Research
RDM encourages research projects.If you are thinking of or are using a CFAM for a research project then advice is free. Please telephone if possible.
If you are considering making use of the on/off-line data mapping procedures of the CFAM3(c) then please contact us to check that the data categories you are thinking of using are appropriate to the display formats available. In some cases free data processing is available to check that your categories are viable. In special circumstances additions may be made to the standard software to cater for a particular need.
The advantage to us is that this enables us to see where the system can be improved.
Confidential pre-submission review of papers to check for correct technical content concerning the monitor is free.
Most people wish to keep their research projects confidential. However, if you wish, we can contact other CFAM users who may be interested in collaboration.
Research projects (general)
These may be broken into two main groups.
Firstly, studies based on a small number of subjects, in which two or more data classes are tested for a statistical difference. In this type of study as few as ten subjects may sometimes be used.
Secondly, studies based on a large number of subjects, in which the intention is to be able to discriminate reliably between several data classes. In this type of study there may be 20 to 30 examples in each data class.
The CFAM software assists both types of project, although the first type will normally best be served by exporting data to conventional statistical software packages. It may be undertaken as a preliminary investigation prior to undertaking the second more detailed type of study discussed below.
To give an example of the type of result that is aimed at, look on this web-site under "fetal". There you will see a data map in which a recording from a fetus during labour is compared against a two-dimensional database , state of awareness versus gestational age, obtained from normal neonates. To generate the data base, a research worker took CFAM recordings from neonates of various gestational ages. Their state of awareness was assessed visually (i.e. not from the CFAM record), and their gestational age was taken from the clinical notes. Sections of CFAM recordings, in which the state of awareness could be determined reliably, were selected in the replay software and their statistics placed in a database under paired headings, state of awareness and a specific gestational age range.
If, in your database, you want a range of say five increments along each axis, then you have 5x5 = 25 data categories to fill. Assuming that you want 20 examples in each category, this means 500 examples have to be found.
Of course, in the given example, a compromise has been made. Although sections of record at a particular state of awareness may be reliable in that context, the gestational age presents a problem. Firstly is it accurate? Secondly, if we place records in gestational age groups each covering a 2 week range then, at the margins between two gestational age ranges, a small error in estimating that age can place the neonate in either of two groups whose mean age differs by two weeks. The CFAM patented map generation software copes with this by not assuming that data are all at the centre of their category ranges.
Data Categories
Sections of records may be placed in data categories. These may later be used to generate maps as in the fetal example mentioned above. The end-user defines the categories and generates a category menu via the mnu_bild programme. Each line in the menu consists of an up to 8 character code followed by a description. When choosing the code to be used please bear in mind that, if at some time you choose to put your database on-line to a patient, then the CFAM will print the code of the nearest matching category in the left hand margin of both the display and the paper trace (if the chart is switched on). Therefore you may not want the code to be understandable to the casual observer.
Eight characters enable you to code an enormous number of different categories. For example, if you use one character to define age ranges, you have 26+10=36 alpha-numeric characters to choose from and hence 36 different age ranges can be defined. Similarly, for sedation you may have a variety of drugs and a variety of infusion rates. It may be that these would fall into a 3x12 or a 6x6 or even a 6x3x2 grid, in which case a single (cryptic) character would suffice to encode these. You may find that a combination of immediately meaningful and cryptic characters will suit you best. For example AN36 may mean awake, normal, 36 weeks gestational age. The first two characters may be from a simple cryptic set, the age isn't. This still leaves you four more characters that can be used cryptically.
As your research project progresses you will almost certainly find that you want to create new categories and new category menus. Don't worry too much about this, the CFAM software allows you to both share and move data between multiple categories and multiple menus.
