Automated data abstraction for quality surveillance and outcome assessment in radiation oncology.

Rigorous radiotherapy quality surveillance and comprehensive outcome assessment require electronic capture and automatic abstraction of clinical, radiation treatment planning, and delivery data. We present the design and implementation framework of an integrated data abstraction, aggregation, and storage, curation, and analytics software: the Health Information Gateway and Exchange (HINGE), which collates data for cancer patients receiving radiotherapy. The HINGE software abstracts structured DICOM-RT data from the treatment planning system (TPS), treatment data from the treatment management system (TMS), and clinical data from the electronic health records (EHRs). HINGE software has disease site-specific "Smart" templates that facilitate the entry of relevant clinical information by physicians and clinical staff in a discrete manner as part of the routine clinical documentation. Radiotherapy data abstracted from these disparate sources and the smart templates are processed for quality and outcome assessment. The predictive data analyses are done on using well-defined clinical and dosimetry quality measures defined by disease site experts in radiation oncology. HINGE application software connects seamlessly to the local IT/medical infrastructure via interfaces and cloud services and performs data extraction and aggregation functions without human intervention. It provides tools to assess variations in radiation oncology practices and outcomes and determines gaps in radiotherapy quality delivered by each provider.

A Machine Learning method for relabeling arbitrary DICOM structure sets to TG-263 defined labels.

PURPOSE: To present a Machine Learning pipeline for automatically relabeling anatomical structure sets in the Digital Imaging and Communications in Medicine (DICOM) format to a standard nomenclature that will enable data abstraction for research and quality improvement. METHODS: DICOM structure sets from approximately 1200 lung and prostate cancer patients across 40 treatment centers were used to build predictive models to automate the relabeling of clinically specified structure labels to standardized labels as defined by the American Association of Physics in Medicine's (AAPM) Task Group 263 (TG-263). Volumetric bitmaps were created based on the delineated volumes and were combined with associated bony anatomy data to build feature vectors. Feature reduction was performed with singular value decomposition and the resulting vectors were used for predicting the label of each structure using five different classifier algorithms on the Apache Spark platform with 5-fold cross-validation. Undersampling methods were used to deal with underlying class imbalance that hindered the performance of classifiers. Experiments were performed on both a curated version of the data, which included only annotated structures, and the non-curated data that included all structures from the original treatment plans. RESULTS: Random Forest provided the highest accuracies with F1 scores of 98.77 for lung and 95.06 for prostate on the curated data sets. Scores were lower with 95.67 for lung and 90.22 for prostate on the non-curated data sets, highlighting some of the challenges of classifying real clinical data. Including bony anatomy data and pooling information from all structures for the same patient both increased accuracies. In some cases, undersampling with k-Means clustering for class balancing improved classifier accuracy but in all experiments it significantly reduced run time compared to random undersampling. CONCLUSION: This work shows that structure sets can be relabeled using our approach with accuracies over 95% for many structure types when presented with curated data. Although accuracies dropped when using the full non-curated data sets, some structure types were still correctly labeled over 90% of the time. With similar results obtained on an external test data set, we can infer that the proposed models are likely to work on other clinical data sets.

VA-Radiation Oncology Quality Surveillance Program.

PURPOSE: We sought to develop a quality surveillance program for approximately 15,000 US veterans treated at the 40 radiation oncology facilities at the Veterans Affairs (VA) hospitals each year. METHODS AND MATERIALS: State-of-the-art technologies were used with the goal to improve clinical outcomes while providing the best possible care to veterans. To measure quality of care and service rendered to veterans, the Veterans Health Administration established the VA Radiation Oncology Quality Surveillance program. The program carries forward the American College of Radiology Quality Research in Radiation Oncology project methodology of assessing the wide variation in practice pattern and quality of care in radiation therapy by developing clinical quality measures (QM) used as quality indices. These QM data provide feedback to physicians by identifying areas for improvement in the process of care and identifying the adoption of evidence-based recommendations for radiation therapy. RESULTS: Disease-site expert panels organized by the American Society for Radiation Oncology (ASTRO) defined quality measures and established scoring criteria for prostate cancer (intermediate and high risk), non-small cell lung cancer (IIIA/B stage), and small cell lung cancer (limited stage) case presentations. Data elements for 1567 patients from the 40 VA radiation oncology practices were abstracted from the electronic medical records and treatment management and planning systems. Overall, the 1567 assessed cases passed 82.4% of all QM. Pass rates for QM for the 773 lung and 794 prostate cases were 78.0% and 87.2%, respectively. Marked variations, however, were noted in the pass rates for QM when tumor site, clinical pathway, or performing centers were separately examined. CONCLUSIONS: The peer-review protected VA-Radiation Oncology Surveillance program based on clinical quality measures allows providers to compare their clinical practice to peers and to make meaningful adjustments in their personal patterns of care unobtrusively.

Determining causal miRNAs and their signaling cascade in diseases using an influence diffusion model.

In recent studies, miRNAs have been found to be extremely influential in many of the essential biological processes. They exhibit a self-regulatory mechanism through which they act as positive/negative regulators of expression of genes and other miRNAs. This has direct implications in the regulation of various pathophysiological conditions, signaling pathways and different types of cancers. Studying miRNA-disease associations has been an extensive area of research; however deciphering miRNA-miRNA network regulatory patterns in several diseases remains a challenge. In this study, we use information diffusion theory to quantify the influence diffusion in a miRNA-miRNA regulation network across multiple disease categories. Our proposed methodology determines the critical disease specific miRNAs which play a causal role in their signaling cascade and hence may regulate disease progression. We extensively validate our framework using existing computational tools from the literature. Furthermore, we implement our framework on a comprehensive miRNA expression data set for alcohol dependence and identify the causal miRNAs for alcohol-dependency in patients which were validated by the phase-shift in their expression scores towards the early stages of the disease. Finally, our computational framework for identifying causal miRNAs implicated in diseases is available as a free online tool for the greater scientific community.

miRsig: a consensus-based network inference methodology to identify pan-cancer miRNA-miRNA interaction signatures.

Decoding the patterns of miRNA regulation in diseases are important to properly realize its potential in diagnostic, prog- nostic, and therapeutic applications. Only a handful of studies computationally predict possible miRNA-miRNA interactions; hence, such interactions require a thorough investigation to understand their role in disease progression. In this paper, we design a novel computational pipeline to predict the common signature/core sets of miRNA-miRNA interactions for different diseases using network inference algorithms on the miRNA-disease expression profiles; the individual predictions of these algorithms were then merged using a consensus-based approach to predict miRNA-miRNA associations. We next selected the miRNA-miRNA associations across particular diseases to generate the corresponding disease-specific miRNA-interaction networks. Next, graph intersection analysis was performed on these networks for multiple diseases to identify the common signature/core sets of miRNA interactions. We applied this pipeline to identify the common signature of miRNA-miRNA inter- actions for cancers. The identified signatures when validated using a manual literature search from PubMed Central and the PhenomiR database, show strong relevance with the respective cancers, providing an indirect proof of the high accuracy of our methodology. We developed miRsig, an online tool for analysis and visualization of the disease-specific signature/core miRNA-miRNA interactions, available at: http://bnet.egr.vcu.edu/miRsig.

miRegulome: a knowledge-base of miRNA regulomics and analysis.

UNLABELLED: miRNAs regulate post transcriptional gene expression by targeting multiple mRNAs and hence can modulate multiple signalling pathways, biological processes, and patho-physiologies. Therefore, understanding of miRNA regulatory networks is essential in order to modulate the functions of a miRNA. The focus of several existing databases is to provide information on specific aspects of miRNA regulation. However, an integrated resource on the miRNA regulome is currently not available to facilitate the exploration and understanding of miRNA regulomics. miRegulome attempts to bridge this gap. The current version of miRegulome v1.0 provides details on the entire regulatory modules of miRNAs altered in response to chemical treatments and transcription factors, based on validated data manually curated from published literature. Modules of miRegulome (upstream regulators, downstream targets, miRNA regulated pathways, functions, diseases, etc) are hyperlinked to an appropriate external resource and are displayed visually to provide a comprehensive understanding. Four analysis tools are incorporated to identify relationships among different modules based on user specified datasets. miRegulome and its tools are helpful in understanding the biology of miRNAs and will also facilitate the discovery of biomarkers and therapeutics. With added features in upcoming releases, miRegulome will be an essential resource to the scientific community. AVAILABILITY: http://bnet.egr.vcu.edu/miRegulome.

DISMIRA: Prioritization of disease candidates in miRNA-disease associations based on maximum weighted matching inference model and motif-based analysis.

BACKGROUND: MicroRNAs (miRNAs) have increasingly been found to regulate diseases at a significant level. The interaction of miRNA and diseases is a complex web of multilevel interactions, given the fact that a miRNA regulates upto 50 or more diseases and miRNAs/diseases work in clusters. The clear patterns of miRNA regulations in a disease are still elusive. METHODS: In this work, we approach the miRNA-disease interactions from a network scientific perspective and devise two approaches - maximum weighted matching model (a graph theoretical algorithm which provides the result by solving an optimization equation of selecting the most prominent set of diseases) and motif-based analyses (which investigates the motifs of the miRNA-disease network and selects the most prominent set of diseases based on their maximum number of participation in motifs, thereby revealing the miRNA-disease interaction dynamics) to determine and prioritize the set of diseases which are most certainly impacted upon the activation of a group of queried miRNAs, in a miRNA-disease network. RESULTS AND CONCLUSION: Our tool, DISMIRA implements the above mentioned approaches and presents an interactive visualization which helps the user in exploring the networking dynamics of miRNAs and diseases by analyzing their neighbors, paths and topological features. A set of miRNAs can be used in this analysis to get the associated diseases for the input group of miRs with ranks and also further analysis can be done to find key miRs or diseases, shortest paths etc. DISMIRA can be accessed online for free at http://bnet.egr.vcu.edu:8080/dismira.