Naval intelligence plays a critical role in multi-domain operations by identifying and tracking vessels of interest, especially suspected "dark ships" operating in an emissions-controlled (EMCON) state. While applying machine learning (ML) to maritime satellite imagery could enable an automated open-ocean search capability for dark ships, ensuring the robustness of ML models to environmental variations in the maritime domain remains a challenge because training sets do not encapsulate all possible environmental conditions. To address the challenge of unsupervised domain adaptation (UDA) in ship classification, i.e. transferring a ML model from a labeled source domain to an unlabeled target domain, we propose employing combinations of semi-supervised learning (SSL) techniques with standalone UDA approaches. Specifically, we incorporate combinations of FixMatch, minimum class confusion, gradient reversal, and mixup augmentation into the standard cross-entropy supervised loss function. These interventions were compared in two domain shift settings, one in which the source and target domains are both comprised of simulated data, and another in which the source domain consists of only simulated data, and the target domain consists of only real data. Experimental results comparing the combinations of interventions to a regularized fine-tuning baseline demonstrate that the greatest improvements in model robustness were achieved when combinations of our SSL strategy (FixMatch) and UDA algorithms were incorporated into training.However, ensuring the robustness of ML models for maritime imagery remains a challenge. We define "robustness" as the performance of a ML model on data generated from a counterfactually-altered version of the data-generating process. 5 In maritime imagery, the data-generating process is defined by numerous environmental and operating factors that produce non-i.i.d. data. For instance, the brightness and texture of the ocean's surface under a clear sky is a complex function of the viewing angle, solar angle, and wind speed. 6 Collecting training data over the full range of conditions can be difficult, so datasets may be biased toward ocean conditions with clear skies and calm seas. Furthermore, the observable features of a ship and its wake can change as a function its velocity relative to the waves, sun, and viewpoint, 7 and collected datasets may not be able to sufficiently sample this space.A potential "robustness tactic", 5 or intervention, to improve maritime image classifier robustness is to employ synthetic training data that accommodates physics-based models for the relevant optics and ocean dynamics that define the data-generating process. However, even the most sophisticated models cannot employ all of the relevant factors that impact the imagery (e.g. clouds, biologics). Naively training on simulated data can impose a different bias in the ML model than if it was trained on "cherry-picked" real-world data. Recently, many unsupervised domain adaptation (UDA) techniques 8,9 have been propos...

BACKGROUNDChanges in medical anatomy teaching have resulted from the application of modern learning theory, transitions to integrated curricula, decreased instructional hours, and a desire for distributed learning across the medEd continuum. In response, anatomists have tried to refine and reduce content, but this change raises the question: What anatomy is essential and for which stage of learning (for USMLE Step 1, clerkships, residency)? Often, anatomy curricula may be based on “what all medical students need,” asked to meet the needs of a “generalist,” or to provide the anatomical foundations for physical and neurological examinations. Recent studies based on the expertise of both clinical and anatomy faculty have offered proposals for curricula at different stages in the learning continuum (Lisk et al., 2013; Lazerus et al., 2014; Tubbs et al., 2014; Smith et al., 2016). Comprehensive clinical anatomy curricula for gross and developmental anatomy were also previously proposed by the AACA (1996, 2000). The present study, which expands on preliminary data presented in 2016, is the first broad assessment of the clinical importance of anatomy by residents in specific fields of medical practice. This presentation will discuss the clinical importance of a comprehensive list of anatomical structures and concepts across all anatomical regions to provide a better understanding of anatomy as applied in a clinical context.METHODSThis study reports survey data regarding the clinical importance of anatomy as assessed by 109 residents: anesthesiology (n=7), emergency medicine (n=14), family medicine (n=12), internal medicine (n=35), med‐peds (n=7), obstetrics & gynecology (n=13), and orthopedics (n=21).RESULTS & CONCLUSIONSThe rankings of the clinical importance of specific anatomy by residents in different fields of practice often conformed to preconceived notions of the field of practice (e.g., ObGyn residents ranked anatomy of the pelvis and perineum in the female as essential, but not the male). In some areas, however, the rankings surprised (e.g., low assessment of back surface anatomy by orthopaedics residents). Further stratification of the data based on variables such as postgraduate year and area of specialization may better define the anatomy considered essential by different groups of residents, especially in fields such as internal medicine and family medicine that encompass a broad range of subfields. The outcome of this study and the high level of specificity provided by respondents provides a database that may be used to for course planning for residency training. Moreover, this data will be useful in undergraduate medical education to inform the development anatomy learning experiences that both fit within curricular parameters and meet the needs of medical student learners.Support or Funding InformationNoneThis abstract is from the Experimental Biology 2018 Meeting. There is no full text article associated with this abstract published in The FASEB Journal.