Visualization for Diagnostic Review of Copy Number Variants in Complex DNA Sequencing Data.

Genomics is at the core of precision medicine, and there are high expectations on genomics-enabled improvement of patient outcomes in the years to come. Around the world, initiatives to increase the use of DNA sequencing in clinical routine are being deployed, such as the use of broad panels in the standard care for oncology patients. Such a development comes at the cost of increased demands on throughput in genomic data analysis. In this paper, we use the task of copy number variant (CNV) analysis as a context for exploring visualization concepts for clinical genomics. CNV calls are generated algorithmically, but time-consuming manual intervention is needed to separate relevant findings from irrelevant ones in the resulting large call candidate lists. We present a visualization environment, named Copycat, to support this review task in a clinical scenario. Key components are a scatter-glyph plot replacing the traditional list visualization, and a glyph representation designed for at-a-glance relevance assessments. Moreover, we present results from a formative evaluation of the prototype by domain specialists, from which we elicit insights to guide both prototype improvements and visualization for clinical genomics in general.

Benefits of spatial uncertainty aggregation for segmentation in digital pathology.

Purpose: Uncertainty estimation has gained significant attention in recent years for its potential to enhance the performance of deep learning (DL) algorithms in medical applications and even potentially address domain shift challenges. However, it is not straightforward to incorporate uncertainty estimation with a DL system to achieve a tangible positive effect. The objective of our work is to evaluate if the proposed spatial uncertainty aggregation (SUA) framework may improve the effectiveness of uncertainty estimation in segmentation tasks. We evaluate if SUA boosts the observed correlation between the uncertainty estimates and false negative (FN) predictions. We also investigate if the observed benefits can translate to tangible improvements in segmentation performance. Approach: Our SUA framework processes negative prediction regions from a segmentation algorithm and detects FNs based on an aggregated uncertainty score. It can be utilized with many existing uncertainty estimation methods to boost their performance. We compare the SUA framework with a baseline of processing individual pixel's uncertainty independently. Results: The results demonstrate that SUA is able to detect FN regions. It achieved Fß=0.5 of 0.92 on the in-domain and 0.85 on the domain-shift test data compared with 0.81 and 0.48 achieved by the baseline uncertainty, respectively. We also demonstrate that SUA yields improved general segmentation performance compared with utilizing the baseline uncertainty. Conclusions: We propose the SUA framework for incorporating and utilizing uncertainty estimates for FN detection in DL segmentation algorithms for histopathology. The evaluation confirms the benefits of our approach compared with assessing pixel uncertainty independently.

The thorny complexities of visualization research for clinical settings: A case study from genomics.

In this perspective article we discuss a certain type of research on visualization for bioinformatics data, namely, methods targeting clinical use. We argue that in this subarea additional complex challenges come into play, particularly so in genomics. We here describe four such challenge areas, elicited from a domain characterization effort in clinical genomics. We also list opportunities for visualization research to address clinical challenges in genomics that were uncovered in the case study. The findings are shown to have parallels with experiences from the diagnostic imaging domain.

VAI-B: a multicenter platform for the external validation of artificial intelligence algorithms in breast imaging.

Purpose: Multiple vendors are currently offering artificial intelligence (AI) computer-aided systems for triage detection, diagnosis, and risk prediction of breast cancer based on screening mammography. There is an imminent need to establish validation platforms that enable fair and transparent testing of these systems against external data. Approach: We developed validation of artificial intelligence for breast imaging (VAI-B), a platform for independent validation of AI algorithms in breast imaging. The platform is a hybrid solution, with one part implemented in the cloud and another in an on-premises environment at Karolinska Institute. Cloud services provide the flexibility of scaling the computing power during inference time, while secure on-premises clinical data storage preserves their privacy. A MongoDB database and a python package were developed to store and manage the data on-premises. VAI-B requires four data components: radiological images, AI inferences, radiologist assessments, and cancer outcomes. Results: To pilot test VAI-B, we defined a case-control population based on 8080 patients diagnosed with breast cancer and 36,339 healthy women based on the Swedish national quality registry for breast cancer. Images and radiological assessments from more than 100,000 mammography examinations were extracted from hospitals in three regions of Sweden. The images were processed by AI systems from three vendors in a virtual private cloud to produce abnormality scores related to signs of cancer in the images. A total of 105,706 examinations have been processed and stored in the database. Conclusions: We have created a platform that will allow downstream evaluation of AI systems for breast cancer detection, which enables faster development cycles for participating vendors and safer AI adoption for participating hospitals. The platform was designed to be scalable and ready to be expanded should a new vendor want to evaluate their system or should a new hospital wish to obtain an evaluation of different AI systems on their images.

Mapping the Landscape of Care Providers' Quality Assurance Approaches for AI in Diagnostic Imaging.

The discussion on artificial intelligence (AI) solutions in diagnostic imaging has matured in recent years. The potential value of AI adoption is well established, as are the potential risks associated. Much focus has, rightfully, been on regulatory certification of AI products, with the strong incentive of being an enabling step for the commercial actors. It is, however, becoming evident that regulatory approval is not enough to ensure safe and effective AI usage in the local setting. In other words, care providers need to develop and implement quality assurance (QA) approaches for AI solutions in diagnostic imaging. The domain of AI-specific QA is still in an early development phase. We contribute to this development by describing the current landscape of QA-for-AI approaches in medical imaging, with focus on radiology and pathology. We map the potential quality threats and review the existing QA approaches in relation to those threats. We propose a practical categorization of QA approaches, based on key characteristics corresponding to means, situation, and purpose. The review highlights the heterogeneity of methods and practices relevant for this domain and points to targets for future research efforts.

Generalization of Deep Learning in Digital Pathology: Experience in Breast Cancer Metastasis Detection.

Poor generalizability is a major barrier to clinical implementation of artificial intelligence in digital pathology. The aim of this study was to test the generalizability of a pretrained deep learning model to a new diagnostic setting and to a small change in surgical indication. A deep learning model for breast cancer metastases detection in sentinel lymph nodes, trained on CAMELYON multicenter data, was used as a base model, and achieved an AUC of 0.969 (95% CI 0.926-0.998) and FROC of 0.838 (95% CI 0.757-0.913) on CAMELYON16 test data. On local sentinel node data, the base model performance dropped to AUC 0.929 (95% CI 0.800-0.998) and FROC 0.744 (95% CI 0.566-0.912). On data with a change in surgical indication (axillary dissections) the base model performance indicated an even larger drop with a FROC of 0.503 (95%CI 0.201-0.911). The model was retrained with addition of local data, resulting in about a 4% increase for both AUC and FROC for sentinel nodes, and an increase of 11% in AUC and 49% in FROC for axillary nodes. Pathologist qualitative evaluation of the retrained model´s output showed no missed positive slides. False positives, false negatives and one previously undetected micro-metastasis were observed. The study highlights the generalization challenge even when using a multicenter trained model, and that a small change in indication can considerably impact the model´s performance.

Generalisation effects of predictive uncertainty estimation in deep learning for digital pathology.

Deep learning (DL) has shown great potential in digital pathology applications. The robustness of a diagnostic DL-based solution is essential for safe clinical deployment. In this work we evaluate if adding uncertainty estimates for DL predictions in digital pathology could result in increased value for the clinical applications, by boosting the general predictive performance or by detecting mispredictions. We compare the effectiveness of model-integrated methods (MC dropout and Deep ensembles) with a model-agnostic approach (Test time augmentation, TTA). Moreover, four uncertainty metrics are compared. Our experiments focus on two domain shift scenarios: a shift to a different medical center and to an underrepresented subtype of cancer. Our results show that uncertainty estimates increase reliability by reducing a model's sensitivity to classification threshold selection as well as by detecting between 70 and 90% of the mispredictions done by the model. Overall, the deep ensembles method achieved the best performance closely followed by TTA.

Presenting artificial intelligence, deep learning, and machine learning studies to clinicians and healthcare stakeholders: an introductory reference with a guideline and a Clinical AI Research (CAIR) checklist proposal.

Background and purpose - Artificial intelligence (AI), deep learning (DL), and machine learning (ML) have become common research fields in orthopedics and medicine in general. Engineers perform much of the work. While they gear the results towards healthcare professionals, the difference in competencies and goals creates challenges for collaboration and knowledge exchange. We aim to provide clinicians with a context and understanding of AI research by facilitating communication between creators, researchers, clinicians, and readers of medical AI and ML research.Methods and results - We present the common tasks, considerations, and pitfalls (both methodological and ethical) that clinicians will encounter in AI research. We discuss the following topics: labeling, missing data, training, testing, and overfitting. Common performance and outcome measures for various AI and ML tasks are presented, including accuracy, precision, recall, F1 score, Dice score, the area under the curve, and ROC curves. We also discuss ethical considerations in terms of privacy, fairness, autonomy, safety, responsibility, and liability regarding data collecting or sharing.Interpretation - We have developed guidelines for reporting medical AI research to clinicians in the run-up to a broader consensus process. The proposed guidelines consist of a Clinical Artificial Intelligence Research (CAIR) checklist and specific performance metrics guidelines to present and evaluate research using AI components. Researchers, engineers, clinicians, and other stakeholders can use these proposal guidelines and the CAIR checklist to read, present, and evaluate AI research geared towards a healthcare setting.

The human-in-the-loop: an evaluation of pathologists' interaction with artificial intelligence in clinical practice.

AIMS: One of the major drivers of the adoption of digital pathology in clinical practice is the possibility of introducing digital image analysis (DIA) to assist with diagnostic tasks. This offers potential increases in accuracy, reproducibility, and efficiency. Whereas stand-alone DIA has great potential benefit for research, little is known about the effect of DIA assistance in clinical use. The aim of this study was to investigate the clinical use characteristics of a DIA application for Ki67 proliferation assessment. Specifically, the human-in-the-loop interplay between DIA and pathologists was studied. METHODS AND RESULTS: We retrospectively investigated breast cancer Ki67 areas assessed with human-in-the-loop DIA and compared them with visual and automatic approaches. The results, expressed as standard deviation of the error in the Ki67 index, showed that visual estimation ('eyeballing') (14.9 percentage points) performed significantly worse (P < 0.05) than DIA alone (7.2 percentage points) and DIA with human-in-the-loop corrections (6.9 percentage points). At the overall level, no improvement resulting from the addition of human-in-the-loop corrections to the automatic DIA results could be seen. For individual cases, however, human-in-the-loop corrections could address major DIA errors in terms of poor thresholding of faint staining and incorrect tumour-stroma separation. CONCLUSION: The findings indicate that the primary value of human-in-the-loop corrections is to address major weaknesses of a DIA application, rather than fine-tuning the DIA quantifications.

Proactive Construction of an Annotated Imaging Database for Artificial Intelligence Training.

Artificial intelligence (AI) holds much promise for enabling highly desired imaging diagnostics improvements. One of the most limiting bottlenecks for the development of useful clinical-grade AI models is the lack of training data. One aspect is the large amount of cases needed and another is the necessity of high-quality ground truth annotation. The aim of the project was to establish and describe the construction of a database with substantial amounts of detail-annotated oncology imaging data from pathology and radiology. A specific objective was to be proactive, that is, to support undefined subsequent AI training across a wide range of tasks, such as detection, quantification, segmentation, and classification, which puts particular focus on the quality and generality of the annotations. The main outcome of this project was the database as such, with a collection of labeled image data from breast, ovary, skin, colon, skeleton, and liver. In addition, this effort also served as an exploration of best practices for further scalability of high-quality image collections, and a main contribution of the study was generic lessons learned regarding how to successfully organize efforts to construct medical imaging databases for AI training, summarized as eight guiding principles covering team, process, and execution aspects.

Measuring Domain Shift for Deep Learning in Histopathology.

The high capacity of neural networks allows fitting models to data with high precision, but makes generalization to unseen data a challenge. If a domain shift exists, i.e. differences in image statistics between training and test data, care needs to be taken to ensure reliable deployment in real-world scenarios. In digital pathology, domain shift can be manifested in differences between whole-slide images, introduced by for example differences in acquisition pipeline - between medical centers or over time. In order to harness the great potential presented by deep learning in histopathology, and ensure consistent model behavior, we need a deeper understanding of domain shift and its consequences, such that a model's predictions on new data can be trusted. This work focuses on the internal representation learned by trained convolutional neural networks, and shows how this can be used to formulate a novel measure - the representation shift - for quantifying the magnitude of model-specific domain shift. We perform a study on domain shift in tumor classification of hematoxylin and eosin stained images, by considering different datasets, models, and techniques for preparing data in order to reduce the domain shift. The results show how the proposed measure has a high correlation with drop in performance when testing a model across a large number of different types of domain shifts, and how it improves on existing techniques for measuring data shift and uncertainty. The proposed measure can reveal how sensitive a model is to domain variations, and can be used to detect new data that a model will have problems generalizing to. We see techniques for measuring, understanding and overcoming the domain shift as a crucial step towards reliable use of deep learning in the future clinical pathology applications.

TissueWand, a Rapid Histopathology Annotation Tool.

BACKGROUND: Recent advancements in machine learning (ML) bring great possibilities for the development of tools to assist with diagnostic tasks within histopathology. However, these approaches typically require a large amount of ground truth training data in the form of image annotations made by human experts. As such annotation work is a very time-consuming task, there is a great need for tools that can assist in this process, saving time while not sacrificing annotation quality. METHODS: In an iterative design process, we developed TissueWand - an interactive tool designed for efficient annotation of gigapixel-sized histopathological images, not being constrained to a predefined annotation task. RESULTS: Several findings regarding appropriate interaction concepts were made, where a key design component was semi-automation based on rapid interaction feedback in a local region. In a user study, the resulting tool was shown to cause substantial speed-up compared to manual work while maintaining quality. CONCLUSIONS: The TissueWand tool shows promise to replace manual methods for early stages of dataset curation where no task-specific ML model yet exists to aid the effort.

Annotations, Ontologies, and Whole Slide Images - Development of an Annotated Ontology-Driven Whole Slide Image Library of Normal and Abnormal Human Tissue.

OBJECTIVE: Digital pathology is today a widely used technology, and the digitalization of microscopic slides into whole slide images (WSIs) allows the use of machine learning algorithms as a tool in the diagnostic process. In recent years, "deep learning" algorithms for image analysis have been applied to digital pathology with great success. The training of these algorithms requires a large volume of high-quality images and image annotations. These large image collections are a potent source of information, and to use and share the information, standardization of the content through a consistent terminology is essential. The aim of this project was to develop a pilot dataset of exhaustive annotated WSI of normal and abnormal human tissue and link the annotations to appropriate ontological information. MATERIALS AND METHODS: Several biomedical ontologies and controlled vocabularies were investigated with the aim of selecting the most suitable ontology for this project. The selection criteria required an ontology that covered anatomical locations, histological subcompartments, histopathologic diagnoses, histopathologic terms, and generic terms such as normal, abnormal, and artifact. WSIs of normal and abnormal tissue from 50 colon resections and 69 skin excisions, diagnosed 2015-2016 at the Department of Clinical Pathology in Linköping, were randomly collected. These images were manually and exhaustively annotated at the level of major subcompartments, including normal or abnormal findings and artifacts. RESULTS: Systemized nomenclature of medicine clinical terms (SNOMED CT) was chosen, and the annotations were linked to its codes and terms. Two hundred WSI were collected and annotated, resulting in 17,497 annotations, covering a total area of 302.19 cm2, equivalent to 107,7 gigapixels. Ninety-five unique SNOMED CT codes were used. The time taken to annotate a WSI varied from 45 s to over 360 min, a total time of approximately 360 h. CONCLUSION: This work resulted in a dataset of 200 exhaustive annotated WSIs of normal and abnormal tissue from the colon and skin, and it has informed plans to build a comprehensive library of annotated WSIs. SNOMED CT was found to be the best ontology for annotation labeling. This project also demonstrates the need for future development of annotation tools in order to make the annotation process more efficient.

Deep learning nuclei detection: A simple approach can deliver state-of-the-art results.

BACKGROUND: Deep convolutional neural networks have become a widespread tool for the detection of nuclei in histopathology images. Many implementations share a basic approach that includes generation of an intermediate map indicating the presence of a nucleus center, which we refer to as PMap. Nevertheless, these implementations often still differ in several parameters, resulting in different detection qualities. METHODS: We identified several essential parameters and configured the basic PMap approach using combinations of them. We thoroughly evaluated and compared various configurations on multiple datasets with respect to detection quality, efficiency and training effort. RESULTS: Post-processing of the PMap was found to have the largest impact on detection quality. Also, two different network architectures were identified that improve either detection quality or runtime performance. The best-performing configuration yields f1-measures of 0.816 on H&E stained images of colorectal adenocarcinomas and 0.819 on Ki-67 stained images of breast tumor tissue. On average, it was fully trained in less than 15,000 iterations and processed 4.15 megapixels per second at prediction time. CONCLUSIONS: The basic PMap approach is greatly affected by certain parameters. Our evaluation provides guidance on their impact and best settings. When configured properly, this simple and efficient approach can yield equal detection quality as more complex and time-consuming state-of-the-art approaches.

Interactive Visualization of 3D Histopathology in Native Resolution.

We present a visualization application that enables effective interactive visual analysis of large-scale 3D histopathology, that is, high-resolution 3D microscopy data of human tissue. Clinical work flows and research based on pathology have, until now, largely been dominated by 2D imaging. As we will show in the paper, studying volumetric histology data will open up novel and useful opportunities for both research and clinical practice. Our starting point is the current lack of appropriate visualization tools in histopathology, which has been a limiting factor in the uptake of digital pathology. Visualization of 3D histology data does pose difficult challenges in several aspects. The full-color datasets are dense and large in scale, on the order of 100,000 × 100,000× 100 voxels. This entails serious demands on both rendering performance and user experience design. Despite this, our developed application supports interactive study of 3D histology datasets at native resolution. Our application is based on tailoring and tuning of existing methods, system integration work, as well as a careful study of domain specific demands emanating from a close participatory design process with domain experts as team members. Results from a user evaluation employing the tool demonstrate a strong agreement among the 14 participating pathologists that 3D histopathology will be a valuable and enabling tool for their work.

Automated quantification of steatosis: agreement with stereological point counting.

BACKGROUND: Steatosis is routinely assessed histologically in clinical practice and research. Automated image analysis can reduce the effort of quantifying steatosis. Since reproducibility is essential for practical use, we have evaluated different analysis methods in terms of their agreement with stereological point counting (SPC) performed by a hepatologist. METHODS: The evaluation was based on a large and representative data set of 970 histological images from human patients with different liver diseases. Three of the evaluated methods were built on previously published approaches. One method incorporated a new approach to improve the robustness to image variability. RESULTS: The new method showed the strongest agreement with the expert. At 20× resolution, it reproduced steatosis area fractions with a mean absolute error of 0.011 for absent or mild steatosis and 0.036 for moderate or severe steatosis. At 10× resolution, it was more accurate than and twice as fast as all other methods at 20× resolution. When compared with SPC performed by two additional human observers, its error was substantially lower than one and only slightly above the other observer. CONCLUSIONS: The results suggest that the new method can be a suitable automated replacement for SPC. Before further improvements can be verified, it is necessary to thoroughly assess the variability of SPC between human observers.

A Design Study Investigating Augmented Reality and Photograph Annotation in a Digitalized Grossing Workstation.

CONTEXT: Within digital pathology, digitalization of the grossing procedure has been relatively underexplored in comparison to digitalization of pathology slides. AIMS: Our investigation focuses on the interaction design of an augmented reality gross pathology workstation and refining the interface so that information and visualizations are easily recorded and displayed in a thoughtful view. SETTINGS AND DESIGN: The work in this project occurred in two phases: the first phase focused on implementation of an augmented reality grossing workstation prototype while the second phase focused on the implementation of an incremental prototype in parallel with a deeper design study. SUBJECTS AND METHODS: Our research institute focused on an experimental and "designerly" approach to create a digital gross pathology prototype as opposed to focusing on developing a system for immediate clinical deployment. STATISTICAL ANALYSIS USED: Evaluation has not been limited to user tests and interviews, but rather key insights were uncovered through design methods such as "rapid ethnography" and "conversation with materials". RESULTS: We developed an augmented reality enhanced digital grossing station prototype to assist pathology technicians in capturing data during examination. The prototype uses a magnetically tracked scalpel to annotate planned cuts and dimensions onto photographs taken of the work surface. This article focuses on the use of qualitative design methods to evaluate and refine the prototype. Our aims were to build on the strengths of the prototype's technology, improve the ergonomics of the digital/physical workstation by considering numerous alternative design directions, and to consider the effects of digitalization on personnel and the pathology diagnostics information flow from a wider perspective. A proposed interface design allows the pathology technician to place images in relation to its orientation, annotate directly on the image, and create linked information. CONCLUSIONS: The augmented reality magnetically tracked scalpel reduces tool switching though limitations in today's augmented reality technology fall short of creating an ideal immersive workflow by requiring the use of a monitor. While this technology catches up, we recommend focusing efforts on enabling the easy creation of layered, complex reports, linking, and viewing information across systems. Reflecting upon our results, we argue for digitalization to focus not only on how to record increasing amounts of data but also how these data can be accessed in a more thoughtful way that draws upon the expertise and creativity of pathology professionals using the systems.