Pilot study to evaluate tools to collect pathologist annotations for validating machine learning algorithms.

Purpose: Validation of artificial intelligence (AI) algorithms in digital pathology with a reference standard is necessary before widespread clinical use, but few examples focus on creating a reference standard based on pathologist annotations. This work assesses the results of a pilot study that collects density estimates of stromal tumor-infiltrating lymphocytes (sTILs) in breast cancer biopsy specimens. This work will inform the creation of a validation dataset for the evaluation of AI algorithms fit for a regulatory purpose. Approach: Collaborators and crowdsourced pathologists contributed glass slides, digital images, and annotations. Here, "annotations" refer to any marks, segmentations, measurements, or labels a pathologist adds to a report, image, region of interest (ROI), or biological feature. Pathologists estimated sTILs density in 640 ROIs from hematoxylin and eosin stained slides of 64 patients via two modalities: an optical light microscope and two digital image viewing platforms. Results: The pilot study generated 7373 sTILs density estimates from 29 pathologists. Analysis of annotations found the variability of density estimates per ROI increases with the mean; the root mean square differences were 4.46, 14.25, and 26.25 as the mean density ranged from 0% to 10%, 11% to 40%, and 41% to 100%, respectively. The pilot study informs three areas of improvement for future work: technical workflows, annotation platforms, and agreement analysis methods. Upgrades to the workflows and platforms will improve operability and increase annotation speed and consistency. Conclusions: Exploratory data analysis demonstrates the need to develop new statistical approaches for agreement. The pilot study dataset and analysis methods are publicly available to allow community feedback. The development and results of the validation dataset will be publicly available to serve as an instructive tool that can be replicated by developers and researchers.

Assigning readers to cases in imaging studies using balanced incomplete block designs.

In many imaging studies, each case is reviewed by human readers and characterized according to one or more features. Often, the inter-reader agreement of the feature indications is of interest in addition to their diagnostic accuracy or association with clinical outcomes. Complete designs in which all participating readers review all cases maximize efficiency and guarantee estimability of agreement metrics for all pairs of readers but often involve a heavy reading burden. Assigning readers to cases using balanced incomplete block designs substantially reduces reading burden by having each reader review only a subset of cases, while still maintaining estimability of inter-reader agreement for all pairs of readers. Methodology for data analysis and power and sample size calculations under balanced incomplete block designs is presented and applied to simulation studies and an actual example. Simulation studies results suggest that such designs may reduce reading burdens by >40% while in most scenarios incurring a <20% increase in the standard errors and a <8% and <20% reduction in power to detect between-modality differences in diagnostic accuracy and κ statistics, respectively.

Paired split-plot designs of multireader multicase studies.

The widely used multireader multicase ROC study design for comparing imaging modalities is the fully crossed (FC) design: every reader reads every case of both modalities. We investigate paired split-plot (PSP) designs that may allow for reduced cost and increased flexibility compared with the FC design. In the PSP design, case images from two modalities are read by the same readers, thereby the readings are paired across modalities. However, within each modality, not every reader reads every case. Instead, both the readers and the cases are partitioned into a fixed number of groups and each group of readers reads its own group of cases-a split-plot design. Using a [Formula: see text]-statistic based variance analysis for AUC (i.e., area under the ROC curve), we show analytically that precision can be gained by the PSP design as compared with the FC design with the same number of readers and readings. Equivalently, we show that the PSP design can achieve the same statistical power as the FC design with a reduced number of readings. The trade-off for the increased precision in the PSP design is the cost of collecting a larger number of truth-verified patient cases than the FC design. This means that one can trade-off between different sources of cost and choose a least burdensome design. We provide a validation study to show the iMRMC software can be reliably used for analyzing data from both FC and PSP designs. Finally, we demonstrate the advantages of the PSP design with a reader study comparing full-field digital mammography with screen-film mammography.

Full receiver operating characteristic curve estimation using two alternative forced choice studies.

Task-based medical image quality is typically measured by the degree to which a human observer can perform a diagnostic task in a psychophysical human observer study. During a typical study, an observer is asked to provide a numerical score quantifying his confidence as to whether an image contains a diagnostic marker or not. Such scores are then used to measure the observers' diagnostic accuracy, summarized by the receiver operating characteristic (ROC) curve and the area under ROC curve. These types of human studies are difficult to arrange, costly, and time consuming. In addition, human observers involved in this type of study should be experts on the image genre to avoid inconsistent scoring through the lengthy study. In two-alternative forced choice (2AFC) studies, known to be faster, two images are compared simultaneously and a single indicator is given. Unfortunately, the 2AFC approach cannot lead to a full ROC curve or a set of image scores. The aim of this work is to propose a methodology in which multiple rounds of the 2AFC studies are used to re-estimate an image confidence score (a.k.a. rating, ranking) and generate the full ROC curve. In the proposed approach, we treat image confidence score as an unknown rating that needs to be estimated and 2AFC as a two-player match game. To achieve this, we use the ELO rating system, which is used for calculating the relative skill levels of players in competitor-versus-competitor games such as chess. The proposed methodology is not limited to ELO, and other rating methods such as TrueSkill™, Chessmetrics, or Glicko can be also used. The presented results, using simulated data, indicate that a full ROC curve can be recovered using several rounds of 2AFC studies and that the best pairing strategy starts with the first round of pairing abnormal versus normal images (as in the classical 2AFC approach) followed by a number of rounds using random pairing. In addition, the proposed method was tested in a pilot human observer study. These pilot results indicate that three to five rounds of 2AFC studies require less human observer time than a full scoring study and that the re-estimated ROC curves and associated area under ROC curve values have high statistical agreement with the full scoring study.

Evaluation environment for digital and analog pathology: a platform for validation studies.

We present a platform for designing and executing studies that compare pathologists interpreting histopathology of whole slide images (WSIs) on a computer display to pathologists interpreting glass slides on an optical microscope. eeDAP is an evaluation environment for digital and analog pathology. The key element in eeDAP is the registration of the WSI to the glass slide. Registration is accomplished through computer control of the microscope stage and a camera mounted on the microscope that acquires real-time images of the microscope field of view (FOV). Registration allows for the evaluation of the same regions of interest (ROIs) in both domains. This can reduce or eliminate disagreements that arise from pathologists interpreting different areas and focuses on the comparison of image quality. We reduced the pathologist interpretation area from an entire glass slide (10 to [Formula: see text]) to small ROIs ([Formula: see text]). We also made possible the evaluation of individual cells. We summarize eeDAP's software and hardware and provide calculations and corresponding images of the microscope FOV and the ROIs extracted from the WSIs. The eeDAP software can be downloaded from the Google code website (project: eeDAP) as a MATLAB source or as a precompiled stand-alone license-free application.

eeDAP: An Evaluation Environment for Digital and Analog Pathology.

PURPOSE: The purpose of this work is to present a platform for designing and executing studies that compare pathologists interpreting histopathology of whole slide images (WSI) on a computer display to pathologists interpreting glass slides on an optical microscope. METHODS: Here we present eeDAP, an evaluation environment for digital and analog pathology. The key element in eeDAP is the registration of the WSI to the glass slide. Registration is accomplished through computer control of the microscope stage and a camera mounted on the microscope that acquires images of the real time microscope view. Registration allows for the evaluation of the same regions of interest (ROIs) in both domains. This can reduce or eliminate disagreements that arise from pathologists interpreting different areas and focuses the comparison on image quality. RESULTS: We reduced the pathologist interpretation area from an entire glass slide (≈10-30 mm)2 to small ROIs <(50 um)2. We also made possible the evaluation of individual cells. CONCLUSIONS: We summarize eeDAP's software and hardware and provide calculations and corresponding images of the microscope field of view and the ROIs extracted from the WSIs. These calculations help provide a sense of eeDAP's functionality and operating principles, while the images provide a sense of the look and feel of studies that can be conducted in the digital and analog domains. The eeDAP software can be downloaded from code.google.com (project: eeDAP) as Matlab source or as a precompiled stand-alone license-free application.

Generalized Roe and Metz receiver operating characteristic model: analytic link between simulated decision scores and empirical AUC variances and covariances.

Modeling and simulation are often used to understand and investigate random quantities and estimators. In 1997, Roe and Metz introduced a simulation model to validate analysis methods for the popular endpoint in reader studies to evaluate medical imaging devices, the reader-averaged area under the receiver operating characteristic (ROC) curve. Here, we generalize the notation of the model to allow more flexibility in recognition that variances of ROC ratings depend on modality and truth state. We also derive and validate equations for computing population variances and covariances for reader-averaged empirical AUC estimates under the generalized model. The equations are one-dimensional integrals that can be calculated using standard numerical integration techniques. This work provides the theoretical foundation and validation for a Java application called iRoeMetz that can simulate multireader multicase ROC studies and numerically calculate the corresponding variances and covariances of the empirical AUC. The iRoeMetz application and source code can be found at the "iMRMC" project on the google code project hosting site. These results and the application can be used by investigators to investigate ROC endpoints, validate analysis methods, and plan future studies.