Archive for the ‘Education & Training’ Category
On Thursday, September 8, the one-hour presentation Overview of Deep Learning in Healthcare was held from 12:00-1:00 PM PDT. The session was recorded and is available for viewing through NIH Videocast.
Abstract: Machine Learning (ML) has become a core technology underlying many modern applications, especially in healthcare. Machine learning techniques provide powerful methods for analyzing large data sets, such as medical images, electronic health records, and genomics. Recent advances in Deep Learning (DL) provide an analysis framework that can be used to automatically classify images and objects with (and occasionally exceeding) human-level accuracy. A key advantage of Deep Learning is its ability to perform unsupervised feature extraction over massive data sets making big data part of the solution — not the problem. Deep Learning is rapidly becoming a key tool at many of the top technology companies around the world.
The talk will introduce DL in the broader context of machine learning and discuss critical factors driving the success of DL with examples of how deep learning is advancing healthcare. It will also outline development and deployment workflows for building powerful DL solutions and provide an overview of relevant open source tool kits, companies, and products. The session will wrap up with a short demo of NVIDIA’s DIGITS training system for rapidly prototyping your own deep learning applications.
Registration is available for the inaugural session of a new bi-monthly NN/LM collaborative webinar series, NN/LM Resource Picks, which will be held on September 28, 12:00-1:00 PM PDT. Every two months, NN/LM and/or NLM staff will highlight NLM resources. Coming up first is Don’t Wait, Communicate About Disaster Preparedness, hosted by the NN/LM Middle Atlantic Region.
One of the core competencies of disaster medicine is knowing how to “identify authoritative sources for information in a disaster or public health emergency.” Librarians and information professionals with this competency can support their communities with high-quality information throughout the disaster cycle of mitigation, preparedness, response, and recovery. The National Library of Medicine has developed a series of courses with an emphasis on disaster health information. The courses are currently being updated and formatted for self-paced study online. Two courses listed below are now available. The courses meet the requirements for the Medical Library Association Disaster Information Specialization, as well as core competencies for public health professionals and others through the Public Health Foundation’s learning management system, TRAIN. By the end of the year, there will be four more courses: US Response to Disasters and Public Health Emergencies; Information Roles in Disaster Management; A Seat at the Table: Working with Local Responders; and Health and Disasters: Understanding the International Context.
Disaster Health Information Sources: The Basics
This class provides a comprehensive overview of the essential resources needed to provide health-related information services for supporting disaster mitigation, planning, response, and recovery. This self-paced course introduces key sources from the National Library of Medicine, federal and nonfederal agencies, and international organizations. Tools for locating, organizing and disseminating disaster health information are covered.
CBRNE (Chemical, Biological, Radiological, Nuclear and Explosive): Health Information Resources
This class provides an overview of the concepts of CBRNE, including a review of National Library of Medicine resources and tools that provide health-related information to support planning, response, and recovery from the effects of these potential hazards.
The 2016 Association of Health Care Journalists (AHCJ)-National Library of Medicine (NLM) Fellows class features seven reporters and editors representing diverse media backgrounds.
The 2016 AHCJ-NLM Fellows are:
- Rachel Bluth, reporter, Kaiser Health News
- Shannon Firth, Washington reporter, MedPage Today
- Julio Ochoa, editor, WUSF-Health News Florida
- David Wahlberg, health/medical reporter, Wisconsin State Journal
- Leigh Ann Winick, medical producer, CBS News
- Paula Andalo, senior managing editor, HolaDoctor
- Laura Beil, independent journalist, Dallas
Now in its eighth year, the program brings journalists selected by AHCJ to NLM for four days of training to better use some of NLM’s health information resources, such as PubMed, PubMed Health, Genetics Home Reference, TOXMAP, ClinicalTrials.gov, and MedlinePlus. This year’s Fellows class will be at NLM September 26-30. AHCJ is an independent, nonprofit organization dedicated to advancing public understanding of health care issues. With more than 1,500 members, AHCJ’s mission is to improve the quality, accuracy and visibility of health care reporting, writing and editing. The association and its Center for Excellence in Health Care Journalism are based at the University of Missouri-Columbia School of Journalism.
The NIH Big Data to Knowledge program has announced a new weekly webinar series beginning Friday, September 9, The BD2K Guide to the Fundamentals of Data Science. The sessions will consist of online lectures given by experts from across the country covering a range of diverse topics in data science. This course is an introductory overview that assumes no prior knowledge or understanding of data science. The series will run through May 19, 2017, at 9:00-10:00 AM Pacific Time. Join from your computer, tablet, or smartphone by visiting: https://global.gotomeeting.com/join/786506213. You may also dial in using your phone: (872) 240-3311, Access Code 786-506-213. Registration is not required. Bookmark the webinar link for easy access! All presentations will be archived and available on the course YouTube channel.
This is a joint effort of the BD2K Training Coordinating Center (TCC), the BD2K Centers Coordination Center (BD2KCCC), and the NIH Office of the Associate Director of Data Science. The initial set of confirmed data science lecturers includes: Mark Musen (Stanford), William Hersh (Oregon Health Sciences), Lucila Ohno-Machado (UCSD), Michel Dumontier (Stanford), Zachary Ives (Penn), Suzanne Sansone (Oxford), Chaitan Baru (NSF), Brian Caffo (Johns Hopkins), and Naomi Elhadad (Columbia).
The National Library of Medicine (NLM) Environmental Health Information Partnership (EnHIP) is a collaboration between NLM and Historically Black Colleges and Universities, a Predominately Black Institution, Hispanic-Serving Institutions, Tribal Colleges and Universities, and an Alaska Native-Serving Institution. A list of EnHIP Member Schools is available, as well as the March 2016 EnHIP Meeting Proceedings. The mission of the EnHIP is to enhance the capacity of minority serving academic institutions to reduce health disparities through the access, use and delivery of environmental health information on their campuses and in their communities. Two member schools are based in the Pacific Southwest Region; Charles Drew University of Medicine and Science in Los Angeles, and Diné College, with various locations in Arizona and New Mexico.
EnHIP began as a pilot project in 1991 as the Toxicology Information Outreach Project (TIOP). During the late 1980s and early 1990s, there were a number of published articles and books that highlighted the adverse effects of environmental hazards on minority and socioeconomically deprived communities. There was a clear need for toxicology and environmental health information to be more readily accessible to health professionals serving these communities. Recognizing this need, NLM launched TIOP to strengthen the capacity of Historically Black Colleges and Universities to train medical and other health professionals in the use of toxicology, environmental, occupational, and hazardous waste information resources. The value and success of the project later led to the longest-standing outreach program of NLM. The name was changed to the Environmental Health Information Outreach Program (EnHIOP) as more schools were added to the program in order to reflect more diversity in the participating institutions. In 2008, the name changed to Environmental Health Information Partnership (EnHIP) to reflect a true partnership with NLM.
Elsevier is sponsoring the inaugural North America Research Intelligence Conference, which will take place at the DoubleTree by Hilton Hotel Berkeley Marina on Thursday, September 29. The conference theme is Enabling Data-informed Strategic Planning for the Research Enterprise. Registration for the conference is free of charge and includes access to every session. Breakfast, lunch, and refreshments will be provided during the day. There is also a Networking Reception on Wednesday, September 28.
Kevin Reed and Alisa Surkis, NYU School of Medicine, are seeking participants to pilot test research data management education materials for medical librarians. This project is currently funded by a grant from the Big Data to Knowledge (BD2K) Initiative at NIH to develop a curriculum for medical librarians to facilitate their teaching research data management at their own institutions. The training modules are open to all librarians, regardless of skill set. There are two components to the training materials:
Part 1: Seven online modules (approximately three hours of content) designed to teach medical librarians about the practice and culture of research and best practices in research data management.
Part 2: A teaching toolkit including slides, scripts, and evaluation materials to teach an in-person introductory research data management class for researchers at your institution.
Kevin and Alisa are currently seeking participants to pilot part 1. Following that, they will seek out a subset of participants to pilot part 2, which will involve structured observations of classes taught by the librarians at their institutions. All participants in piloting part 1 will be given access to the materials in part 2, regardless of whether or not they are part of the piloting of those materials.
Kevin and Alisa have been teaching research data management to medical librarians at the past three MLA annual meetings, based on their experiences in providing research data management services at NYU School of Medicine. Hopefully, the materials created for this project will make the core elements of that class more broadly available to facilitate the teaching of research data management at medical libraries across the United States. Anyone intending to take the modules should contact Kevin or Alisa to confirm participation. There’s no need to wait for a reply to begin taking the modules. They are available to answer questions at any time.
BD2K has announced the following new opportunities:
Proceedings of the two-day Best Practices of Biomedical Research: Improving Reproducibility and Transparency of Preclinical Research conference, held June 9-10, 2016, are now available for viewing through the NIH VideoCasting site:
Reproducibility of biomedical research, which is the ability to conduct projects that lead to the same results multiple times, was the focus of this conference featuring the nation’s leading experts, sponsored by the National Library of Medicine, the Friends of the National Library of Medicine, and Research!America. Discussions included insights from Food and Drug Administration Commissioner Dr. Robert Califf, as well as presentations by John Ioannidis, MD, Professor of Medicine and Health Research and Policy, Stanford University; Christopher Austin, MD, Director, National Center for Advancing Translational Sciences, NIH; and Jon R. Lorsch, PhD, Director, National Institute of General Medical Sciences, NIH.
Topics covered in the conference include:
- The challenge of reproducibility
- Due diligence in acquiring science
- Ethics and institutional responsibility
- Open science and data sharing
- Scientific rigor and open science
- Best strategies for reproducible research
- Best practices of reproducible research