In some areas such as the medical field, ML-assisted predictions or decisions can drastically impact human life. For example, breast cancer [131] can be devastating if not diagnosed in time (or at all). The use of black-box predictors in these crucial cases has deceived more than once: a classical example of which is the use of the COMPAS system by the USA judiciary system for predicting criminal recidivism [133]. Other cases where fairness has been jeopardized by the use of black-boxes are numerous: job and loan applications biased toward men [40]; mortgage-approval biased toward white applicants [122]; higher credit card limits for men [172]; etc. With time, it became clear that interpretability is crucial when it comes to understanding how a predictor behaves and thus preventing unfortunate events; as pointed out by Goodman and Flaxman [70]: "If we do not know how ML [predictors] work, we cannot check or regulate them to ensure that they do not encode discrimination against minorities [...], we will not be able to learn from instances in which it is mistaken."

Foundation models hold promise for transforming AI in healthcare by providing modular components that are easily adaptable to downstream healthcare tasks, making AI development more scalable and cost-effective. Structured EHR foundation models, trained on coded medical records from millions of patients, demonstrated benefits including increased performance with fewer training labels, and improved robustness to distribution shifts. However, questions remain on the feasibility of sharing these models across different hospitals and their performance for local task adaptation. This multi-center study examined the adaptability of a recently released structured EHR foundation model ($FM_{SM}$), trained on longitudinal medical record data from 2.57M Stanford Medicine patients. Experiments were conducted using EHR data at The Hospital for Sick Children and MIMIC-IV. We assessed both adaptability via continued pretraining on local data, and task adaptability compared to baselines of training models from scratch at each site, including a local foundation model. We evaluated the performance of these models on 8 clinical prediction tasks. In both datasets, adapting the off-the-shelf $FM_{SM}$ matched the performance of GBM models locally trained on all data while providing a 13% improvement in settings with few task-specific training labels. With continued pretraining on local data, label efficiency substantially improved, such that $FM_{SM}$ required fewer than 1% of training examples to match the fully trained GBM's performance. Continued pretraining was also 60 to 90% more sample-efficient than training local foundation models from scratch. Our findings show that adapting shared EHR foundation models across hospitals provides improved prediction performance at less cost, underscoring the utility of base foundation models as modular components to streamline the development of healthcare AI.

Self-supervised learning for time-series data holds potential similar to that recently unleashed in Natural Language Processing and Computer Vision. While most existing works in this area focus on contrastive learning, we propose a conceptually simple yet powerful non-contrastive approach, based on the data2vec self-distillation framework. The core of our method is a student-teacher scheme that predicts the latent representation of an input time series from masked views of the same time series. This strategy avoids strong modality-specific assumptions and biases typically introduced by the design of contrastive sample pairs. We demonstrate the competitiveness of our approach for classification and forecasting as downstream tasks, comparing with state-of-the-art self-supervised learning methods on the UCR and UEA archives as well as the ETT and Electricity datasets.

Breast cancer diagnosis challenges both patients and clinicians, with early detection being crucial for effective treatment. Ultrasound imaging plays a key role in this, but its utility is hampered by the need for precise lesion segmentation-a task that is both time-consuming and labor-intensive. To address these challenges, we propose a new framework: a morphology-enhanced, Class Activation Map (CAM)-guided model, which is optimized using a computer vision foundation model known as SAM. This innovative framework is specifically designed for weakly supervised lesion segmentation in early-stage breast ultrasound images. Our approach uniquely leverages image-level annotations, which removes the requirement for detailed pixel-level annotation. Initially, we perform a preliminary segmentation using breast lesion morphology knowledge. Following this, we accurately localize lesions by extracting semantic information through a CAM-based heatmap. These two elements are then fused together, serving as a prompt to guide the SAM in performing refined segmentation. Subsequently, post-processing techniques are employed to rectify topological errors made by the SAM. Our method not only simplifies the segmentation process but also attains accuracy comparable to supervised learning methods that rely on pixel-level annotation. Our framework achieves a Dice score of 74.39% on the test set, demonstrating compareable performance with supervised learning methods. Additionally, it outperforms a supervised learning model, in terms of the Hausdorff distance, scoring 24.27 compared to Deeplabv3+'s 32.22. These experimental results showcase its feasibility and superior performance in integrating weakly supervised learning with SAM. The code is made available at: https://github.com/YueXin18/MorSeg-CAM-SAM.

Semi-supervised learning (SSL) methods effectively leverage unlabeled data to improve model generalization. However, SSL models often underperform in open-set scenarios, where unlabeled data contain outliers from novel categories that do not appear in the labeled set. In this paper, we study the challenging and realistic open-set SSL setting, where the goal is to both correctly classify inliers and to detect outliers. Intuitively, the inlier classifier should be trained on inlier data only. However, we find that inlier classification performance can be largely improved by incorporating high-confidence pseudo-labeled data, regardless of whether they are inliers or outliers. Also, we propose to utilize non-linear transformations to separate the features used for inlier classification and outlier detection in the multi-task learning framework, preventing adverse effects between them. Additionally, we introduce pseudo-negative mining, which further boosts outlier detection performance. The three ingredients lead to what we call Simple but Strong Baseline (SSB) for open-set SSL. In experiments, SSB greatly improves both inlier classification and outlier detection performance, outperforming existing methods by a large margin. Our code will be released at https://github.com/YUE-FAN/SSB.

AI-assisted nuclei segmentation in histopathological images is a crucial task in the diagnosis and treatment of cancer diseases. It decreases the time required to manually screen microscopic tissue images and can resolve the conflict between pathologists during diagnosis. Deep Learning has proven useful in such a task. However, lack of labeled data is a significant barrier for deep learning-based approaches. In this study, we propose a novel approach to nuclei segmentation that leverages the available labelled and unlabelled data. The proposed method combines the strengths of both transductive and inductive learning, which have been previously attempted separately, into a single framework. Inductive learning aims at approximating the general function and generalizing to unseen test data, while transductive learning has the potential of leveraging the unlabelled test data to improve the classification. To the best of our knowledge, this is the first study to propose such a hybrid approach for medical image segmentation. Moreover, we propose a novel two-stage transductive inference scheme. We evaluate our approach on MoNuSeg benchmark to demonstrate the efficacy and potential of our method.

Structured prediction problems are one of the fundamental tools in machine learning. In order to facilitate algorithm development for their numerical solution, we collect in one place a large number of datasets in easy to read formats for a diverse set of problem classes. We provide archival links to datasets, description of the considered problems and problem formats, and a short summary of problem characteristics including size, number of instances etc. For reference we also give a non-exhaustive selection of algorithms proposed in the literature for their solution. We hope that this central repository will make benchmarking and comparison to established works easier. We welcome submission of interesting new datasets and algorithms for inclusion in our archive.

In-context learning has been applied to knowledge-rich tasks such as question answering. In such scenarios, in-context examples are used to trigger a behaviour in the language model: namely, it should surface information stored in its parametric knowledge. We study the construction of in-context example sets, with a focus on the parametric knowledge of the model regarding in-context examples. We identify 'known' examples, where models can correctly answer from its parametric knowledge, and 'unknown' ones. Our experiments show that prompting with 'unknown' examples decreases the performance, potentially as it encourages hallucination rather than searching its parametric knowledge. Constructing an in-context example set that presents both known and unknown information performs the best across diverse settings. We perform analysis on three multi-answer question answering datasets, which allows us to further study answer set ordering strategies based on the LM's knowledge about each answer. Together, our study sheds lights on how to best construct in-context example sets for knowledge-rich tasks.

Sepsis is a syndrome that develops in the body in response to the presence of an infection. Characterized by severe organ dysfunction, sepsis is one of the leading causes of mortality in Intensive Care Units (ICUs) worldwide. These complications can be reduced through early application of antibiotics. Hence, the ability to anticipate the onset of sepsis early is crucial to the survival and well-being of patients. Current machine learning algorithms deployed inside medical infrastructures have demonstrated poor performance and are insufficient for anticipating sepsis onset early. Recently, deep learning methodologies have been proposed to predict sepsis, but some fail to capture the time of onset (e.g., classifying patients' entire visits as developing sepsis or not) and others are unrealistic for deployment in clinical settings (e.g., creating training instances using a fixed time to onset, where the time of onset needs to be known apriori). In this paper, we first propose a novel but realistic prediction framework that predicts each morning whether sepsis onset will occur within the next 24 hours using the most recent data collected the previous night, when patient-provider ratios are higher due to cross-coverage resulting in limited observation to each patient. However, as we increase the prediction rate into daily, the number of negative instances will increase, while that of positive instances remain the same. This causes a severe class imbalance problem making it hard to capture these rare sepsis cases. To address this, we propose a nightly profile representation learning (NPRL) approach. We prove that NPRL can theoretically alleviate the rare event problem and our empirical study using data from a level-1 trauma center demonstrates the effectiveness of our proposal.

Recognizing fallacies is crucial for ensuring the quality and validity of arguments across various domains. However, computational fallacy recognition faces challenges due to the diverse genres, domains, and types of fallacies found in datasets. This leads to a highly multiclass, and even multi-label, setup with substantial class imbalance. In this study, we aim to enhance existing models for fallacy recognition by incorporating additional context and by leveraging large language models to generate synthetic data, thus increasing the representation of the infrequent classes. We experiment with GPT3.5 to generate synthetic examples and we examine the impact of prompt settings for this. Moreover, we explore zero-shot and few-shot scenarios to evaluate the effectiveness of using the generated examples for training smaller models within a unified fallacy recognition framework. Furthermore, we analyze the overlap between the synthetic data and existing fallacy datasets. Finally, we investigate the usefulness of providing supplementary context for detecting fallacy types that need such context, e.g., diversion fallacies. Our evaluation results demonstrate consistent improvements across fallacy types, datasets, and generators.