Search this site

    Potentially disruptive platforms

    Potentially disruptive process platforms are a number of SOPs associated with methodologies and applications to improve productivity and quality in biomedical research and technology. They are characterized by being a springboard toward a leap in productivity and quality in biomedical research and technology that can be applied to multiple content areas. These disruptive platforms were either developed within the Research on Research Network or developed elsewhere and then applied to biomedical research and technology within our network.

    1. Blending - Blending Framework is a series of observations regarding methods to create potentially disruptive process platforms for research and technology. The method is based on a series of sequential steps, starting with an evaluation of current resources available for research and technology such as skills, funding, problem/opportunities, and the individuals connected to each of these resources. Subsequent steps involve the blending of overarching themes, creation of the so called blending narrative, and multiple iterations over time to refine the process. The Blending Framework is described in an upcoming e-book released through Amazon's Kindle.
    2. Cloud writing - Cloud writing using tools such as Google docs allows for researchers and innovators from all over the world to work on the same document, having the latest version available to them 24/7. This allows different tasks to be conducted in parallel, such as one researcher working on data analysis, another on building the Introduction and Methods, and yet another working on something like the administrative aspects of the proposal.
    3. Data analysis toolkit - originally based on the concept of layers of Information Pietrobon et cols. input - variables that go in, methods p&p; output - interpretation (text) of tables, figures, and computer output; references - progressive order difficulty, problem-based based on enactive cog principle; 
    4. Database standardization - Database standardization allows for different studies to collect information on clinical variables using an internationally accepted terminology. Advantages include the potential for exchange of data sets across collaborators for larger for studies with enhanced statistical power and generalizability. For more information see the caDSR site.
    5. Global language and iterative translation - Global language is a standard way of writing texts in English as well as other languages that allows for a more efficient understanding and automated translation from other languages into English. In the Research on Research Network we have been combining these standards with automated translation using platforms such as the Translator Toolkit for scientific articles, funding proposals and patent submissions, with good results for some languages. For more information see the The Global English Style Guide.
    6. InnovationMatch - InnovationMatch is a platform to enhance the commercialization of research-based technologies, constituted by two components. First, it is composed by a database containing publicly available information about research and technologies created within universities, information about corporate and start-up companies, government programs to incentivize innovation, as well as information on potential investors. Publicly available information is enhanced by non-publicly available information from each of the previously mentioned stakeholders. This database is then used to create a match among stakeholders with similar interests towards commercialization. Second, this initial information exchange is used for in-person contacts among all stakeholders to move commercialization forward. For more information see the slides summarizing the project.
    7. Leveraging combined biomedical research and technology - Leveraging research from technology is a series of best practices to simultaneously advance a combined research and technology program. See details from the slides on this platform.
    8. Linked Open Data and database matching - This platform focuses on augmenting biomedical databases with variables from other databases containing primarily behavioral and socio-demographic variables. Based on Linked Open Data principles, this method does not require that the patients be common across the two databases. This system is currently in initial development, but its principles are already being applied to an initial biomedical database.
    9. Manuscript templates - Manuscript templates describe a common argument structure for different sections of research manuscripts as well as funding proposals. It assumes that the Introduction and the Discussion section of a manuscript are relatively stable across manuscripts, while the Methods and Results sections are design-driven. It takes as its basis a mix of existing international reporting standards as well as a series of action research experiments conducted within the Research on Research Network. For more information see the article by Pietrobon et cols.
    10. Hangout video conference - Google Hangout is a freely available video conference that allows for up to 10 participants to share video, voice, and screen sharing. It is also possible to include other people through calls made through the phone (free within the US, discounted prices outside). Video recording is usually done through an external software such as Camtasia or Screenflow and then privately posted through a social network such as Google Plus.
    11. Mechanical Turk - Mechanical Turk, or mturk for short, is a Web application created by Amazon.com that allows researchers to conduct a series of studies among normal subjects from the general population, with very fast recruitment and at minimal cost. Within our group we have used this platform for behavioral studies analyzed through Structural Equation Modeling, scale validation using confirmatory factor analysis and Item Response Theory, enhancement of existing databases through Linked Open Data, experimental behavioral studies, preference measurement for cost-effectiveness studies, and the testing of health games. For more information, see the article by Buhrmester et cols.
    12. Question Diagram - Question Diagram, a platform to facilitate the design of hypothesis-driven biomedical research questions. Its role is to both streamline the communication of research questions among the participants of a given project as well as to ensure that a research question is internally consistent. For more information see the article by Pietrobon et cols.
    13. R - R language is an open-source, object-oriented statistical language with a variety of packages addressing cutting edge methods in statistics and data mining. Of importance, R plays a very important role in reproducible research, one of the main contemporary movement to ensure that one can not only access the data behind research papers but also reproduce the results as displayed in the final manuscript. For more information see the R site.
    14. Reproducible research - is a research movement in response to recent articles demonstrating that research studies tend not to be reproducible even when their data are made available (Ioannidis, 2008). In response to this problem, the reproducible research movement proposes methods such as embedding analytical scripts within the final manuscript using methods such as sweave or Emacs org.mode while making data sets available in publicly available repositories such as dryad.
    15. Scientific dissemination integrated with research and technology networks - Recent studies have demonstrated that there is a significant discrepancy between how scientific studies are communicated to scientific peers and how they come across to the general public -- see Gonon, 2011 for one out of many examples. This phenomenon can occur for a variety of factors, among them language that is not clear by not using simplifying mechanisms such as metaphors, or the misuse of uncertainty language giving the impression of more certainty about a scientific finding than researchers believe should be warranted. Our group is now developing a combination of large, detailed corpora of metaphors and uncertainty language in biomedical research, to be followed by cognitive experiments and ultimately a distance course on scientific dissemination for researchers and science journalists.
    16. Shared calendars - With the increasing popularity of distributed research and technology teams across not only different locations within a campus but also different locations around the world, scheduling times for meetings can take a substantial amount of time. Applications such as Google Calendar allows for meetings with multiple participants across different time zones to be schedule at time where all are free within less than a minute. 
    17. Social networking for internal project communication and broad scientific dissemination - Communication within large research groups is always a challenge, with multiple ongoing projects going on while most people are not even aware that they exist. Move that to the context of a distributed research network and the problem is compounded. The same communication issue happens when researchers attempt to communicate with the general public. This is why we have been using more and more standardized tools for communication within our network and also reason to the general public. Examples for both publics include the use of Google Plus, which allows for quick posts -- public or targeted at a specific group -- along with images and videos. A second mechanism is Google Sites, which allows you to have a very functional site in literally a few minutes. 
    18. Targeted literature reviews -  originally based on the concept of literature matrices Pietrobon et cols. tag with working project, manuscript section, and text block - create link and add to manuscript section as comment annotate pdf only for relevant sections, characteristics: ability to work in groups, extract reference from any source, storing and sharing jointly annotated full text papers, automatic generation of manuscript references  
    19. Text messaging and social network analysis for follow-up and compliance - 
    20. Training and associated head hunting - dual role high performance researchers -- objective is to create people that will both focus on a given biomedical research area as well as focus on the creation of PDP
    21. Virtual Delphi - The objective of the traditional Delphi method is to get a group of people who have experience with a certain topic and reach a consensus regarding a certain aspect of this topic. An example would be to have a group of healthcare providers to agree on the most commonly affected activities of daily living impaired among burn patients. This approach is usually conducted using only surveys, thus making it difficult to initiate a discussion on the underlying reasons for the lack of agreement. Virtual Delphi is a PDP where the meeting is held on a video conference where discussions can be held in real time, while voting is still anonymous. Meetings are recorded for further mixed method evaluation, and real-time graphics display the latest level of consensus within that group.
    22. Workflow dissonance classification - Today it is a common expectation that the implementation of systems for electronic capture of clinical data will lead to a significant slowing the clinical workflow. Be they electronic data capture systems for trials and registries, or electronic health records for clinical care, clinical pace on the clinic is invariably affected. This slowing is usually also accompanied by lower data quality as fields are skipped or entire blocks of data collection left blank altogether. Within this platform, we propose that one of the main underlying causes behind these problems is a dissonance between clinical and data collection workflows. A methodology for workflow retrieval and a dissonance classification were proposed and are now being applied to different scenarios to improve quality and productivity.