With the increasing interest in deploying Artificial Intelligence in medicine, we previously introduced HAIM (Holistic AI in Medicine), a framework that fuses multimodal data to solve downstream clinical tasks. However, HAIM uses data in a task-agnostic manner and lacks explainability. To address these limitations, we introduce xHAIM (Explainable HAIM), a novel framework leveraging Generative AI to enhance both prediction and explainability through four structured steps: (1) automatically identifying task-relevant patient data across modalities, (2) generating comprehensive patient summaries, (3) using these summaries for improved predictive modeling, and (4) providing clinical explanations by linking predictions to patient-specific medical knowledge. Evaluated on the HAIM-MIMIC-MM dataset, xHAIM improves average AUC from 79.9% to 90.3% across chest pathology and operative tasks. Importantly, xHAIM transforms AI from a black-box predictor into an explainable decision support system, enabling clinicians to interactively trace predictions back to relevant patient data, bridging AI advancements with clinical utility.
In an era of rapid technological advancements and macroeconomic shifts, worker reallocation is necessary, yet responses to labor market shocks remain sluggish, making it crucial to identify bottlenecks in occupational transitions to understand labor market dynamics and improve mobility. In this study, we analyze French occupational data to uncover patterns of worker mobility and pinpoint specific occupations that act as bottlenecks which impede rapid reallocation. We introduce two metrics, transferability and accessibility, to quantify the diversity of occupational transitions and find that bottlenecks can be explained by a condensation effect of occupations with high accessibility but low transferability. Transferability measures the variety of transitions from an occupation to others, while accessibility assesses the variety of transitions into an occupation. We provide a comprehensive framework for analyzing occupational complexity and mobility patterns, offering insights into potential barriers and pathways for efficient retraining programs. We argue that our approach can inform policymakers and stakeholders aiming to enhance labor market efficiency and support workforce adapta
The success of precision medicine requires computational models that can effectively process and interpret diverse physiological signals across heterogeneous patient populations. While foundation models have demonstrated remarkable transfer capabilities across various domains, their effectiveness in handling individual-specific physiological signals - crucial for precision medicine - remains largely unexplored. This work introduces a systematic pipeline for rapidly and efficiently evaluating foundation models' transfer capabilities in medical contexts. Our pipeline employs a three-stage approach. First, it leverages physiological simulation software to generate diverse, clinically relevant scenarios, particularly focusing on data-scarce medical conditions. This simulation-based approach enables both targeted capability assessment and subsequent model fine-tuning. Second, the pipeline projects these simulated signals through the foundation model to obtain embeddings, which are then evaluated using linear methods. This evaluation quantifies the model's ability to capture three critical aspects: physiological feature independence, temporal dynamics preservation, and medical scenario d
Occupational mobility is an emergent strategy to cope with technological unemployment by facilitating efficient labor redeployment. However, previous studies analyzing networks show that the boundaries to smooth mobility are constrained by a fragmented structure in the occupation network. In this study, positing that this structure will significantly change due to automation, we propose the skill automation view, which asserts that automation substitutes for skills, not for occupations, and simulate a scenario of skill automation drawing on percolation theory. We sequentially remove skills from the occupation-skill bipartite network and investigate the structural changes in the projected occupation network. The results show that the accumulation of small changes (the emergence of bridges between occupations due to skill automation) triggers significant structural changes in the occupation network. The structural changes accelerate as the components integrate into a new giant component. This result suggests that automation mitigates the bottlenecks to smooth occupational mobility.
What does Artificial Intelligence (AI) have to contribute to health care? And what should we be looking out for if we are worried about its risks? In this paper we offer a survey, and initial evaluation, of hopes and fears about the applications of artificial intelligence in medicine. AI clearly has enormous potential as a research tool, in genomics and public health especially, as well as a diagnostic aid. It's also highly likely to impact on the organisational and business practices of healthcare systems in ways that are perhaps under-appreciated. Enthusiasts for AI have held out the prospect that it will free physicians up to spend more time attending to what really matters to them and their patients. We will argue that this claim depends upon implausible assumptions about the institutional and economic imperatives operating in contemporary healthcare settings. We will also highlight important concerns about privacy, surveillance, and bias in big data, as well as the risks of over trust in machines, the challenges of transparency, the deskilling of healthcare practitioners, the way AI reframes healthcare, and the implications of AI for the distribution of power in healthcare ins
This study examines the clinical decision-making processes in Traditional East Asian Medicine (TEAM) by reinterpreting pattern identification (PI) through the lens of dimensionality reduction. Focusing on the Eight Principle Pattern Identification (EPPI) system and utilizing empirical data from the Shang-Han-Lun, we explore the necessity and significance of prioritizing the Exterior-Interior pattern in diagnosis and treatment selection. We test three hypotheses: whether the Ext-Int pattern contains the most information about patient symptoms, represents the most abstract and generalizable symptom information, and facilitates the selection of appropriate herbal prescriptions. Employing quantitative measures such as the abstraction index, cross-conditional generalization performance, and decision tree regression, our results demonstrate that the Exterior-Interior pattern represents the most abstract and generalizable symptom information, contributing to the efficient mapping between symptom and herbal prescription spaces. This research provides an objective framework for understanding the cognitive processes underlying TEAM, bridging traditional medical practices with modern computat
In its broadest definition, systems biology is the application of a `systems' way of thinking about and doing cell biology. By implication, this also invites us to consider a systems approach in the context of medicine and the treatment of disease. In particular, the idea that systems biology can form the basis of a personalised, predictive medicine will require that much closer attention is paid to the analytic properties of the feedback loops which will be set up by a personalised approach to healthcare. To emphasize the role that feedback theory will play in understanding personalised medicine, we use the term feedback medicine to describe the issues outlined.In these notes we consider feedback and control systems concepts applied to two important themes in medical systems biology - personalised medicine and combinatorial intervention. In particular, we formulate a feedback control interpretation for the administration of medicine, and relate them to various forms of medical treatment.
Recent studies indicate that Generative Pre-trained Transformer 4 with Vision (GPT-4V) outperforms human physicians in medical challenge tasks. However, these evaluations primarily focused on the accuracy of multi-choice questions alone. Our study extends the current scope by conducting a comprehensive analysis of GPT-4V's rationales of image comprehension, recall of medical knowledge, and step-by-step multimodal reasoning when solving New England Journal of Medicine (NEJM) Image Challenges - an imaging quiz designed to test the knowledge and diagnostic capabilities of medical professionals. Evaluation results confirmed that GPT-4V performs comparatively to human physicians regarding multi-choice accuracy (81.6% vs. 77.8%). GPT-4V also performs well in cases where physicians incorrectly answer, with over 78% accuracy. However, we discovered that GPT-4V frequently presents flawed rationales in cases where it makes the correct final choices (35.5%), most prominent in image comprehension (27.2%). Regardless of GPT-4V's high accuracy in multi-choice questions, our findings emphasize the necessity for further in-depth evaluations of its rationales before integrating such multimodal AI m
Medicine, including fields in healthcare and life sciences, has seen a flurry of quantum-related activities and experiments in the last few years (although biology and quantum theory have arguably been entangled ever since Schrödinger's cat). The initial focus was on biochemical and computational biology problems; recently, however, clinical and medical quantum solutions have drawn increasing interest. The rapid emergence of quantum computing in health and medicine necessitates a mapping of the landscape. In this review, clinical and medical proof-of-concept quantum computing applications are outlined and put into perspective. These consist of over 40 experimental and theoretical studies. The use case areas span genomics, clinical research and discovery, diagnostics, and treatments and interventions. Quantum machine learning (QML) in particular has rapidly evolved and shown to be competitive with classical benchmarks in recent medical research. Near-term QML algorithms have been trained with diverse clinical and real-world data sets. This includes studies in generating new molecular entities as drug candidates, diagnosing based on medical image classification, predicting patient pe
3D data from high-resolution volumetric imaging is a central resource for diagnosis and treatment in modern medicine. While the fast development of AI enhances imaging and analysis, commonly used visualization methods lag far behind. Recent research used extended reality (XR) for perceiving 3D images with visual depth perception and touch but used restrictive haptic devices. While unrestricted touch benefits volumetric data examination, implementing natural haptic interaction with XR is challenging. The research question is whether a multisensory XR application with intuitive haptic interaction adds value and should be pursued. In a study, 24 experts for biomedical images in research and medicine explored 3D medical shapes with 3 applications: a multisensory virtual reality (VR) prototype using haptic gloves, a simple VR prototype using controllers, and a standard PC application. Results of standardized questionnaires showed no significant differences between all application types regarding usability and no significant difference between both VR applications regarding presence. Participants agreed to statements that VR visualizations provide better depth information, using the hand
Scientific and technological advances in medicine and systems biology have unequivocally shown that health and disease must be viewed in the context of the interplay among multiple molecular and environmental factors. Understanding the effects of cellular interconnection on disease progression may lead to the identification of novel disease genes and pathways, and hence influence precision diagnostics and therapeutics. To accomplish this goal, the emerging field of network medicine applies network science approaches to investigate disease pathogenesis, integrating information from relevant Omics databases, including protein-protein interaction, correlation-based, gene regulatory, and Bayesian networks. However, this requires analysing and computing large amounts of data. Moreover, if we are to efficiently search for new drugs and new drug combinations, there is a pressing need for computational methods that could allow us to access the immense chemical compound space until now largely unexplored. Finally, at the microscopic level, drug-target chemistry simulation is ultimately a quantum problem, and hence it requires a quantum solution. As we will discuss, quantum computing may be
The Oxford English Dictionary defines precision medicine as "medical care designed to optimize efficiency or therapeutic benefit for particular groups of patients, especially by using genetic or molecular profiling." It is not an entirely new idea: physicians from ancient times have recognized that medical treatment needs to consider individual variations in patient characteristics. However, the modern precision medicine movement has been enabled by a confluence of events: scientific advances in fields such as genetics and pharmacology, technological advances in mobile devices and wearable sensors, and methodological advances in computing and data sciences. This chapter is about bandit algorithms: an area of data science of special relevance to precision medicine. With their roots in the seminal work of Bellman, Robbins, Lai and others, bandit algorithms have come to occupy a central place in modern data science ( Lattimore and Szepesvari, 2020). Bandit algorithms can be used in any situation where treatment decisions need to be made to optimize some health outcome. Since precision medicine focuses on the use of patient characteristics to guide treatment, contextual bandit algorith
The theoretical base of the research of occupational injuries is the idea of the process as Markov chain of random variables. However the exact proof of this position was not carried out whereas the experimental passing of the hypothesis is connected always with the determined confidence limits and consequently it gives the space for alternative assumptions. In this research some databases of occupational injuries had been studied using spectral analysis techniques and the presentation of the occupational injuries as the temporal sequence of the cases ("telegraph wave" process type). Databases had corresponding chapters such as "enterprise" with number of employees about 7000, "big enterprise" (the number of employees about 35000), "whole branch of industry", "whole enterprises of industrial region" receiving during 10 years from different countries having distinguish system of the work organization (Russia and Italy). The behaviour of spectra on principal is not changed when vary the length of realization, resolution, smoothing, upper boundary frequency, country and year of datas. All spectra showed that the occupational injuries process has a not Markov, but deterministic polyhar
The last decade has seen an explosion in models that describe phenomena in systems medicine. Such models are especially useful for studying signaling pathways, such as the Wnt pathway. In this chapter we use the Wnt pathway to showcase current mathematical and statistical techniques that enable modelers to gain insight into (models of) gene regulation, and generate testable predictions. We introduce a range of modeling frameworks, but focus on ordinary differential equation (ODE) models since they remain the most widely used approach in systems biology and medicine and continue to offer great potential. We present methods for the analysis of a single model, comprising applications of standard dynamical systems approaches such as nondimensionalization, steady state, asymptotic and sensitivity analysis, and more recent statistical and algebraic approaches to compare models with data. We present parameter estimation and model comparison techniques, focusing on Bayesian analysis and coplanarity via algebraic geometry. Our intention is that this (non exhaustive) review may serve as a useful starting point for the analysis of models in systems medicine.
Time-to-event prediction, e.g. cancer survival analysis or hospital length of stay, is a highly prominent machine learning task in medical and healthcare applications. However, only a few interpretable machine learning methods comply with its challenges. To facilitate a comprehensive explanatory analysis of survival models, we formally introduce time-dependent feature effects and global feature importance explanations. We show how post-hoc interpretation methods allow for finding biases in AI systems predicting length of stay using a novel multi-modal dataset created from 1235 X-ray images with textual radiology reports annotated by human experts. Moreover, we evaluate cancer survival models beyond predictive performance to include the importance of multi-omics feature groups based on a large-scale benchmark comprising 11 datasets from The Cancer Genome Atlas (TCGA). Model developers can use the proposed methods to debug and improve machine learning algorithms, while physicians can discover disease biomarkers and assess their significance. We hope the contributed open data and code resources facilitate future work in the emerging research direction of explainable survival analysis.