Preparing healthcare leaders of the digital age with an integrative artificial intelligence curriculum: a pilot study.

Artificial intelligence (AI) is rapidly being introduced into the clinical workflow of many specialties. Despite the need to train physicians who understand the utility and implications of AI and mitigate a growing skills gap, no established consensus exists on how to best introduce AI concepts to medical students during preclinical training. This study examined the effectiveness of a pilot Digital Health Scholars (DHS) non-credit enrichment elective that paralleled the Dartmouth Geisel School of Medicine's first-year preclinical curriculum with a focus on introducing AI algorithms and their applications in the concurrently occurring systems-blocks. From September 2022 to March 2023, ten self-selected first-year students enrolled in the elective curriculum run in parallel with four existing curricular blocks (Immunology, Hematology, Cardiology, and Pulmonology). Each DHS block consisted of a journal club, a live-coding demonstration, and an integration session led by a researcher in that field. Students' confidence in explaining the content objectives (high-level knowledge, implications, and limitations of AI) was measured before and after each block and compared using Mann-Whitney U tests. Students reported significant increases in confidence in describing the content objectives after all four blocks (Immunology: U = 4.5, p = 0.030; Hematology: U = 1.0, p = 0.009; Cardiology: U = 4.0, p = 0.019; Pulmonology: U = 4.0, p = 0.030) as well as an average overall satisfaction level of 4.29/5 in rating the curriculum content. Our study demonstrates that a digital health enrichment elective that runs in parallel to an institution's preclinical curriculum and embeds AI concepts into relevant clinical topics can enhance students' confidence in describing the content objectives that pertain to high-level algorithmic understanding, implications, and limitations of the studied models. Building on this elective curricular design, further studies with a larger enrollment can help determine the most effective approach in preparing future physicians for the AI-enhanced clinical workflow.

Nonmetastatic Axillary Lymph Nodes Have Distinct Morphology and Immunophenotype in Obese Patients with Breast Cancer at Risk for Metastasis.

Obese patients with breast cancer have worse outcomes than their normal weight counterparts, with a 50% to 80% increased rate of axillary nodal metastasis. Recent studies suggest a link between increased lymph node adipose tissue and breast cancer nodal metastasis. Further investigation into potential mechanisms underlying this link may reveal potential prognostic utility of fat-enlarged lymph nodes in patients with breast cancer. This study used a deep learning model to identify morphologic differences in nonmetastatic axillary nodes between obese, node-positive, and node-negative patients with breast cancer. The model was developed using nested cross-validation on 180 cases and achieved an area under the receiver operator characteristic curve of 0.67 in differentiating patients using hematoxylin and eosin-stained whole slide images. The morphologic analysis of the predictive regions showed an increased average adipocyte size (P = 0.004), increased white space between lymphocytes (P < 0.0001), and increased red blood cells (P < 0.001) in nonmetastatic lymph nodes of node-positive patients. Preliminary immunohistochemistry analysis on a subset of 30 patients showed a trend of decreased CD3 expression and increased leptin expression in fat-replaced axillary lymph nodes of obese, node-positive patients. These findings suggest a novel direction to further investigate the interaction between lymph node adiposity, lymphatic dysfunction, and breast cancer nodal metastases, highlighting a possible prognostic tool for obese patients with breast cancer.

Automated classification of fat-infiltrated axillary lymph nodes on screening mammograms.

OBJECTIVE: Fat-infiltrated axillary lymph nodes (LNs) are unique sites for ectopic fat deposition. Early studies showed a strong correlation between fatty LNs and obesity-related diseases. Confirming this correlation requires large-scale studies, hindered by scarce labeled data. With the long-term goal of developing a rapid and generalizable tool to aid data labeling, we developed an automated deep learning (DL)-based pipeline to classify the status of fatty LNs on screening mammograms. METHODS: Our internal data set included 886 mammograms from a tertiary academic medical institution, with a binary status of the fat-infiltrated LNs based on the size and morphology of the largest visible axillary LN. A two-stage DL model training and fine-tuning pipeline was developed to classify the fat-infiltrated LN status using the internal training and development data set. The model was evaluated on a held-out internal test set and a subset of the Digital Database for Screening Mammography. RESULTS: Our model achieved 0.97 (95% CI: 0.94-0.99) accuracy and 1.00 (95% CI: 1.00-1.00) area under the receiver operator characteristic curve on 264 internal testing mammograms, and 0.82 (95% CI: 0.77-0.86) accuracy and 0.87 (95% CI: 0.82-0.91) area under the receiver operator characteristic curve on 70 external testing mammograms. CONCLUSION: This study confirmed the feasibility of using a DL model for fat-infiltrated LN classification. The model provides a practical tool to identify fatty LNs on mammograms and to allow for future large-scale studies to evaluate the role of fatty LNs as an imaging biomarker of obesity-associated pathologies. ADVANCES IN KNOWLEDGE: Our study is the first to classify fatty LNs using an automated DL approach.

HistoPerm: A permutation-based view generation approach for improving histopathologic feature representation learning.

Deep learning has been effective for histology image analysis in digital pathology. However, many current deep learning approaches require large, strongly- or weakly labeled images and regions of interest, which can be time-consuming and resource-intensive to obtain. To address this challenge, we present HistoPerm, a view generation method for representation learning using joint embedding architectures that enhances representation learning for histology images. HistoPerm permutes augmented views of patches extracted from whole-slide histology images to improve classification performance. We evaluated the effectiveness of HistoPerm on 2 histology image datasets for Celiac disease and Renal Cell Carcinoma, using 3 widely used joint embedding architecture-based representation learning methods: BYOL, SimCLR, and VICReg. Our results show that HistoPerm consistently improves patch- and slide-level classification performance in terms of accuracy, F1-score, and AUC. Specifically, for patch-level classification accuracy on the Celiac disease dataset, HistoPerm boosts BYOL and VICReg by 8% and SimCLR by 3%. On the Renal Cell Carcinoma dataset, patch-level classification accuracy is increased by 2% for BYOL and VICReg, and by 1% for SimCLR. In addition, on the Celiac disease dataset, models with HistoPerm outperform the fully supervised baseline model by 6%, 5%, and 2% for BYOL, SimCLR, and VICReg, respectively. For the Renal Cell Carcinoma dataset, HistoPerm lowers the classification accuracy gap for the models up to 10% relative to the fully supervised baseline. These findings suggest that HistoPerm can be a valuable tool for improving representation learning of histopathology features when access to labeled data is limited and can lead to whole-slide classification results that are comparable to or superior to fully supervised methods.

Patient Engagement in a Multimodal Digital Phenotyping Study of Opioid Use Disorder.

BACKGROUND: Multiple digital data sources can capture moment-to-moment information to advance a robust understanding of opioid use disorder (OUD) behavior, ultimately creating a digital phenotype for each patient. This information can lead to individualized interventions to improve treatment for OUD. OBJECTIVE: The aim is to examine patient engagement with multiple digital phenotyping methods among patients receiving buprenorphine medication for OUD. METHODS: The study enrolled 65 patients receiving buprenorphine for OUD between June 2020 and January 2021 from 4 addiction medicine programs in an integrated health care delivery system in Northern California. Ecological momentary assessment (EMA), sensor data, and social media data were collected by smartphone, smartwatch, and social media platforms over a 12-week period. Primary engagement outcomes were meeting measures of minimum phone carry (≥8 hours per day) and watch wear (≥18 hours per day) criteria, EMA response rates, social media consent rate, and data sparsity. Descriptive analyses, bivariate, and trend tests were performed. RESULTS: The participants' average age was 37 years, 47% of them were female, and 71% of them were White. On average, participants met phone carrying criteria on 94% of study days, met watch wearing criteria on 74% of days, and wore the watch to sleep on 77% of days. The mean EMA response rate was 70%, declining from 83% to 56% from week 1 to week 12. Among participants with social media accounts, 88% of them consented to providing data; of them, 55% of Facebook, 54% of Instagram, and 57% of Twitter participants provided data. The amount of social media data available varied widely across participants. No differences by age, sex, race, or ethnicity were observed for any outcomes. CONCLUSIONS: To our knowledge, this is the first study to capture these 3 digital data sources in this clinical population. Our findings demonstrate that patients receiving buprenorphine treatment for OUD had generally high engagement with multiple digital phenotyping data sources, but this was more limited for the social media data. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): RR2-10.3389/fpsyt.2022.871916.

Non-Metastatic Axillary Lymph Nodes Have Distinct Morphology and Immunophenotype in Obese Breast Cancer patients at Risk for Metastasis.

Obese patients have worse breast cancer outcomes than normal weight women including a 50% to 80% increased rate of axillary nodal metastasis. Recent studies have shown a potential link between increased lymph node adipose tissue and breast cancer nodal metastasis. Further investigation into potential mechanisms underlying this link may reveal potential prognostic utility of fat-enlarged lymph nodes in breast cancer patients. In this study, a deep learning framework was developed to identify morphological differences of non-metastatic axillary nodes between node-positive and node-negative obese breast cancer patients. Pathology review of the model-selected patches found an increase in the average size of adipocytes (p-value=0.004), an increased amount of white space between lymphocytes (p-value<0.0001), and an increased amount of red blood cells (p-value<0.001) in non-metastatic lymph nodes of node-positive breast cancer patients. Our downstream immunohistology (IHC) analysis showed a decrease of CD3 expression and increase of leptin expression in fat-replaced axillary lymph nodes in obese node-positive patients. In summary, our findings suggest a novel direction to further investigate the crosstalk between lymph node adiposity, lymphatic dysfunction, and breast cancer nodal metastases.

MHAttnSurv: Multi-head attention for survival prediction using whole-slide pathology images.

Whole slide images (WSI) based survival prediction has attracted increasing interest in pathology. Despite this, extracting prognostic information from WSIs remains a challenging task due to their enormous size and the scarcity of pathologist annotations. Previous studies have utilized multiple instance learning approach to combine information from several randomly sampled patches, but this approach may not be adequate as different visual patterns may contribute unequally to prognosis prediction. In this study, we introduce a multi-head attention mechanism that allows each attention head to independently explore the utility of various visual patterns on a tumor slide, thereby enabling more comprehensive information extraction from WSIs. We evaluated our approach on four cancer types from The Cancer Genome Atlas database. Our model achieved an average c-index of 0.640, outperforming three existing state-of-the-art approaches for WSI-based survival prediction on these datasets. Visualization of attention maps reveals that the attention heads synergistically focus on different morphological patterns, providing additional evidence for the effectiveness of multi-head attention in survival prediction.

Infectious waste management during a pandemic: A stochastic location-routing problem with chance-constrained time windows.

The COVID-19 pandemic has presented tremendous challenges to the world, one of which is the management of infectious waste generated by healthcare activities. Finding cost-efficient services with minimum threats to public health has become a top priority. The pandemic has induced extreme uncertainties, not only in the amount of generated waste, but also in the associated service times. With this in mind, the present study develops a mixed-integer linear programming (MILP) model for the location-routing problem with time windows (LRPTW). To handle the uncertainty in the amount of generated waste, three scenarios are defined respectively reflecting different severity levels of a pandemic. Furthermore, chance constraints are applied to deal with the variation of the service times at small generation nodes, and time windows at the transfer facilities. The complexity of the resulting mathematical model motivated the application of a branch-and-price (B&P) algorithm along with an ɛ -constraint technique. A case study of the situation of Wuhan, China, during the initial COVID-19 outbreak is employed to examine the performance and applicability of the proposed model. Our numerical tests indicate that the B&P algorithm outperforms CPLEX in the computational times by more than 83% in small-sized problem instances and reduces the gaps by at least 70% in large-scale ones. Through a comparison with the current and deterministic systems, our proposed stochastic system can timely adjust itself to fulfill nearly four times the demand of other systems in an extreme pandemic scenario, while maintaining a cost-efficient operation with no outbreak.

Graph Convolutional Neural Networks for Histologic Classification of Pancreatic Cancer.

CONTEXT.­: Pancreatic ductal adenocarcinoma has some of the worst prognostic outcomes among various cancer types. Detection of histologic patterns of pancreatic tumors is essential to predict prognosis and decide the treatment for patients. This histologic classification can have a large degree of variability even among expert pathologists. OBJECTIVE.­: To detect aggressive adenocarcinoma and less aggressive pancreatic tumors from nonneoplasm cases using a graph convolutional network-based deep learning model. DESIGN.­: Our model uses a convolutional neural network to extract detailed information from every small region in a whole slide image. Then, we use a graph architecture to aggregate the extracted features from these regions and their positional information to capture the whole slide-level structure and make the final prediction. RESULTS.­: We evaluated our model on an independent test set and achieved an F1 score of 0.85 for detecting neoplastic cells and ductal adenocarcinoma, significantly outperforming other baseline methods. CONCLUSIONS.­: If validated in prospective studies, this approach has a great potential to assist pathologists in identifying adenocarcinoma and other types of pancreatic tumors in clinical settings.

Detection of Colorectal Adenocarcinoma and Grading Dysplasia on Histopathologic Slides Using Deep Learning.

Colorectal cancer (CRC) is one of the most common types of cancer among men and women. The grading of dysplasia and the detection of adenocarcinoma are important clinical tasks in the diagnosis of CRC and shape the patients' follow-up plans. This study evaluated the feasibility of deep learning models for the classification of colorectal lesions into four classes: benign, low-grade dysplasia, high-grade dysplasia, and adenocarcinoma. To this end, a deep neural network was developed on a training set of 655 whole slide images of digitized colorectal resection slides from a tertiary medical institution; and the network was evaluated on an internal test set of 234 slides, as well as on an external test set of 606 adenocarcinoma slides from The Cancer Genome Atlas database. The model achieved an overall accuracy, sensitivity, and specificity of 95.5%, 91.0%, and 97.1%, respectively, on the internal test set, and an accuracy and sensitivity of 98.5% for adenocarcinoma detection task on the external test set. Results suggest that such deep learning models can potentially assist pathologists in grading colorectal dysplasia, detecting adenocarcinoma, prescreening, and prioritizing the reviewing of suspicious cases to improve the turnaround time for patients with a high risk of CRC. Furthermore, the high sensitivity on the external test set suggests the model's generalizability in detecting colorectal adenocarcinoma on whole slide images across different institutions.

Improving Representation Learning for Histopathologic Images with Cluster Constraints.

Recent advances in whole-slide image (WSI) scanners and computational capabilities have significantly propelled the application of artificial intelligence in histopathology slide analysis. While these strides are promising, current supervised learning approaches for WSI analysis come with the challenge of exhaustively labeling high-resolution slides-a process that is both labor-intensive and timeconsuming. In contrast, self-supervised learning (SSL) pretraining strategies are emerging as a viable alternative, given that they don't rely on explicit data annotations. These SSL strategies are quickly bridging the performance disparity with their supervised counterparts. In this context, we introduce an SSL framework. This framework aims for transferable representation learning and semantically meaningful clustering by synergizing invariance loss and clustering loss in WSI analysis. Notably, our approach outperforms common SSL methods in downstream classification and clustering tasks, as evidenced by tests on the Camelyon16 and a pancreatic cancer dataset. The code and additional details are accessible at https://github.com/wwyi1828/CluSiam.

Association between fat-infiltrated axillary lymph nodes on screening mammography and cardiometabolic disease.

Objective: Ectopic fat deposition within and around organs is a stronger predictor of cardiometabolic disease status than body mass index (BMI). Fat deposition within the lymphatic system is poorly understood. This study examined the association between the prevalence of cardiometabolic disease and ectopic fat deposition within axillary lymph nodes (LNs) visualized on screening mammograms. Methods: A cross-sectional study was conducted on 834 women presenting for full-field digital screening mammography. The status of fat-infiltrated LNs was assessed based on the size and morphology of axillary LNs from screening mammograms. The prevalence of cardiometabolic disease was retrieved from the electronic medical records, including type 2 diabetes mellitus (T2DM), hypertension, dyslipidemia, high blood glucose, cardiovascular disease, stroke, and non-alcoholic fatty liver disease. Results: Fat-infiltrated axillary LNs were associated with a high prevalence of T2DM among all women (adjusted odds ratio: 3.92, 95% CI: [2.40, 6.60], p-value < 0.001) and in subgroups of women with and without obesity. Utilizing the status of fatty LNs improved the classification of T2DM status in addition to age and BMI (1.4% improvement in the area under the receiver operating characteristic curve). Conclusion: Fat-infiltrated axillary LNs visualized on screening mammograms were associated with the prevalence of T2DM. If further validated, fat-infiltrated axillary LNs may represent a novel imaging biomarker of T2DM in women undergoing screening mammography.

Bladder cancer prognosis using deep neural networks and histopathology images.

Background: Recent studies indicate that bladder cancer is among the top 10 most common cancers in the world (Saginala et al. 2022). Bladder cancer frequently reoccurs, and prognostic judgments may vary among clinicians. As a favorable prognosis may help to inform less aggressive treatment plans, classification of histopathology slides is essential for the accurate prognosis and effective treatment of bladder cancer patients. Developing automated and accurate histopathology image analysis methods can help pathologists determine the prognosis of patients with bladder cancer. Materials and methods: In this study, we introduced Bladder4Net, a deep learning pipeline, to classify whole-slide histopathology images of bladder cancer into two classes: low-risk (combination of PUNLMP and low-grade tumors) and high-risk (combination of high-grade and invasive tumors). This pipeline consists of four convolutional neural network (CNN)-based classifiers to address the difficulties of identifying PUNLMP and invasive classes. We evaluated our pipeline on 182 independent whole-slide images from the New Hampshire Bladder Cancer Study (NHBCS) (Karagas et al., 1998; Sverrisson et al., 2014; Sverrisson et al., 2014) collected from 1994 to 2004 and 378 external digitized slides from The Cancer Genome Atlas (TCGA) database (https://www.cancer.gov/tcga). Results: The weighted average F1-score of our approach was 0.91 (95% confidence interval (CI): 0.86-0.94) on the NHBCS dataset and 0.99 (95% CI: 0.97-1.00) on the TCGA dataset. Additionally, we computed Kaplan-Meier survival curves for patients who were predicted as high risk versus those predicted as low risk. For the NHBCS test set, patients predicted as high risk had worse overall survival than those predicted as low risk, with a log-rank p-value of 0.004. Conclusions: If validated through prospective trials, our model could be used in clinical settings to improve patient care.

Quantifying epithelial-mesenchymal heterogeneity and EMT scoring in tumor samples via tyramide signal amplification (TSA).

Tumor heterogeneity presents an ongoing challenge to disease progression and treatment in many solid tumor types. Understanding the roots of intra-tumoral heterogeneity and how it may relate to the high incidence of metastasis is critical in overcoming disease relapse and chemoresistance. The epithelial-to-mesenchymal transition is a dynamic cellular program that is co-opted by cancer cells to enhance, among others, migratory and invasive cell traits. It is a key contributor to heterogeneity, chemo-resistance, and metastasis in many carcinoma-types, with the intermediate or hybrid EMT state playing a critical role due to its increased tumor-initiating potential. A critical component in utilizing this knowledge in patient treatment is to first detect and score the impact of EMT in a patient sample. Here, we provide a detailed protocol to detect EMT states and quantify the resulting epithelial-mesenchymal heterogeneity within tumors using a novel multiplexed immunostaining approach and analysis method. This protocol and concept can easily be adapted using custom panels of markers to explore other sources of tumoral heterogeneity in addition to EMT.

Phenotypic heterogeneity driven by plasticity of the intermediate EMT state governs disease progression and metastasis in breast cancer.

The epithelial-to-mesenchymal transition (EMT) is frequently co-opted by cancer cells to enhance migratory and invasive cell traits. It is a key contributor to heterogeneity, chemoresistance, and metastasis in many carcinoma types, where the intermediate EMT state plays a critical tumor-initiating role. We isolate multiple distinct single-cell clones from the SUM149PT human breast cell line spanning the EMT spectrum having diverse migratory, tumor-initiating, and metastatic qualities, including three unique intermediates. Using a multiomics approach, we identify CBFß as a key regulator of metastatic ability in the intermediate state. To quantify epithelial-mesenchymal heterogeneity within tumors, we develop an advanced multiplexed immunostaining approach using SUM149-derived orthotopic tumors and find that the EMT state and epithelial-mesenchymal heterogeneity are predictive of overall survival in a cohort of stage III breast cancer. Our model reveals previously unidentified insights into the complex EMT spectrum and its regulatory networks, as well as the contributions of epithelial-mesenchymal plasticity (EMP) in tumor heterogeneity in breast cancer.

Predicting oncogene mutations of lung cancer using deep learning and histopathologic features on whole-slide images.

Lung cancer is a leading cause of death in both men and women globally. The recent development of tumor molecular profiling has opened opportunities for targeted therapies for lung adenocarcinoma (LUAD) patients. However, the lack of access to molecular profiling or cost and turnaround time associated with it could hinder oncologists' willingness to order frequent molecular tests, limiting potential benefits from precision medicine. In this study, we developed a weakly supervised deep learning model for predicting somatic mutations of LUAD patients based on formalin-fixed paraffin-embedded (FFPE) whole-slide images (WSIs) using LUAD subtypes-related histological features and recent advances in computer vision. Our study was performed on a total of 747 hematoxylin and eosin (H&E) stained FFPE LUAD WSIs and the genetic mutation data of 232 patients who were treated at Dartmouth-Hitchcock Medical Center (DHMC). We developed our convolutional neural network-based models to analyze whole slides and predict five major genetic mutations, i.e., BRAF, EGFR, KRAS, STK11, and TP53. We additionally used 111 cases from the LUAD dataset of the CPTAC-3 study for external validation. Our model achieved an AUROC of 0.799 (95% CI: 0.686-0.904) and 0.686 (95% CI: 0.620-0.752) for predicting EGFR genetic mutations on the DHMC and CPTAC-3 test sets, respectively. Predicting TP53 genetic mutations also showed promising outcomes. Our results demonstrated that H&E stained FFPE LUAD whole slides could be utilized to predict oncogene mutations, such as EGFR, indicating that somatic mutations could present subtle morphological characteristics in histology slides, where deep learning-based feature extractors can learn such latent information.

Enhancing research data infrastructure to address the opioid epidemic: the Opioid Overdose Network (O2-Net).

Opioid Overdose Network is an effort to generalize and adapt an existing research data network, the Accrual to Clinical Trials (ACT) Network, to support design of trials for survivors of opioid overdoses presenting to emergency departments (ED). Four institutions (Medical University of South Carolina [MUSC], Dartmouth Medical School [DMS], University of Kentucky [UK], and University of California San Diego [UCSD]) worked to adapt the ACT network. The approach that was taken to enhance the ACT network focused on 4 activities: cloning and extending the ACT infrastructure, developing an e-phenotype and corresponding registry, developing portable natural language processing tools to enhance data capture, and developing automated documentation templates to enhance extended data capture. Overall, initial results suggest that tailoring of existing multipurpose federated research networks to specific tasks is feasible; however, substantial efforts are required for coordination of the subnetwork and development of new tools for extension of available data. The initial output of the project was a new approach to decision support for the prescription of naloxone for home use in the ED, which is under further study within the network.

The Feasibility and Utility of Harnessing Digital Health to Understand Clinical Trajectories in Medication Treatment for Opioid Use Disorder: D-TECT Study Design and Methodological Considerations.

Introduction: Across the U.S., the prevalence of opioid use disorder (OUD) and the rates of opioid overdoses have risen precipitously in recent years. Several effective medications for OUD (MOUD) exist and have been shown to be life-saving. A large volume of research has identified a confluence of factors that predict attrition and continued substance use during substance use disorder treatment. However, much of this literature has examined a small set of potential moderators or mediators of outcomes in MOUD treatment and may lead to over-simplified accounts of treatment non-adherence. Digital health methodologies offer great promise for capturing intensive, longitudinal ecologically-valid data from individuals in MOUD treatment to extend our understanding of factors that impact treatment engagement and outcomes. Methods: This paper describes the protocol (including the study design and methodological considerations) from a novel study supported by the National Drug Abuse Treatment Clinical Trials Network at the National Institute on Drug Abuse (NIDA). This study (D-TECT) primarily seeks to evaluate the feasibility of collecting ecological momentary assessment (EMA), smartphone and smartwatch sensor data, and social media data among patients in outpatient MOUD treatment. It secondarily seeks to examine the utility of EMA, digital sensing, and social media data (separately and compared to one another) in predicting MOUD treatment retention, opioid use events, and medication adherence [as captured in electronic health records (EHR) and EMA data]. To our knowledge, this is the first project to include all three sources of digitally derived data (EMA, digital sensing, and social media) in understanding the clinical trajectories of patients in MOUD treatment. These multiple data streams will allow us to understand the relative and combined utility of collecting digital data from these diverse data sources. The inclusion of EHR data allows us to focus on the utility of digital health data in predicting objectively measured clinical outcomes. Discussion: Results may be useful in elucidating novel relations between digital data sources and OUD treatment outcomes. It may also inform approaches to enhancing outcomes measurement in clinical trials by allowing for the assessment of dynamic interactions between individuals' daily lives and their MOUD treatment response. Clinical Trial Registration: Identifier: NCT04535583.

AI-RADS: Successes and challenges of a novel artificial intelligence curriculum for radiologists across different delivery formats.

Introduction: Artificial intelligence and data-driven predictive modeling have become increasingly common tools integrated in clinical practice, heralding a new chapter of medicine in the digital era. While these techniques are poised to affect nearly all aspects of medicine, medical education as an institution has languished behind; this has raised concerns that the current training infrastructure is not adequately preparing future physicians for this changing clinical landscape. Our institution attempted to ameliorate this by implementing a novel artificial intelligence in radiology curriculum, "AI-RADS," in two different educational formats: a 7-month lecture series and a one-day workshop intensive. Methods: The curriculum was structured around foundational algorithms within artificial intelligence. As most residents have little computer science training, algorithms were initially presented as a series of simple observations around a relatable problem (e.g., fraud detection, movie recommendations, etc.). These observations were later re-framed to illustrate how a machine could apply the underlying concepts to perform clinically relevant tasks in the practice of radiology. Secondary lessons in basic computing, such as data representation/abstraction, were integrated as well. The lessons were ordered such that these algorithms were logical extensions of each other. The 7-month curriculum consisted of seven lectures paired with seven journal clubs, resulting in an AI-focused session every two weeks. The workshop consisted of six hours of content modified for the condensed format, with a final integrative activity. Results: Both formats of the AI-RADS curriculum were well received by learners, with the 7-month version and workshop garnering 9.8/10 and 4.3/5 ratings, respectively, for overall satisfaction. In both, there were increases in perceived understanding of artificial intelligence. In the 7-lecture course, 6/7 lectures achieved statistically significant (P < 0.02) differences, with the final lecture approaching significance (P = 0.07). In the one-day workshop, there was a significant increase in perceived understanding (P = 0.03). Conclusion: As artificial intelligence becomes further enmeshed in clinical practice, it will become critical for physicians to have a basic understanding of how these tools work. Our AI-RADS curriculum demonstrates that it is successful in increasing learner perceived understanding in both an extended and condensed format.