Informatics Nerd

University of Tennessee, Knoxville researchers, led by Dr. Xiaopeng Zhao, have developed a new algorithm to improve ECGs that are conducted remotely over a smartphone.  First, since I find telemedicine fascinating but don't follow the field closely, I wasn't even aware that you could conduct an ECG on a smartphone.  Knowing that, the algorithm described improves the ability to conduct one remotely (and possibly by someone who is not a trained expert) by attempting to analyze the possibility of noise in the signal, and even tries to provide suggestions about electrode placement. This is amazing, fascinating, and my big question is how soon this can be rolled out into care!  Obviously proper evaluation needs to be done, but this has great implications for providing care in remote settings, not only internationally but even domestically.  The implications for medical research may also be addressed if studies could benefit from field conduction of ECGs. From a techn...

There's an interesting summary on CMIO , about an equally interesting article over at JAMIA regarding design and user acceptance of an EHR over one year.  We should applaud the authors for proving a long-held tenant of developing EHRs - that a well-designed EHR will be well-accepted (and increasingly so) over time.  Part of the explanation for an increased acceptance over time can be attributed to a learning effect, and the fact that the users are getting comfortable using the system.  This holds true especially for well-designed software.  I can only speculate that poorly designed software would also see better (relatively speaking) acceptance as users just get worn down on using poorly designed software.  Definitely another reason to design it right the first time. The tangent from here that my mind takes is that when technology changes and new innovations are made, how do we get users to buy into changing what they've grown to like?  Assuming that the...

After much slacking and shuffling on my part, I've finally gone ahead and signed up for a GitHub account.  The first project I'll be getting updated is the UIMA.NET support library.  The purpose of this project is to allow .NET developers (like myself) to have an easy-to-use assembly for accessing UIMA documents.  This is being developed specifically around cTAKES NLP engine to begin with, so perhaps the title is slightly misleading, but I'm hoping that it will grow over time. Right now the library offers basic functionality for identifying cTAKES objects within a document and being able to extract those items or merge them with other types of objects.  Ideally there will be more advanced LINQ support, but I see that as being a few months out before I'll probably have time to get to it.