Large language models (LLMs) hold immense 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. In this work, 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 then conduct an empirical case study with Med-PaLM 2, resulting in the largest human evaluation study in this area to date. Our contributions include a multifactorial framework for human assessment of LLM-generated answers for biases, and EquityMedQA, a collection of seven newly-released datasets comprising both manually-curated and LLM-generated questions enriched for adversarial queries. Both our human assessment framework and dataset design process are grounded in an iterative participatory approach and review of possible biases in Med-PaLM 2 answers to adversarial queries. Through our empirical study, we find that the use of a collection of datasets curated through a variety of methodologies, coupled with a thorough evaluation protocol that leverages multiple assessment rubric designs and diverse rater groups, 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. We emphasize that while our framework can identify specific forms of bias, it is not sufficient to holistically assess whether the deployment of an AI system promotes equitable health outcomes. We hope the broader community leverages and builds on these tools and methods towards realizing a shared goal of LLMs that promote accessible and equitable healthcare for all.
Foundation large language models (LLMs) have shown an impressive ability to solve tasks across a wide range of fields including health. To effectively solve personalized health tasks, LLMs need the ability to ingest a diversity of data modalities that are relevant to an individual's health status. In this paper, we take a step towards creating multimodal LLMs for health that are grounded in individual-specific data by developing a framework (HeLM: Health Large Language Model for Multimodal Understanding) that enables LLMs to use high-dimensional clinical modalities to estimate underlying disease risk. HeLM encodes complex data modalities by learning an encoder that maps them into the LLM's token embedding space and for simple modalities like tabular data by serializing the data into text. Using data from the UK Biobank, we show that HeLM can effectively use demographic and clinical features in addition to high-dimensional time-series data to estimate disease risk. For example, HeLM achieves an AUROC of 0.75 for asthma prediction when combining tabular and spirogram data modalities compared with 0.49 when only using tabular data. Overall, we find that HeLM outperforms or performs at parity with classical machine learning approaches across a selection of eight binary traits. Furthermore, we investigate the downstream uses of this model such as its generalizability to out-of-distribution traits and its ability to power conversations around individual health and wellness.
Despite considerable progress in maternal healthcare, maternal and perinatal deaths remain high in low-to-middle income countries. Fetal ultrasound is an important component of antenatal care, but shortage of adequately trained healthcare workers has limited its adoption. We developed and validated an artificial intelligence (AI) system that uses novice-acquired "blind sweep" ultrasound videos to estimate gestational age (GA) and fetal malpresentation. We further addressed obstacles that may be encountered in low-resourced settings. Using a simplified sweep protocol with real-time AI feedback on sweep quality, we have demonstrated the generalization of model performance to minimally trained novice ultrasound operators using low cost ultrasound devices with on-device AI integration. The GA model was non-inferior to standard fetal biometry estimates with as few as two sweeps, and the fetal malpresentation model had high AUC-ROCs across operators and devices. Our AI models have the potential to assist in upleveling the capabilities of lightly trained ultrasound operators in low resource settings.
Model explanation techniques play a critical role in understanding the source of a model's performance and making its decisions transparent. Here we investigate if explanation techniques can also be used as a mechanism for scientific discovery. We make three contributions: first, we propose a framework to convert predictions from explanation techniques to a mechanism of discovery. Second, we show how generative models in combination with black-box predictors can be used to generate hypotheses (without human priors) that can be critically examined. Third, with these techniques we study classification models for retinal images predicting Diabetic Macular Edema (DME), where recent work showed that a CNN trained on these images is likely learning novel features in the image. We demonstrate that the proposed framework is able to explain the underlying scientific mechanism, thus bridging the gap between the model's performance and human understanding.
Colorectal Cancer (CRC) is a global health problem, resulting in 900K deaths per year. Colonoscopy is the tool of choice for preventing CRC, by detecting polyps before they become cancerous, and removing them. However, colonoscopy is hampered by the fact that endoscopists routinely miss an average of 22-28% of polyps. While some of these missed polyps appear in the endoscopist's field of view, others are missed simply because of substandard coverage of the procedure, i.e. not all of the colon is seen. This paper attempts to rectify the problem of substandard coverage in colonoscopy through the introduction of the C2D2 (Colonoscopy Coverage Deficiency via Depth) algorithm which detects deficient coverage, and can thereby alert the endoscopist to revisit a given area. More specifically, C2D2 consists of two separate algorithms: the first performs depth estimation of the colon given an ordinary RGB video stream; while the second computes coverage given these depth estimates. Rather than compute coverage for the entire colon, our algorithm computes coverage locally, on a segment-by-segment basis; C2D2 can then indicate in real-time whether a particular area of the colon has suffered from deficient coverage, and if so the endoscopist can return to that area. Our coverage algorithm is the first such algorithm to be evaluated in a large-scale way; while our depth estimation technique is the first calibration-free unsupervised method applied to colonoscopies. The C2D2 algorithm achieves state of the art results in the detection of deficient coverage: it is 2.4 times more accurate than human experts.
Much work aims to explain a model's prediction on a static input. We consider explanations in a temporal setting where a stateful dynamical model produces a sequence of risk estimates given an input at each time step. When the estimated risk increases, the goal of the explanation is to attribute the increase to a few relevant inputs from the past. While our formal setup and techniques are general, we carry out an in-depth case study in a clinical setting. The goal here is to alert a clinician when a patient's risk of deterioration rises. The clinician then has to decide whether to intervene and adjust the treatment. Given a potentially long sequence of new events since she last saw the patient, a concise explanation helps her to quickly triage the alert. We develop methods to lift static attribution techniques to the dynamical setting, where we identify and address challenges specific to dynamics. We then experimentally assess the utility of different explanations of clinical alerts through expert evaluation.
In a wide array of areas, algorithms are matching and surpassing the performance of human experts, leading to consideration of the roles of human judgment and algorithmic prediction in these domains. The discussion around these developments, however, has implicitly equated the specific task of prediction with the general task of automation. We argue here that automation is broader than just a comparison of human versus algorithmic performance on a task; it also involves the decision of which instances of the task to give to the algorithm in the first place. We develop a general framework that poses this latter decision as an optimization problem, and we show how basic heuristics for this optimization problem can lead to performance gains even on heavily-studied applications of AI in medicine. Our framework also serves to highlight how effective automation depends crucially on estimating both algorithmic and human error on an instance-by-instance basis, and our results show how improvements in these error estimation problems can yield significant gains for automation as well.
TensorFlow.js is a library for building and executing machine learning algorithms in JavaScript. TensorFlow.js models run in a web browser and in the Node.js environment. The library is part of the TensorFlow ecosystem, providing a set of APIs that are compatible with those in Python, allowing models to be ported between the Python and JavaScript ecosystems. TensorFlow.js has empowered a new set of developers from the extensive JavaScript community to build and deploy machine learning models and enabled new classes of on-device computation. This paper describes the design, API, and implementation of TensorFlow.js, and highlights some of the impactful use cases.
Predictive modeling with electronic health record (EHR) data is anticipated to drive personalized medicine and improve healthcare quality. Constructing predictive statistical models typically requires extraction of curated predictor variables from normalized EHR data, a labor-intensive process that discards the vast majority of information in each patient's record. We propose a representation of patients' entire, raw EHR records based on the Fast Healthcare Interoperability Resources (FHIR) format. We demonstrate that deep learning methods using this representation are capable of accurately predicting multiple medical events from multiple centers without site-specific data harmonization. We validated our approach using de-identified EHR data from two U.S. academic medical centers with 216,221 adult patients hospitalized for at least 24 hours. In the sequential format we propose, this volume of EHR data unrolled into a total of 46,864,534,945 data points, including clinical notes. Deep learning models achieved high accuracy for tasks such as predicting in-hospital mortality (AUROC across sites 0.93-0.94), 30-day unplanned readmission (AUROC 0.75-0.76), prolonged length of stay (AUROC 0.85-0.86), and all of a patient's final discharge diagnoses (frequency-weighted AUROC 0.90). These models outperformed state-of-the-art traditional predictive models in all cases. We also present a case-study of a neural-network attribution system, which illustrates how clinicians can gain some transparency into the predictions. We believe that this approach can be used to create accurate and scalable predictions for a variety of clinical scenarios, complete with explanations that directly highlight evidence in the patient's chart.