A data-trend log and bar graph display for physiological monitoring
✍ Scribed by Homer R. Warner
- Publisher
- Elsevier Science
- Year
- 1970
- Tongue
- English
- Weight
- 217 KB
- Volume
- 3
- Category
- Article
- ISSN
- 0010-4809
No coin nor oath required. For personal study only.
✦ Synopsis
The task of presenting data derived from measurements of physiological variables on a patient in an intensive care unit to a physician in a form that will best serve him in making decisions about the patient, is a challenging one. The physician would like to see the most recent data in most detail and at the same time relate these measurements to data obtained earlier in time. To accomplish this, a bar graph display is used which represents the variable values as a function of time expressed as ~512 ((I, -t,,) ,'fi), where Ln:! is the interger part of the logorithm to the base 2, t, is the present time, t, is the time of measurement, and ti is the time between measurements (Fig. 1).
Since most variables are not measured continuously in an intensive care ward, time is plotted discontinuously in the form of a series of bar graphs, one for each integral value of the independent variable, Lutz ((t,t,):f;). The height of the nth bar represents the mean value of (n/2) measurements and no measurement is represented in more than one bar. For example, only the most recent measurement is represented by the left most bar, the second bar is the average of the two measurements preceding this one, the third bar is an average of the four preceding these two, etc. (see Table 1). The width of each bar is plotted proportional to the number of intervals the measurement represents. In this way, up to six days of data gathered at the rate of one measurement every 15 min can be displayed on a single plot as shown in Fig. 1 and the detail in the most recently acquired data is preserved. Furthermore, only n + 1 words of core memory are required to store 2" measurements that are to be presented in this form. An algorhythm for arrangement of successive data values in this type of table is presented at the end of the paper and Table 1 shows the contents of each cell in the table after 12 entries. To provide scaling for a bar graph, horizontal lines are drawn through the maximum and minimum bars and labeled, and the short horizontal lines represent the maximum and minimum values contributing to the average represented by the height of each