked using the singularity strip method to obtain new scalings regarding the approach to the real axis and the influence of normal, hypo and hyper dissipation.

Non-Gaussian entangled states play a crucial role in harnessing quantum advantage in continuous-variable quantum information. However, how to fully characterize N-partite (N > 3) non-Gaussian entanglement without quantum state tomography remains elusive, leading to a very limited understanding of the underlying entanglement mechanism. Here, we propose several necessary and sufficient conditions for the positive-partial-transposition separability of multimode nonlinear quantum states resulting from high-order Hamiltonians and successive beam splitting operations. When applied to the initial state, the beam-splitter operations induce the emergence of different types of entanglement mechanisms, including pairwise high-order entanglement, collective high-order entanglement and the crossover between the two. We show numerically that for the four-mode scenario, the threshold for the existence of entanglement for any bipartition does not exceed the entanglement of the original state at fixed high-order moments. These results provide a new perspective for understanding multipartite nonlinear entanglement and will promote their application in quantum information processing., Comment: Comments are welcome

An accurate differential diagnosis (DDx) is a cornerstone of medical care, often reached through an iterative process of interpretation that combines clinical history, physical examination, investigations and procedures. Interactive interfaces powered by Large Language Models (LLMs) present new opportunities to both assist and automate aspects of this process. In this study, we introduce an LLM optimized for diagnostic reasoning, and evaluate its ability to generate a DDx alone or as an aid to clinicians. 20 clinicians evaluated 302 challenging, real-world medical cases sourced from the New England Journal of Medicine (NEJM) case reports. Each case report was read by two clinicians, who were randomized to one of two assistive conditions: either assistance from search engines and standard medical resources, or LLM assistance in addition to these tools. All clinicians provided a baseline, unassisted DDx prior to using the respective assistive tools. Our LLM for DDx exhibited standalone performance that exceeded that of unassisted clinicians (top-10 accuracy 59.1% vs 33.6%, [p = 0.04]). Comparing the two assisted study arms, the DDx quality score was higher for clinicians assisted by our LLM (top-10 accuracy 51.7%) compared to clinicians without its assistance (36.1%) (McNemar's Test: 45.7, p < 0.01) and clinicians with search (44.4%) (4.75, p = 0.03). Further, clinicians assisted by our LLM arrived at more comprehensive differential lists than those without its assistance. Our study suggests that our LLM for DDx has potential to improve clinicians' diagnostic reasoning and accuracy in challenging cases, meriting further real-world evaluation for its ability to empower physicians and widen patients' access to specialist-level expertise.

Radiology reports are an instrumental part of modern medicine, informing key clinical decisions such as diagnosis and treatment. The worldwide shortage of radiologists, however, restricts access to expert care and imposes heavy workloads, contributing to avoidable errors and delays in report delivery. While recent progress in automated report generation with vision-language models offer clear potential in ameliorating the situation, the path to real-world adoption has been stymied by the challenge of evaluating the clinical quality of AI-generated reports. In this study, we build a state-of-the-art report generation system for chest radiographs, $\textit{Flamingo-CXR}$, by fine-tuning a well-known vision-language foundation model on radiology data. To evaluate the quality of the AI-generated reports, a group of 16 certified radiologists provide detailed evaluations of AI-generated and human written reports for chest X-rays from an intensive care setting in the United States and an inpatient setting in India. At least one radiologist (out of two per case) preferred the AI report to the ground truth report in over 60$\%$ of cases for both datasets. Amongst the subset of AI-generated reports that contain errors, the most frequently cited reasons were related to the location and finding, whereas for human written reports, most mistakes were related to severity and finding. This disparity suggested potential complementarity between our AI system and human experts, prompting us to develop an assistive scenario in which Flamingo-CXR generates a first-draft report, which is subsequently revised by a clinician. This is the first demonstration of clinician-AI collaboration for report writing, and the resultant reports are assessed to be equivalent or preferred by at least one radiologist to reports written by experts alone in 80$\%$ of in-patient cases and 60$\%$ of intensive care cases.

Technologies play an increasingly important role in sports and become a real competitive advantage for the athletes who benefit from it. Among them, the use of motion capture is developing in various sports to optimize sporting gestures. Unfortunately, traditional motion capture systems are expensive and constraining. Recently developed computer vision-based approaches also struggle in certain sports, like swimming, due to the aquatic environment. One of the reasons for the gap in performance is the lack of labeled datasets with swimming videos. In an attempt to address this issue, we introduce SwimXYZ, a synthetic dataset of swimming motions and videos. SwimXYZ contains 3.4 million frames annotated with ground truth 2D and 3D joints, as well as 240 sequences of swimming motions in the SMPL parameters format. In addition to making this dataset publicly available, we present use cases for SwimXYZ in swimming stroke clustering and 2D pose estimation., Comment: ACM MIG 2023

Non-Gaussian quantum states are a known necessary resource for reaching a quantum advantage and for violating Bell inequalities in continuous variable systems. As one kind of manifestation of quantum correlations, Einstein-Podolsky-Rosen (EPR) steering enables verification of shared entanglement even when one of the subsystems is not characterized. However, how to detect and classify such an effect for non-Gaussian states is far from being well understood. Here, we present an efficient non-Gaussian steering criterion based on the high-order observables and conduct a systematic investigation into the hierarchy of non-Gaussian steering criteria. Moreover, we apply our criterion to three experimentally-relevant non-Gaussian states under realistic conditions and, in particular, propose a feasible scheme to create multi-component cat states with tunable size by performing a suitable high-order quadrature measurement on the steering party. Our work reveals the fundamental characteristics of non-Gaussianity and quantum correlations, and offers new insights to explore their applications in quantum information processing.

We propose an automated nonlinear model reduction and mesh adaptation framework for rapid and reliable solution of parameterized advection-dominated problems, with emphasis on compressible flows. The key features of our approach are threefold: (i) a metric-based mesh adaptation technique to generate an accurate mesh for a range of parameters, (ii) a general (i.e., independent of the underlying equations) registration procedure for the computation of a mapping $\Phi$ that tracks moving features of the solution field, and (iii) an hyper-reduced least-square Petrov-Galerkin reduced-order model for the rapid and reliable estimation of the mapped solution. We discuss a general paradigm -- which mimics the refinement loop considered in mesh adaptation -- to simultaneously construct the high-fidelity and the reduced-order approximations, and we discuss actionable strategies to accelerate the offline phase. We present extensive numerical investigations for a quasi-1D nozzle problem and for a two-dimensional inviscid flow past a Gaussian bump to display the many features of the methodology and to assess the performance for problems with discontinuous solutions.

Medicine is inherently multimodal, with rich data modalities spanning text, imaging, genomics, and more. Generalist biomedical artificial intelligence (AI) systems that flexibly encode, integrate, and interpret this data at scale can potentially enable impactful applications ranging from scientific discovery to care delivery. To enable the development of these models, we first curate MultiMedBench, a new multimodal biomedical benchmark. MultiMedBench encompasses 14 diverse tasks such as medical question answering, mammography and dermatology image interpretation, radiology report generation and summarization, and genomic variant calling. We then introduce Med-PaLM Multimodal (Med-PaLM M), our proof of concept for a generalist biomedical AI system. Med-PaLM M is a large multimodal generative model that flexibly encodes and interprets biomedical data including clinical language, imaging, and genomics with the same set of model weights. Med-PaLM M reaches performance competitive with or exceeding the state of the art on all MultiMedBench tasks, often surpassing specialist models by a wide margin. We also report examples of zero-shot generalization to novel medical concepts and tasks, positive transfer learning across tasks, and emergent zero-shot medical reasoning. To further probe the capabilities and limitations of Med-PaLM M, we conduct a radiologist evaluation of model-generated (and human) chest X-ray reports and observe encouraging performance across model scales. In a side-by-side ranking on 246 retrospective chest X-rays, clinicians express a pairwise preference for Med-PaLM M reports over those produced by radiologists in up to 40.50% of cases, suggesting potential clinical utility. While considerable work is needed to validate these models in real-world use cases, our results represent a milestone towards the development of generalist biomedical AI systems.

Historically, the resolution of optical imaging systems was dictated by diffraction, and the Rayleigh criterion was long considered an unsurpassable limit. In superresolution microscopy, this limit is overcome by manipulating the emission properties of the object. However, in passive imaging, when sources are uncontrolled, reaching sub-Rayleigh resolution remains a challenge. Here, we implement a quantum-metrolgy-inspired approach for estimating the separation between two incoherent sources, achieving a sensitivity five orders of magnitude beyond the Rayleigh limit. Using a spatial mode demultiplexer, we examine scenes with bright and faint sources, through intensity measurements in the Hermite-Gauss basis. Analysing sensitivity and accuracy over an extensive range of separations, we demonstrate the remarkable effectiveness of demultiplexing for sub-Rayleigh separation estimation. These results effectively render the Rayleigh limit obsolete for passive imaging.