During the diagnostic process, doctors incorporate multimodal information including imaging and the medical history - and similarly medical AI development has increasingly become multimodal. In this paper we tackle a more subtle challenge: doctors take a targeted medical history to obtain only the most pertinent pieces of information; how do we enable AI to do the same? We develop a wrapper method named MINT (Make your model INTeractive) that automatically determines what pieces of information are most valuable at each step, and ask for only the most useful information. We demonstrate the efficacy of MINT wrapping a skin disease prediction model, where multiple images and a set of optional answers to $25$ standard metadata questions (i.e., structured medical history) are used by a multi-modal deep network to provide a differential diagnosis. We show that MINT can identify whether metadata inputs are needed and if so, which question to ask next. We also demonstrate that when collecting multiple images, MINT can identify if an additional image would be beneficial, and if so, which type of image to capture. We showed that MINT reduces the number of metadata and image inputs needed by 82% and 36.2% respectively, while maintaining predictive performance. Using real-world AI dermatology system data, we show that needing fewer inputs can retain users that may otherwise fail to complete the system submission and drop off without a diagnosis. Qualitative examples show MINT can closely mimic the step-by-step decision making process of a clinical workflow and how this is different for straight forward cases versus more difficult, ambiguous cases. Finally we demonstrate how MINT is robust to different underlying multi-model classifiers and can be easily adapted to user requirements without significant model re-training.

Conformal Prediction (CP) allows to perform rigorous uncertainty quantification by constructing a prediction set $C(X)$ satisfying $\mathbb{P}(Y \in C(X))\geq 1-\alpha$ for a user-chosen $\alpha \in [0,1]$ by relying on calibration data $(X_1,Y_1),...,(X_n,Y_n)$ from $\mathbb{P}=\mathbb{P}^{X} \otimes \mathbb{P}^{Y|X}$. It is typically implicitly assumed that $\mathbb{P}^{Y|X}$ is the "true" posterior label distribution. However, in many real-world scenarios, the labels $Y_1,...,Y_n$ are obtained by aggregating expert opinions using a voting procedure, resulting in a one-hot distribution $\mathbb{P}_{vote}^{Y|X}$. For such ``voted'' labels, CP guarantees are thus w.r.t. $\mathbb{P}_{vote}=\mathbb{P}^X \otimes \mathbb{P}_{vote}^{Y|X}$ rather than the true distribution $\mathbb{P}$. In cases with unambiguous ground truth labels, the distinction between $\mathbb{P}_{vote}$ and $\mathbb{P}$ is irrelevant. However, when experts do not agree because of ambiguous labels, approximating $\mathbb{P}^{Y|X}$ with a one-hot distribution $\mathbb{P}_{vote}^{Y|X}$ ignores this uncertainty. In this paper, we propose to leverage expert opinions to approximate $\mathbb{P}^{Y|X}$ using a non-degenerate distribution $\mathbb{P}_{agg}^{Y|X}$. We develop Monte Carlo CP procedures which provide guarantees w.r.t. $\mathbb{P}_{agg}=\mathbb{P}^X \otimes \mathbb{P}_{agg}^{Y|X}$ by sampling multiple synthetic pseudo-labels from $\mathbb{P}_{agg}^{Y|X}$ for each calibration example $X_1,...,X_n$. In a case study of skin condition classification with significant disagreement among expert annotators, we show that applying CP w.r.t. $\mathbb{P}_{vote}$ under-covers expert annotations: calibrated for $72\%$ coverage, it falls short by on average $10\%$; our Monte Carlo CP closes this gap both empirically and theoretically.

For safety, AI systems in health undergo thorough evaluations before deployment, validating their predictions against a ground truth that is assumed certain. However, this is actually not the case and the ground truth may be uncertain. Unfortunately, this is largely ignored in standard evaluation of AI models but can have severe consequences such as overestimating the future performance. To avoid this, we measure the effects of ground truth uncertainty, which we assume decomposes into two main components: annotation uncertainty which stems from the lack of reliable annotations, and inherent uncertainty due to limited observational information. This ground truth uncertainty is ignored when estimating the ground truth by deterministically aggregating annotations, e.g., by majority voting or averaging. In contrast, we propose a framework where aggregation is done using a statistical model. Specifically, we frame aggregation of annotations as posterior inference of so-called plausibilities, representing distributions over classes in a classification setting, subject to a hyper-parameter encoding annotator reliability. Based on this model, we propose a metric for measuring annotation uncertainty and provide uncertainty-adjusted metrics for performance evaluation. We present a case study applying our framework to skin condition classification from images where annotations are provided in the form of differential diagnoses. The deterministic adjudication process called inverse rank normalization (IRN) from previous work ignores ground truth uncertainty in evaluation. Instead, we present two alternative statistical models: a probabilistic version of IRN and a Plackett-Luce-based model. We find that a large portion of the dataset exhibits significant ground truth uncertainty and standard IRN-based evaluation severely over-estimates performance without providing uncertainty estimates.

Reliable detection of out-of-distribution (OOD) inputs is increasingly understood to be a precondition for deployment of machine learning systems. This paper proposes and investigates the use of contrastive training to boost OOD detection performance. Unlike leading methods for OOD detection, our approach does not require access to examples labeled explicitly as OOD, which can be difficult to collect in practice. We show in extensive experiments that contrastive training significantly helps OOD detection performance on a number of common benchmarks. By introducing and employing the Confusion Log Probability (CLP) score, which quantifies the difficulty of the OOD detection task by capturing the similarity of inlier and outlier datasets, we show that our method especially improves performance in the `near OOD' classes -- a particularly challenging setting for previous methods.

Access to sufficient annotated data is a common challenge in training deep neural networks on medical images. As annotating data is expensive and time-consuming, it is difficult for an individual medical center to reach large enough sample sizes to build their own, personalized models. As an alternative, data from all centers could be pooled to train a centralized model that everyone can use. However, such a strategy is often infeasible due to the privacy-sensitive nature of medical data. Recently, federated learning (FL) has been introduced to collaboratively learn a shared prediction model across centers without the need for sharing data. In FL, clients are locally training models on site-specific datasets for a few epochs and then sharing their model weights with a central server, which orchestrates the overall training process. Importantly, the sharing of models does not compromise patient privacy. A disadvantage of FL is the dependence on a central server, which requires all clients to agree on one trusted central body, and whose failure would disrupt the training process of all clients. In this paper, we introduce BrainTorrent, a new FL framework without a central server, particularly targeted towards medical applications. BrainTorrent presents a highly dynamic peer-to-peer environment, where all centers directly interact with each other without depending on a central body. We demonstrate the overall effectiveness of FL for the challenging task of whole brain segmentation and observe that the proposed server-less BrainTorrent approach does not only outperform the traditional server-based one but reaches a similar performance to a model trained on pooled data.

In this paper we propose a novel augmentation technique that improves not only the performance of deep neural networks on clean test data, but also significantly increases their robustness to random transformations, both affine and projective. Inspired by ManiFool, the augmentation is performed by a line-search manifold-exploration method that learns affine geometric transformations that lead to the misclassification on an image, while ensuring that it remains on the same manifold as the training data. This augmentation method populates any training dataset with images that lie on the border of the manifolds between two-classes and maximizes the variance the network is exposed to during training. Our method was thoroughly evaluated on the challenging tasks of fine-grained skin lesion classification from limited data, and breast tumor classification of mammograms. Compared with traditional augmentation methods, and with images synthesized by Generative Adversarial Networks our method not only achieves state-of-the-art performance but also significantly improves the network's robustness.