The need for uncertainty quantification in machine-assisted medical decision making

A generalized deep learning framework for whole-slide image segmentation and analysis

Histopathology tissue analysis is considered the gold standard in cancer diagnosis and prognosis. Whole-slide imaging (WSI), i.e., the scanning and digitization of entire histology slides, are now being adopted across the world in pathology labs. Trained histopathologists can provide an accurate diagnosis of biopsy specimens based on WSI data. Given the dimensionality of WSIs and the increase in the number of potential cancer cases, analyzing these images is a time-consuming process. Automated segmentation of tumorous tissue helps in elevating the precision, speed, and reproducibility of research. In the recent past, deep learning-based techniques have provided state-of-the-art results in a wide variety of image analysis tasks, including the analysis of digitized slides. However, deep learning-based solutions pose many technical challenges, including the large size of WSI data, heterogeneity in images, and complexity of features. In this study, we propose a generalized deep learning-based framework for histopathology tissue analysis to address these challenges. Our framework is, in essence, a sequence of individual techniques in the preprocessing-training-inference pipeline which, in conjunction, improve the efficiency and the generalizability of the analysis. The combination of techniques we have introduced includes an ensemble segmentation model, division of the WSI into smaller overlapping patches while addressing class imbalances, efficient techniques for inference, and an efficient, patch-based uncertainty estimation framework. Our ensemble consists of DenseNet-121, Inception-ResNet-V2, and DeeplabV3Plus, where all the networks were trained end to end for every task. We demonstrate the efficacy and improved generalizability of our framework by evaluating it on a variety of histopathology tasks including breast cancer metastases (CAMELYON), colon cancer (DigestPath), and liver cancer (PAIP). Our proposed framework has state-of-the-art performance across all these tasks and is ranked within the top 5 currently for the challenges based on these datasets. The entire framework along with the trained models and the related documentation are made freely available at GitHub and PyPi. Our framework is expected to aid histopathologists in accurate and efficient initial diagnosis. Moreover, the estimated uncertainty maps will help clinicians to make informed decisions and further treatment planning or analysis.

Uncertainty quantification in classical molecular dynamics

Molecular dynamics simulation is now a widespread approach for understanding complex systems on the atomistic scale. It finds applications from physics and chemistry to engineering, life and medical science. In the last decade, the approach has begun to advance from being a computer-based means of rationalizing experimental observations to producing apparently credible predictions for a number of real-world applications within industrial sectors such as advanced materials and drug discovery. However, key aspects concerning the reproducibility of the method have not kept pace with the speed of its uptake in the scientific community. Here, we present a discussion of uncertainty quantification for molecular dynamics simulation designed to endow the method with better error estimates that will enable it to be used to report actionable results. The approach adopted is a standard one in the field of uncertainty quantification, namely using ensemble methods, in which a sufficiently large number of replicas are run concurrently, from which reliable statistics can be extracted. Indeed, because molecular dynamics is intrinsically chaotic, the need to use ensemble methods is fundamental and holds regardless of the duration of the simulations performed. We discuss the approach and illustrate it in a range of applications from materials science to ligand-protein binding free energy estimation. This article is part of the theme issue 'Reliability and reproducibility in computational science: implementing verification, validation and uncertainty quantification in silico'.

Artificial Intelligence in Cancer Research and Precision Medicine

Artificial intelligence (AI) is rapidly reshaping cancer research and personalized clinical care. Availability of high-dimensionality datasets coupled with advances in high-performance computing, as well as innovative deep learning architectures, has led to an explosion of AI use in various aspects of oncology research. These applications range from detection and classification of cancer, to molecular characterization of tumors and their microenvironment, to drug discovery and repurposing, to predicting treatment outcomes for patients. As these advances start penetrating the clinic, we foresee a shifting paradigm in cancer care becoming strongly driven by AI. SIGNIFICANCE: AI has the potential to dramatically affect nearly all aspects of oncology-from enhancing diagnosis to personalizing treatment and discovering novel anticancer drugs. Here, we review the recent enormous progress in the application of AI to oncology, highlight limitations and pitfalls, and chart a path for adoption of AI in the cancer clinic.

Threshold switching memristor-based stochastic neurons for probabilistic computing

Biological neurons exhibit dynamic excitation behavior in the form of stochastic firing, rather than stiffly giving out spikes upon reaching a fixed threshold voltage, which empowers the brain to perform probabilistic inference in the face of uncertainty. However, owing to the complexity of the stochastic firing process in biological neurons, the challenge of fabricating and applying stochastic neurons with bio-realistic dynamics to probabilistic scenarios remains to be fully addressed. In this work, a novel CuS/GeSe conductive-bridge threshold switching memristor is fabricated and singled out to realize electronic stochastic neurons, which is ascribed to the similarity between the stochastic switching behavior observed in the device and that of biological ion channels. The corresponding electric circuit of a stochastic neuron is then constructed and the probabilistic firing capacity of the neuron is utilized to implement Bayesian inference in a spiking neural network (SNN). The application prospects are demonstrated on the example of a tumor diagnosis task, where common fatal diagnostic errors of a conventional artificial neural network are successfully circumvented. Moreover, in comparison to deterministic neuron-based SNNs, the stochastic neurons enable SNNs to deliver an estimate of the uncertainty in their predictions, and the fidelity of the judgement is drastically improved by 81.2%.

Secure and Robust Machine Learning for Healthcare: A Survey

Recent years have witnessed widespread adoption of machine learning (ML)/deep learning (DL) techniques due to their superior performance for a variety of healthcare applications ranging from the prediction of cardiac arrest from one-dimensional heart signals to computer-aided diagnosis (CADx) using multi-dimensional medical images. Notwithstanding the impressive performance of ML/DL, there are still lingering doubts regarding the robustness of ML/DL in healthcare settings (which is traditionally considered quite challenging due to the myriad security and privacy issues involved), especially in light of recent results that have shown that ML/DL are vulnerable to adversarial attacks. In this paper, we present an overview of various application areas in healthcare that leverage such techniques from security and privacy point of view and present associated challenges. In addition, we present potential methods to ensure secure and privacy-preserving ML for healthcare applications. Finally, we provide insight into the current research challenges and promising directions for future research.

Machine Learning Against Terrorism: How Big Data Collection and Analysis Influences the Privacy-Security Dilemma

Rapid advancements in machine learning techniques allow mass surveillance to be applied on larger scales and utilize more and more personal data. These developments demand reconsideration of the privacy-security dilemma, which describes the tradeoffs between national security interests and individual privacy concerns. By investigating mass surveillance techniques that use bulk data collection and machine learning algorithms, we show why these methods are unlikely to pinpoint terrorists in order to prevent attacks. The diverse characteristics of terrorist attacks-especially when considering lone-wolf terrorism-lead to irregular and isolated (digital) footprints. The irregularity of data affects the accuracy of machine learning algorithms and the mass surveillance that depends on them which can be explained by three kinds of known problems encountered in machine learning theory: class imbalance, the curse of dimensionality, and spurious correlations. Proponents of mass surveillance often invoke the distinction between collecting data and metadata, in which the latter is understood as a lesser breach of privacy. Their arguments commonly overlook the ambiguity in the definitions of data and metadata and ignore the ability of machine learning techniques to infer the former from the latter. Given the sparsity of datasets used for machine learning in counterterrorism and the privacy risks attendant with bulk data collection, policymakers and other relevant stakeholders should critically re-evaluate the likelihood of success of the algorithms and the collection of data on which they depend.

Uncertainty quantification in drug design

Machine learning and artificial intelligence are increasingly being applied to the drug-design process as a result of the development of novel algorithms, growing access, the falling cost of computation and the development of novel technologies for generating chemically and biologically relevant data. There has been recent progress in fields such as molecular de novo generation, synthetic route prediction and, to some extent, property predictions. Despite this, most research in these fields has focused on improving the accuracy of the technologies, rather than on quantifying the uncertainty in the predictions. Uncertainty quantification will become a key component in autonomous decision making and will be crucial for integrating machine learning and chemistry automation to create an autonomous design-make-test-analyse cycle. This review covers the empirical, frequentist and Bayesian approaches to uncertainty quantification, and outlines how they can be used for drug design. We also outline the impact of uncertainty quantification on decision making.

Machine learning for metabolic engineering: A review

Machine learning provides researchers a unique opportunity to make metabolic engineering more predictable. In this review, we offer an introduction to this discipline in terms that are relatable to metabolic engineers, as well as providing in-depth illustrative examples leveraging omics data and improving production. We also include practical advice for the practitioner in terms of data management, algorithm libraries, computational resources, and important non-technical issues. A variety of applications ranging from pathway construction and optimization, to genetic editing optimization, cell factory testing, and production scale-up are discussed. Moreover, the promising relationship between machine learning and mechanistic models is thoroughly reviewed. Finally, the future perspectives and most promising directions for this combination of disciplines are examined.

Identifying Ethical Considerations for Machine Learning Healthcare Applications

Along with potential benefits to healthcare delivery, machine learning healthcare applications (ML-HCAs) raise a number of ethical concerns. Ethical evaluations of ML-HCAs will need to structure the overall problem of evaluating these technologies, especially for a diverse group of stakeholders. This paper outlines a systematic approach to identifying ML-HCA ethical concerns, starting with a conceptual model of the pipeline of the conception, development, implementation of ML-HCAs, and the parallel pipeline of evaluation and oversight tasks at each stage. Over this model, we layer key questions that raise value-based issues, along with ethical considerations identified in large part by a literature review, but also identifying some ethical considerations that have yet to receive attention. This pipeline model framework will be useful for systematic ethical appraisals of ML-HCA from development through implementation, and for interdisciplinary collaboration of diverse stakeholders that will be required to understand and subsequently manage the ethical implications of ML-HCAs.

Towards Knowledge Uncertainty Estimation for Open Set Recognition

Uncertainty is ubiquitous and happens in every single prediction of Machine Learning models. The ability to estimate and quantify the uncertainty of individual predictions is arguably relevant, all the more in safety-critical applications. Real-world recognition poses multiple challenges since a model’s knowledge about physical phenomenon is not complete, and observations are incomplete by definition. However, Machine Learning algorithms often assume that train and test data distributions are the same and that all testing classes are present during training. A more realistic scenario is the Open Set Recognition, where unknown classes can be submitted to an algorithm during testing. In this paper, we propose a Knowledge Uncertainty Estimation (KUE) method to quantify knowledge uncertainty and reject out-of-distribution inputs. Additionally, we quantify and distinguish aleatoric and epistemic uncertainty with the classical information-theoretical measures of entropy by means of ensemble techniques. We performed experiments on four datasets with different data modalities and compared our results with distance-based classifiers, SVM-based approaches and ensemble techniques using entropy measures. Overall, the effectiveness of KUE in distinguishing in- and out-distribution inputs obtained better results in most cases and was at least comparable in others. Furthermore, a classification with rejection option based on a proposed combination strategy between different measures of uncertainty is an application of uncertainty with proven results.

Beyond K-complex binary scoring during sleep: probabilistic classification using deep learning

STUDY OBJECTIVES

K-complexes (KCs) are a recognized electroencephalography marker of sensory processing and a defining feature of sleep stage 2. KC frequency and morphology may also be reflective of sleep quality, aging, and a range of sleep and sensory processing deficits. However, manual scoring of K-complexes is impractical, time-consuming, and thus costly and currently not well-standardized. Although automated KC detection methods have been developed, performance and uptake remain limited.

METHODS

The proposed algorithm is based on a deep neural network and Gaussian process, which gives the input waveform a probability of being a KC ranging from 0% to 100%. The algorithm was trained on half a million synthetic KCs derived from manually scored sleep stage 2 KCs from the Montreal Archive of Sleep Study containing 19 healthy young participants. Algorithm performance was subsequently assessed on 700 independent recordings from the Cleveland Family Study using sleep stages 2 and 3 data.

RESULTS

The developed algorithm showed an F1 score (a measure of binary classification accuracy) of 0.78 and thus outperforms currently available KC scoring algorithms with F1 = 0.2-0.6. The probabilistic approach also captured expected variability in KC shape and amplitude within individuals and across age groups.

CONCLUSIONS

An automated probabilistic KC classification is well suited and effective for systematic KC detection for a more in-depth exploration of potential relationships between KCs during sleep and clinical outcomes such as health impacts and daytime symptomatology.

Environmental Adaptation and Differential Replication in Machine Learning

When deployed in the wild, machine learning models are usually confronted with an environment that imposes severe constraints. As this environment evolves, so do these constraints. As a result, the feasible set of solutions for the considered need is prone to change in time. We refer to this problem as that of environmental adaptation. In this paper, we formalize environmental adaptation and discuss how it differs from other problems in the literature. We propose solutions based on differential replication, a technique where the knowledge acquired by the deployed models is reused in specific ways to train more suitable future generations. We discuss different mechanisms to implement differential replications in practice, depending on the considered level of knowledge. Finally, we present seven examples where the problem of environmental adaptation can be solved through differential replication in real-life applications.

A machine learning Automated Recommendation Tool for synthetic biology

Synthetic biology allows us to bioengineer cells to synthesize novel valuable molecules such as renewable biofuels or anticancer drugs. However, traditional synthetic biology approaches involve ad-hoc engineering practices, which lead to long development times. Here, we present the Automated Recommendation Tool (ART), a tool that leverages machine learning and probabilistic modeling techniques to guide synthetic biology in a systematic fashion, without the need for a full mechanistic understanding of the biological system. Using sampling-based optimization, ART provides a set of recommended strains to be built in the next engineering cycle, alongside probabilistic predictions of their production levels. We demonstrate the capabilities of ART on simulated data sets, as well as experimental data from real metabolic engineering projects producing renewable biofuels, hoppy flavored beer without hops, fatty acids, and tryptophan. Finally, we discuss the limitations of this approach, and the practical consequences of the underlying assumptions failing.

There Is Hope After All: Quantifying Opinion and Trustworthiness in Neural Networks

Artificial Intelligence (AI) plays a fundamental role in the modern world, especially when used as an autonomous decision maker. One common concern nowadays is "how trustworthy the AIs are." Human operators follow a strict educational curriculum and performance assessment that could be exploited to quantify how much we entrust them. To quantify the trust of AI decision makers, we must go beyond task accuracy especially when facing limited, incomplete, misleading, controversial or noisy datasets. Toward addressing these challenges, we describe DeepTrust, a Subjective Logic (SL) inspired framework that constructs a probabilistic logic description of an AI algorithm and takes into account the trustworthiness of both dataset and inner algorithmic workings. DeepTrust identifies proper multi-layered neural network (NN) topologies that have high projected trust probabilities, even when trained with untrusted data. We show that uncertain opinion of data is not always malicious while evaluating NN's opinion and trustworthiness, whereas the disbelief opinion hurts trust the most. Also trust probability does not necessarily correlate with accuracy. DeepTrust also provides a projected trust probability of NN's prediction, which is useful when the NN generates an over-confident output under problematic datasets. These findings open new analytical avenues for designing and improving the NN topology by optimizing opinion and trustworthiness, along with accuracy, in a multi-objective optimization formulation, subject to space and time constraints.

Improved Accuracy in Optical Diagnosis of Colorectal Polyps Using Convolutional Neural Networks with Visual Explanations

BACKGROUND & AIMS

Narrow-band imaging (NBI) can be used to determine whether colorectal polyps are adenomatous or hyperplastic. We investigated whether an artificial intelligence (AI) system can increase the accuracy of characterizations of polyps by endoscopists of different skill levels.

METHODS

We developed convolutional neural networks (CNNs) for evaluation of diminutive colorectal polyps, based on efficient neural architecture searches via parameter sharing with augmentation using NBIs of diminutive (≤5 mm) polyps, collected from October 2015 through October 2017 at the Seoul National University Hospital, Healthcare System Gangnam Center (training set). We trained the CNN using images from 1100 adenomatous polyps and 1050 hyperplastic polyps from 1379 patients. We then tested the system using 300 images of 180 adenomatous polyps and 120 hyperplastic polyps, obtained from January 2018 to May 2019. We compared the accuracy of 22 endoscopists of different skill levels (7 novices, 4 experts, and 11 NBI-trained experts) vs the CNN in evaluation of images (adenomatous vs hyperplastic) from 180 adenomatous and 120 hyperplastic polyps. The endoscopists then evaluated the polyp images with knowledge of the CNN-processed results. We conducted mixed-effect logistic and linear regression analyses to determine the effects of AI assistance on the accuracy of analysis of diminutive colorectal polyps by endoscopists (primary outcome).

RESULTS

The CNN distinguished adenomatous vs hyperplastic diminutive polyps with 86.7% accuracy, based on histologic analysis as the reference standard. Endoscopists distinguished adenomatous vs hyperplastic diminutive polyps with 82.5% overall accuracy (novices, 73.8% accuracy; experts, 83.8% accuracy; and NBI-trained experts, 87.6% accuracy). With knowledge of the CNN-processed results, the overall accuracy of the endoscopists increased to 88.5% (P < .05). With knowledge of the CNN-processed results, the accuracy of novice endoscopists increased to 85.6% (P < .05). The CNN-processed results significantly reduced endoscopist time of diagnosis (from 3.92 to 3.37 seconds per polyp, P = .042).

CONCLUSIONS

We developed a CNN that significantly increases the accuracy of evaluation of diminutive colorectal polyps (as adenomatous vs hyperplastic) and reduces the time of diagnosis by endoscopists. This AI assistance system significantly increased the accuracy of analysis by novice endoscopists, who achieved near-expert levels of accuracy without extra training. The CNN assistance system can reduce the skill-level dependence of endoscopists and costs.

Demystifying Brain Tumor Segmentation Networks: Interpretability and Uncertainty Analysis

The accurate automatic segmentation of gliomas and its intra-tumoral structures is important not only for treatment planning but also for follow-up evaluations. Several methods based on 2D and 3D Deep Neural Networks (DNN) have been developed to segment brain tumors and to classify different categories of tumors from different MRI modalities. However, these networks are often black-box models and do not provide any evidence regarding the process they take to perform this task. Increasing transparency and interpretability of such deep learning techniques is necessary for the complete integration of such methods into medical practice. In this paper, we explore various techniques to explain the functional organization of brain tumor segmentation models and to extract visualizations of internal concepts to understand how these networks achieve highly accurate tumor segmentations. We use the BraTS 2018 dataset to train three different networks with standard architectures and outline similarities and differences in the process that these networks take to segment brain tumors. We show that brain tumor segmentation networks learn certain human-understandable disentangled concepts on a filter level. We also show that they take a top-down or hierarchical approach to localizing the different parts of the tumor. We then extract visualizations of some internal feature maps and also provide a measure of uncertainty with regards to the outputs of the models to give additional qualitative evidence about the predictions of these networks. We believe that the emergence of such human-understandable organization and concepts might aid in the acceptance and integration of such methods in medical diagnosis.

Augmented decision-making for acral lentiginous melanoma detection using deep convolutional neural networks

BACKGROUND

Several studies have achieved high-level performance of melanoma detection using convolutional neural networks (CNNs). However, few have described the extent to which the implementation of CNNs improves the diagnostic performance of the physicians.

OBJECTIVE

This study is aimed at developing a CNN for detecting acral lentiginous melanoma (ALM) and investigating whether its implementation can improve the initial decision for ALM detection made by the physicians.

METHODS

A CNN was trained using 1072 dermoscopic images of acral benign nevi, ALM and intermediate tumours. To investigate whether the implementation of CNN can improve the initial decision for ALM detection, 60 physicians completed a three-stage survey. In Stage I, they were asked for their decisions solely on the basis of dermoscopic images provided to them. In Stage II, they were also provided with clinical information. In Stage III, they were provided with the additional diagnosis and probability predicted by the CNN.

RESULTS

The accuracy of ALM detection in the participants was 74.7% (95% confidence interval [CI], 72.6-76.8%) in Stage I and 79.0% (95% CI, 76.7-81.2%) in Stage II. In Stage III, it was 86.9% (95% CI, 85.3-88.4%), which exceeds the accuracy delivered in Stage I by 12.2%p (95% CI, 10.1-14.3%p) and Stage II by 7.9%p (95% CI, 6.0-9.9%p). Moreover, the concordance between the participants considerably increased (Fleiss-κ of 0.436 [95% CI, 0.437-0.573] in Stage I, 0.506 [95% CI, 0.621-0.749] in Stage II and 0.684 [95% CI, 0.621-0.749] in Stage III).

CONCLUSIONS

Augmented decision-making improved the performance of and concordance between the clinical decisions of a diverse group of experts. This study demonstrates the potential use of CNNs as an adjoining, decision-supporting system for physicians' decisions.

Interpretable multimodal deep learning for real-time pan-tissue pan-disease pathology search on social media

Pathologists are responsible for rapidly providing a diagnosis on critical health issues. Challenging cases benefit from additional opinions of pathologist colleagues. In addition to on-site colleagues, there is an active worldwide community of pathologists on social media for complementary opinions. Such access to pathologists worldwide has the capacity to improve diagnostic accuracy and generate broader consensus on next steps in patient care. From Twitter we curate 13,626 images from 6,351 tweets from 25 pathologists from 13 countries. We supplement the Twitter data with 113,161 images from 1,074,484 PubMed articles. We develop machine learning and deep learning models to (i) accurately identify histopathology stains, (ii) discriminate between tissues, and (iii) differentiate disease states. Area Under Receiver Operating Characteristic (AUROC) is 0.805-0.996 for these tasks. We repurpose the disease classifier to search for similar disease states given an image and clinical covariates. We report precision@k = 1 = 0.7618 ± 0.0018 (chance 0.397 ± 0.004, mean ±stdev ). The classifiers find that texture and tissue are important clinico-visual features of disease. Deep features trained only on natural images (e.g., cats and dogs) substantially improved search performance, while pathology-specific deep features and cell nuclei features further improved search to a lesser extent. We implement a social media bot (@pathobot on Twitter) to use the trained classifiers to aid pathologists in obtaining real-time feedback on challenging cases. If a social media post containing pathology text and images mentions the bot, the bot generates quantitative predictions of disease state (normal/artifact/infection/injury/nontumor, preneoplastic/benign/low-grade-malignant-potential, or malignant) and lists similar cases across social media and PubMed. Our project has become a globally distributed expert system that facilitates pathological diagnosis and brings expertise to underserved regions or hospitals with less expertise in a particular disease. This is the first pan-tissue pan-disease (i.e., from infection to malignancy) method for prediction and search on social media, and the first pathology study prospectively tested in public on social media. We will share data through http://pathobotology.org . We expect our project to cultivate a more connected world of physicians and improve patient care worldwide.

Ethical considerations about artificial intelligence for prognostication in intensive care

Background

Prognosticating the course of diseases to inform decision-making is a key component of intensive care medicine. For several applications in medicine, new methods from the field of artificial intelligence (AI) and machine learning have already outperformed conventional prediction models. Due to their technical characteristics, these methods will present new ethical challenges to the intensivist.

Results

In addition to the standards of data stewardship in medicine, the selection of datasets and algorithms to create AI prognostication models must involve extensive scrutiny to avoid biases and, consequently, injustice against individuals or groups of patients. Assessment of these models for compliance with the ethical principles of beneficence and non-maleficence should also include quantification of predictive uncertainty. Respect for patients’ autonomy during decision-making requires transparency of the data processing by AI models to explain the predictions derived from these models. Moreover, a system of continuous oversight can help to maintain public trust in this technology. Based on these considerations as well as recent guidelines, we propose a pathway to an ethical implementation of AI-based prognostication. It includes a checklist for new AI models that deals with medical and technical topics as well as patient- and system-centered issues.

Conclusion

AI models for prognostication will become valuable tools in intensive care. However, they require technical refinement and a careful implementation according to the standards of medical ethics.

Unlocking the Power of Artificial Intelligence and Big Data in Medicine

Data-driven science and its corollaries in machine learning and the wider field of artificial intelligence have the potential to drive important changes in medicine. However, medicine is not a science like any other: It is deeply and tightly bound with a large and wide network of legal, ethical, regulatory, economical, and societal dependencies. As a consequence, the scientific and technological progresses in handling information and its further processing and cross-linking for decision support and predictive systems must be accompanied by parallel changes in the global environment, with numerous stakeholders, including citizen and society. What can be seen at the first glance as a barrier and a mechanism slowing down the progression of data science must, however, be considered an important asset. Only global adoption can transform the potential of big data and artificial intelligence into an effective breakthroughs in handling health and medicine. This requires science and society, scientists and citizens, to progress together.

AI Meets Exascale Computing: Advancing Cancer Research With Large-Scale High Performance Computing

The application of data science in cancer research has been boosted by major advances in three primary areas: (1) Data: diversity, amount, and availability of biomedical data; (2) Advances in Artificial Intelligence (AI) and Machine Learning (ML) algorithms that enable learning from complex, large-scale data; and (3) Advances in computer architectures allowing unprecedented acceleration of simulation and machine learning algorithms. These advances help build in silico ML models that can provide transformative insights from data including: molecular dynamics simulations, next-generation sequencing, omics, imaging, and unstructured clinical text documents. Unique challenges persist, however, in building ML models related to cancer, including: (1) access, sharing, labeling, and integration of multimodal and multi-institutional data across different cancer types; (2) developing AI models for cancer research capable of scaling on next generation high performance computers; and (3) assessing robustness and reliability in the AI models. In this paper, we review the National Cancer Institute (NCI) -Department of Energy (DOE) collaboration, Joint Design of Advanced Computing Solutions for Cancer (JDACS4C), a multi-institution collaborative effort focused on advancing computing and data technologies to accelerate cancer research on three levels: molecular, cellular, and population. This collaboration integrates various types of generated data, pre-exascale compute resources, and advances in ML models to increase understanding of basic cancer biology, identify promising new treatment options, predict outcomes, and eventually prescribe specialized treatments for patients with cancer.

A Bayesian graph convolutional network for reliable prediction of molecular properties with uncertainty quantification

Deep neural networks have been increasingly used in various chemical fields. In the nature of a data-driven approach, their performance strongly depends on data used in training. Therefore, models developed in data-deficient situations can cause highly uncertain predictions, leading to vulnerable decision making. Here, we show that Bayesian inference enables more reliable prediction with quantitative uncertainty analysis. Decomposition of the predictive uncertainty into model- and data-driven uncertainties allows us to elucidate the source of errors for further improvements. For molecular applications, we devised a Bayesian graph convolutional network (GCN) and evaluated its performance for molecular property predictions. Our study on the classification problem of bio-activity and toxicity shows that the confidence of prediction can be quantified in terms of the predictive uncertainty, leading to more accurate virtual screening of drug candidates than standard GCNs. The result of log P prediction illustrates that data noise affects the data-driven uncertainty more significantly than the model-driven one. Based on this finding, we could identify artefacts that arose from quantum mechanical calculations in the Harvard Clean Energy Project dataset. Consequently, the Bayesian GCN is critical for molecular applications under data-deficient conditions.

The Bionic Radiologist: avoiding blurry pictures and providing greater insights

Radiology images and reports have long been digitalized. However, the potential of the more than 3.6 billion radiology examinations performed annually worldwide has largely gone unused in the effort to digitally transform health care. The Bionic Radiologist is a concept that combines humanity and digitalization for better health care integration of radiology. At a practical level, this concept will achieve critical goals: (1) testing decisions being made scientifically on the basis of disease probabilities and patient preferences; (2) image analysis done consistently at any time and at any site; and (3) treatment suggestions that are closely linked to imaging results and are seamlessly integrated with other information. The Bionic Radiologist will thus help avoiding missed care opportunities, will provide continuous learning in the work process, and will also allow more time for radiologists' primary roles: interacting with patients and referring physicians. To achieve that potential, one has to cope with many implementation barriers at both the individual and institutional levels. These include: reluctance to delegate decision making, a possible decrease in image interpretation knowledge and the perception that patient safety and trust are at stake. To facilitate implementation of the Bionic Radiologist the following will be helpful: uncertainty quantifications for suggestions, shared decision making, changes in organizational culture and leadership style, maintained expertise through continuous learning systems for training, and role development of the involved experts. With the support of the Bionic Radiologist, disparities are reduced and the delivery of care is provided in a humane and personalized fashion.

Accurate Hit Estimation for Iterative Screening Using Venn-ABERS Predictors

Iterative screening has emerged as a promising approach to increase the efficiency of high-throughput screening (HTS) campaigns in drug discovery. By learning from a subset of the compound library, inferences on what compounds to screen next can be made by predictive models. One of the challenges of iterative screening is to decide how many iterations to perform. This is mainly related to difficulties in estimating the prospective hit rate in any given iteration. In this article, a novel method based on Venn-ABERS predictors is proposed. The method provides accurate estimates of the number of hits retrieved in any given iteration during an HTS campaign. The estimates provide the necessary information to support the decision on the number of iterations needed to maximize the screening outcome. Thus, this method offers a prospective screening strategy for early-stage drug discovery.