Current care in multiple sclerosis (MS) primarily relies on infrequently obtained data such as magnetic resonance imaging, clinical laboratory tests or clinical history, resulting in subtle changes that may occur between visits being missed. Mobile technology enables continual collection of data and can pave the path for predicting complex aspects of MS such as symptoms and disease courses. To this end, we conducted a first-of-its-kind observational study called MS Mosaic. First, we developed and publicly launched a mobile app for collecting longitudinal data from MS subjects in the United States. Second, we ran the study across 3 years in order to capture complex patterns for this slow progressing disease. Finally, we retrospectively developed three classical ML methods and two deep learning models to accurately and continually predict the incidence of five high-severity symptoms (fatigue, sensory disturbance, walking instability, depression or anxiety and cramps/spasms) three months in advance.

Deep Generative Models are frequently used to learn continuous representations of complex data distributions using a finite number of samples. For any generative model, including pre-trained foundation models with GAN, Transformer or Diffusion architectures, generation performance can vary significantly based on which part of the learned data manifold is sampled. In this paper we study the post-training local geometry of the learned manifold and its relationship to generation outcomes for models ranging from toy settings to the latent decoder of the near state-of-the-art Stable Diffusion 1.4 Text-to-Image model. Building on the theory of continuous piecewise-linear (CPWL) generators, we characterize the local geometry in terms of three geometric descriptors - scaling ($\psi$), rank ($\nu$), and complexity ($\delta$). We provide quantitative and qualitative evidence showing that for a given latent, the local descriptors are indicative of generation aesthetics, artifacts, diversity, and memorization. Finally we demonstrate that training a reward model using the local geometry allows us to control the log-likelihood of a generated sample under the learned distribution, and improve the qualitative aspects of an image.

Recent advancements in large language model(LLM) performance on medical multiple choice question (MCQ) benchmarks have stimulated interest from healthcare providers and patients globally. Particularly in low-and middle-income countries (LMICs) facing acute physician shortages and lack of specialists, LLMs offer a potentially scalable pathway to enhance healthcare access and reduce costs. However, their effectiveness in the Global South, especially across the African continent, remains to be established. In this work, we introduce AfriMed-QA, the first large scale Pan-African English multi-specialty medical Question-Answering (QA) dataset, 15,000 questions (open and closed-ended) sourced from over 60 medical schools across 16 countries, covering 32 medical specialties. We further evaluate 30 LLMs across multiple axes including correctness and demographic bias. Our findings show significant performance variation across specialties and geographies, MCQ performance clearly lags USMLE (MedQA). We find that biomedical LLMs underperform general models and smaller edge-friendly LLMs struggle to achieve a passing score. Interestingly, human evaluations show a consistent consumer preference for LLM answers and explanations when compared with clinician answers.

Artificial Intelligence (AI) for health has the potential to significantly change and improve healthcare. However in most African countries, identifying culturally and contextually attuned approaches for deploying these solutions is not well understood. To bridge this gap, we conduct a qualitative study to investigate the best practices, fairness indicators, and potential biases to mitigate when deploying AI for health in African countries, as well as explore opportunities where artificial intelligence could make a positive impact in health. We used a mixed methods approach combining in-depth interviews (IDIs) and surveys. We conduct 1.5-2 hour long IDIs with 50 experts in health, policy, and AI across 17 countries, and through an inductive approach we conduct a qualitative thematic analysis on expert IDI responses. We administer a blinded 30-minute survey with case studies to 672 general population participants across 5 countries in Africa and analyze responses on quantitative scales, statistically comparing responses by country, age, gender, and level of familiarity with AI. We thematically summarize open-ended responses from surveys. Our results find generally positive attitudes, high levels of trust, accompanied by moderate levels of concern among general population participants for AI usage for health in Africa. This contrasts with expert responses, where major themes revolved around trust/mistrust, ethical concerns, and systemic barriers to integration, among others. This work presents the first-of-its-kind qualitative research study of the potential of AI for health in Africa from an algorithmic fairness angle, with perspectives from both experts and the general population. We hope that this work guides policymakers and drives home the need for further research and the inclusion of general population perspectives in decision-making around AI usage.

Disaggregated evaluation across subgroups is critical for assessing the fairness of machine learning models, but its uncritical use can mislead practitioners. We show that equal performance across subgroups is an unreliable measure of fairness when data are representative of the relevant populations but reflective of real-world disparities. Furthermore, when data are not representative due to selection bias, both disaggregated evaluation and alternative approaches based on conditional independence testing may be invalid without explicit assumptions regarding the bias mechanism. We use causal graphical models to characterize fairness properties and metric stability across subgroups under different data generating processes. Our framework suggests complementing disaggregated evaluations with explicit causal assumptions and analysis to control for confounding and distribution shift, including conditional independence testing and weighted performance estimation. These findings have broad implications for how practitioners design and interpret model assessments given the ubiquity of disaggregated evaluation.

Concept erasure techniques have recently gained significant attention for their potential to remove unwanted concepts from text-to-image models. While these methods often demonstrate promising results in controlled settings, their robustness in real-world applications and suitability for deployment remain uncertain. In this work, we (1) identify a critical gap in evaluating sanitized models, particularly in assessing their performance across diverse concept dimensions, and (2) systematically analyze the failure modes of text-to-image models post-erasure. We focus on the unintended consequences of concept removal on non-target concepts across different levels of interconnected relationships including visually similar, binomial, and semantically related concepts. To address this, we introduce EraseBench, a comprehensive benchmark for evaluating post-erasure performance. EraseBench includes over 100 curated concepts, targeted evaluation prompts, and a robust set of metrics to assess both effectiveness and side effects of erasure. Our findings reveal a phenomenon of concept entanglement, where erasure leads to unintended suppression of non-target concepts, causing spillover degradation that manifests as distortions and a decline in generation quality.

We introduce a large language model (LLM) capable of processing speech inputs and show that tuning it further with reinforcement learning on human preference (RLHF) enables it to adapt better to disordered speech than traditional fine-tuning. Our method replaces low-frequency text tokens in an LLM's vocabulary with audio tokens and enables the model to recognize speech by fine-tuning it on speech with transcripts. We then use RL with rewards based on syntactic and semantic accuracy measures generalizing the LLM further to recognize disordered speech. While the resulting LLM does not outperform existing systems for speech recognition, we find that tuning with reinforcement learning using custom rewards leads to substantially better performance than supervised fine-tuning of the language model, specifically when adapting to speech in a different setting. This presents a compelling alternative tuning strategy for speech recognition using large language models.

The variability in the visual interpretation of cardiotocograms (CTGs) poses substantial challenges in obstetric care. Despite recent strides in automated CTG interpretation for early detection of fetal hypoxia, the comparative efficacy of objective versus subjective ground truth labels and robustness to temporal distribution shifts remains underexplored. Using a published convolutional neural network (CNN), we predict fetal compromise from CTG recordings, incorporating pre-processing and hyperparameter tuning. We use an open-source dataset of CTGs from 552 patients at University Hospital Brno, Czech Republic. Models trained with objective umbilical cord blood pH measurements (abnormal: pH < 7.20) outperformed those trained with subjective clinician-assigned Apgar scores (abnormal: Apgar < 7), demonstrating greater consistency and robustness to temporal shifts. This reflects the heterogeneity of Apgar scores, making them a more complex classification target. Additionally, aligning training signal intervals with the timing of outcome measurement exhibited superior performance, holding particular relevance for intermittent CTG measurement scenarios.

The AI research community plays a vital role in shaping the scientific, engineering, and societal goals of AI research. In this position paper, we argue that focusing on the highly contested topic of `artificial general intelligence' (`AGI') undermines our ability to choose effective goals. We identify six key traps – obstacles to productive goal setting – that are aggravated by AGI discourse: Illusion of Consensus, Supercharging Bad Science, Presuming Value-Neutrality, Goal Lottery, Generality Debt, and Normalized Exclusion. To avoid these traps, we argue that the AI research community needs to (1) prioritize specificity in engineering and societal goals, (2) center pluralism about multiple worthwhile approaches to multiple valuable goals, and (3) foster innovation through greater inclusion of disciplines and communities. Therefore, the AI research community needs to stop treating `AGI' as the north-star goal of AI research.

We introduce an artificial intelligence model to personalize treatment in major depression, which was deployed in the Artificial Intelligence in Depression: Medication Enhancement Study. We predict probabilities of remission across multiple pharmacological treatments, validate model predictions, and examine them for biases. Data from 9042 adults with moderate to severe major depression from antidepressant clinical trials were used to train a deep learning model. On the held-out test-set, the model demonstrated an AUC of 0.65, outperformed a null model (p = 0.01). The model increased population remission rate in hypothetical and actual improvement testing. While the model identified escitalopram as generally outperforming other drugs (consistent with the input data), there was otherwise significant variation in drug rankings. The model did not amplify potentially harmful biases. We demonstrate the first model capable of predicting outcomes for 10 treatments, intended to be used at or near the start of treatment to personalize treatment selection.

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 our 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 by 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 artificial intelligence (AI) system promotes equitable health outcomes, we hope that it can be leveraged and built upon toward a shared goal of LLMs that promote accessible and equitable healthcare. Identifying a complex panel of bias dimensions to be evaluated, a framework is proposed to assess how prone large language models are to biased reasoning, with possible consequences on equity-related harms, and is applied to a large-scale and diverse user survey on Med-PaLM 2.

With growing application of machine learning (ML) technologies in healthcare, there have been calls for developing techniques to understand and mitigate biases these systems may exhibit. Fair-ness considerations in the development of ML-based solutions for health have particular implications for Africa, which already faces inequitable power imbalances between the Global North and South.This paper seeks to explore fairness for global health, with Africa as a case study. We conduct a scoping review to propose axes of disparities for fairness consideration in the African context and delineate where they may come into play in different ML-enabled medical modalities. We then conduct qualitative research studies with 672 general population study participants and 28 experts inML, health, and policy focused on Africa to obtain corroborative evidence on the proposed axes of disparities. Our analysis focuses on colonialism as the attribute of interest and examines the interplay between artificial intelligence (AI), health, and colonialism. Among the pre-identified attributes, we found that colonial history, country of origin, and national income level were specific axes of disparities that participants believed would cause an AI system to be biased.However, there was also divergence of opinion between experts and general population participants. Whereas experts generally expressed a shared view about the relevance of colonial history for the development and implementation of AI technologies in Africa, the majority of the general population participants surveyed did not think there was a direct link between AI and colonialism. Based on these findings, we provide practical recommendations for developing fairness-aware ML solutions for health in Africa.

Recommender systems are widely used to help people find items that are tailored to their interests. These interests are often influenced by social networks, making it important to use social network information effectively in recommender systems. This is especially true for demographic groups with interests that differ from the majority. This paper introduces STUDY, a Socially-aware Temporally caUsal Decoder recommender sYstem. STUDY introduces a new socially-aware recommender system architecture that is significantly more efficient to learn and train than existing methods. STUDY performs joint inference over socially connected groups in a single forward pass of a modified transformer decoder network. We demonstrate the benefits of STUDY in the recommendation of books for students who are dyslexic, or struggling readers. Dyslexic students often have difficulty engaging with reading material, making it critical to recommend books that are tailored to their interests. We worked with our non-profit partner Learning Ally to evaluate STUDY on a dataset of struggling readers. STUDY was able to generate recommendations that more accurately predicted student engagement, when compared with existing methods.

INTRODUCTION: The pharmacological treatment of Major Depressive Disorder (MDD) relies on a trial-and-error approach. We introduce an artificial intelligence (AI) model aiming to personalize treatment and improve outcomes, which was deployed in the Artificial Intelligence in Depression Medication Enhancement (AIDME) Study. OBJECTIVES: 1) Develop a model capable of predicting probabilities of remission across multiple pharmacological treatments for adults with at least moderate major depression. 2) Validate model predictions and examine them for amplification of harmful biases. METHODS: Data from previous clinical trials of antidepressant medications were standardized into a common framework and included 9,042 adults with moderate to severe major depression. Feature selection retained 25 clinical and demographic variables. Using Bayesian optimization, a deep learning model was trained on the training set, refined using the validation set, and tested once on the held-out test set. RESULTS: In the evaluation on the held-out test set, the model demonstrated achieved an AUC of 0.65. The model outperformed a null model on the test set (p = 0.01). The model demonstrated clinical utility, achieving an absolute improvement in population remission rate in hypothetical and actual improvement testing. While the model did identify one drug (escitalopram) as generally outperforming the other drugs (consistent with the input data), there was otherwise significant variation in drug rankings. On bias testing, the model did not amplify potentially harmful biases. CONCLUSIONS: We demonstrate the first model capable of predicting outcomes for 10 different treatment options for patients with MDD, intended to be used at or near the start of treatment to personalize treatment. The model was put into clinical practice during the AIDME randomized controlled trial whose results are reported separately.

The inherent variability in the visual interpretation of cardiotocograms (CTGs) by obstetric clinical experts, both intra- and inter-observer, presents a substantial challenge in obstetric care. In response, we investigate automated CTG interpretation as a potential solution to enhance the early detection of fetal hypoxia during labor, which has the potential to reduce unnecessary operative interventions and improve overall maternal and neonatal care. This study employs deep learning techniques to reduce the subjectivity associated with visual CTG interpretation. Our results demonstrate that using objective umbilical cord blood pH outcome measurements, rather than clinician-defined Apgar scores, yields more consistent and robust model performance. Additionally, through a series of ablation studies, we explore the impact of temporal distribution shifts on the performance of these deep learning models. We examine tradeoffs between performance and fairness, specifically evaluating performance across demographic and clinical subgroups. Finally, we discuss the practical implications of our findings for the real-world deployment of such systems, emphasizing their potential utility in medical settings with limited resources. ### Competing Interest Statement All authors except N.C. are currently employed by Google, a for-profit entity which may stand to benefit financially from the models developed and presented in the paper. Authors have no other competing interests ### Funding Statement This study was funded by Google Research ### Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: The study used (or will use) ONLY openly available human data that were originally located at CTU-CHB Intrapartum Cardiotocography Database on Physionet: I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes We used an open source dataset found at: All additional data produced in the present study are available upon reasonable request to the authors

Prompting and in-context learning (ICL) have become efficient learning paradigms for large language models (LLMs). However, LLMs suffer from prompt brittleness and various bias factors in the prompt, including but not limited to the formatting, the choice verbalizers, and the ICL examples. To address this problem that results in unexpected performance degradation, calibration methods have been developed to mitigate the effects of these biases while recovering LLM performance. In this work, we first conduct a systematic analysis of the existing calibration methods, where we both provide a unified view and reveal the failure cases. Inspired by these analyses, we propose Batch Calibration (BC), a simple yet intuitive method that controls the contextual bias from the batched input, unifies various prior approaches, and effectively addresses the aforementioned issues. BC is zero-shot, inference-only, and incurs negligible additional costs. In the few-shot setup, we further extend BC to allow it to learn the contextual bias from labeled data. We validate the effectiveness of BC with PaLM 2-(S, M, L) and CLIP models and demonstrate state-of-the-art performance over previous calibration baselines across more than 10 natural language understanding and image classification tasks.

While large language models (LLMs) have shown promise for medical question answering, there is limited work focused on tropical and infectious disease-specific exploration. We build on an opensource tropical and infectious diseases (TRINDs) dataset, expanding it to include demographic and semantic clinical and consumer augmentations yielding 11000+ prompts. We evaluate LLM performance on these, comparing generalist and medical LLMs, as well as LLM outcomes to human experts. We demonstrate through systematic experimentation, the benefit of contextual information such as demographics, location, gender, risk factors for optimal LLM response. Finally we develop a prototype of TRINDs-LM, a research tool that provides a playground to navigate how context impacts LLM outputs for health.

Reward models play a key role in aligning language model applications towards human preferences. However, this setup creates an incentive for the language model to exploit errors in the reward model to achieve high estimated reward, a phenomenon often termed \emph{reward hacking}. A natural mitigation is to train an ensemble of reward models, aggregating over model outputs to obtain a more robust reward estimate. We explore the application of reward ensembles to alignment at both training time (through reinforcement learning) and inference time (through reranking). First, we show that reward models are \emph{underspecified}: reward models that perform similarly in-distribution can yield very different rewards when used in alignment, due to distribution shift. Second, underspecification results in overoptimization, where alignment to one reward model does not improve reward as measured by another reward model trained on the same data. Third, overoptimization is mitigated by the use of reward ensembles, and ensembles that vary by their \emph{pretraining} seeds lead to better generalization than ensembles that differ only by their \emph{fine-tuning} seeds, with both outperforming individual reward models. However, even pretrain reward ensembles do not eliminate reward hacking: we show several qualitative reward hacking phenomena that are not mitigated by ensembling because all reward models in the ensemble exhibit similar error patterns.

Multiple sclerosis is a disease that affects the brain and spinal cord, it can lead to severe disability and has no known cure. The majority of prior work in machine learning for multiple sclerosis has been centered around using Magnetic Resonance Imaging scans or laboratory tests; these modalities are both expensive to acquire and can be unreliable. In a recent paper it was shown that disease progression can be predicted effectively using performance outcome measures and demographic data. In our work we build on this to investigate the modeling side, using continuous time models to predict progression. We benchmark four continuous time models using a publicly available multiple sclerosis dataset. We find that the best continuous model is often able to outperform the best benchmarked discrete time model. We also carry out an extensive ablation to discover the sources of performance gains, we find that standardizing existing features leads to a larger performance increase than interpolating missing features.

With growing machine learning (ML) applications in healthcare, there have been calls for fairness in ML to understand and mitigate ethical concerns these systems may pose. Fairness has implications for global health in Africa, which already has inequitable power imbalances between the Global North and South. This paper seeks to explore fairness for global health, with Africa as a case study. We propose fairness attributes for consideration in the African context and delineate where they may come into play in different ML-enabled medical modalities. This work serves as a basis and call for action for furthering research into fairness in global health.