Session P73.7

A Graphical User Interface for the Study of Heart Rate Variability

PP Domitrovich*

Washington University School of Medicine
St. Louis, MO, USA

A truly windows-based graphical user interface (GUI) for Heart Rate Variability (HRV) is an easy-to-use, educational, and valuable research tool. The development of such a research tool is desirable, but not a trivial endeavor. One wants to be able to maneuver effortlessly throughout the 24-hour interbeat-interval file, and to obtain many different views of HRV simultaneously.
One begins with the document-view architecture of the Microsoft Visual Studio, which is based upon the Microsoft Foundation Classes written in C++. To obtain simultaneous multiple views, one can implement the dynamic splitter-window class in such a way that 4 panes are provided. Corresponding to each pane, there is one view class that governs the associated display. The splitter windows have moveable bars, allowing resizing of all windows within the main window. The mainframe class governs the main window, and the splitter-window class must be instantiated in the mainframe class, along with all toolbars attached to the main window. The document class is where one serializes the input into the program. In this case, an interbeat-interval file is serialized by rewriting the time, beat label, and RR interval as a structure, and then creating a binary file of a series of these structures. The structure also contains an additional element of information defining the segment to which the beat belongs. The entire 24-hour file is normally divided into 5-minute segments, although 359 various segment lengths are allowed. HRV is viewed per segment in this program, and movement forward or backward by 1 segment and by 1 hour is required. Within the document class, the input file is serialized, and de-serialization is implemented segment by segment upon demand. Data for a particular segment is extracted and dumped to a file to be shared with the 4 views and the mainframe. The document class controls the effects of button presses in the additional toolbar that controls movement through the various segments. HRV is dealt with in 2 classes: a general HRV class, and an HRV class for spectral analyses. The 4 views instantiate the HRV classes and produce graphical HRV output. Each view contains a set of various programs, and a total 270 different 4-paned views are possible. Utilizing property sheets and dialogue boxes, 59 adjustable parameters are available. One view contains an interactively editable heart-rate-arousal detector and heart-rate-arousal duration detector. The Lomb periodogram, the fast-Fourier transform (FFT), Poincaré plots, heart-rate tachograms, HRV in text format, phase-plane plots, multi-line heart-rate tachograms, and 2-minute-averaged Lomb and FFT plots are available. A unique representation of approximate entropy is also available.
The HRV output of this program has been utilized in the publication of several abstracts. A sophisticated application built into this program has also produced several abstracts.

(Abstract Control Number: 189)