1. Origin of the sample ======================= This page contains the instructions to download a sample of normal Nerve Conduction Studies. The sample was obtained at Hospital Carlos III, Madrid, Spain. All the subjects gave written informed consent. The whole procedure was approved by the Ethics Committee of the Hospital Carlos III. 2. Global description ===================== The database comprises 250 files belonging to 50 subjects. The files included are: "Normal_v00.tar.gz" and "Normal_v00.zip": The sample in different compression formats (see README.TXT) "data_v00.txt": An ASCII file containing information on subjects, measurements... (see README.TXT) "README.TXT" An ASCII file containing information on this sample Some additional tools are included "Normal_v00_R.gz": The whose set of values in R format "Edfp2ascii.jar": A program that allows converting EDF+ files in ASCII files (values and annotations) 3. License ========== Contents of this section are accessible under the same Copying policy of Physionet (http://www.physionet.org/copying.shtml). I would like if you could include a reference to http://www.neurotraces.com but it is not mandatory. 4. Description of the files (and how to use them) ================================================= 4.1. Normal_v00.tar.gz and Normal_v00.zip ============================================ Once decompressed it creates a directory called Normal_v00 Each subject (c01 to c50 described as cNN) include - Median Motor Nerve Conduction (cNN_MM.EDF) - Antidromic Sensory Median Nerve Conduction to second finger (cNN_MS2.EDF) - Antidromic Sensory Median Nerve Conduction to first finger (cNN_MS1.EDF) - Antidromic Sensory Ulnar Nerve Conduction to fifth finger (cNN_CS.EDF) - Antidromic Radial Ulnar Nerve Conduction to first finger (cNN_RS.EDF) cNN_MM.EDF and cNN_MS2.EDF files contain records with stimulus at wrist and elbow cNN_MS1.EDF, cNN_CS.EDF and cNN_RS.EDF files contain only records with stimulus at wrist 4.2. data_v00.txt ================= This file contains additional information on the files. Each subject is described in seven lines Motor Median Nerve with stimulus at wrist (key: 1) Motor Median Nerve with stimulus at elbow (key: 2) Antidromic Second Finger Sensory Median Nerve with stimulus at wrist (key: 3) Antidromic Second Finger Sensory Median Nerve with stimulus at elbow (key: 4) Antidromic Fifth Finger Sensory Ulnar Nerve with stimulus at wrist (key: 5) Antidromic First Finger Sensory Radial Nerve with stimulus at wrist (key: 6) Antidromic First Finger Sensory Median Nerve with stimulus at wrist (key: 7) So, the whole file includes 350 lines (50 subjects x 7 traces) Each trace is characterized by the next fields (separated by tabs) file name of the file (e.g. "c01_MM.EDF") rec number of record (0/1) key type of record (1/2/3/4/5/6/7) (see previous paragraph) sex Male or Female (M/F) age Age (in year) height Height (in m) weight Weight (in Kg) BMI Body Mass Index (Weight / Height^2) dist1 Distance between(elbow and wrist) per-I Perimeter of first finger (in mm) per-II Perimeter of second finger (in mm) per-V Perimeter of fifth finger (in mm) temp Temperature at wrist (in degree C) lat1 Onset Latency of negative peak (in ms) lat2 Peak Latency of the first Negative peak (in ms) lat3 first cross by 0 after the first negative peak (in ms) lat4 Peak latency of the fist positive peak (in ms) amp1 Amplitude at "lat2" in relation with "lat1" (in mV-uV) amp2 Amplitude at "lat4" in relation with "lat2" (in mV-uV) area Area of the first negative peak (in mVms-uVms) dist2 Distance from stimulus to recording (in cm) vel Conduction Velocity (in m/s) dif difference Ulnar/Median and Radial/Median (in ms) NA Not available 4.3. Edfp2ascii.jar =================== This is a java program that allows to export data from EDF+ (you need to have java installed). This is an example of how to use it: ~/tmp> java -cp Edfp2ascii.jar Edfp2ascii usage: java Edfp2ascii filename record signal ~/tmp> java -cp Edfp2ascii.jar Edfp2ascii Normal_v00/c01_MM.EDF 0 0 0.0 -0.2 5.0E-5 -0.2 1.0E-4 -0.2 1.5E-4 -0.2 2.0E-4 ... 0.04985 -0.2 0.0499 -0.2 0.04995 -0.2 ~/tmp> java -cp Edfp2ascii.jar Edfp2ascii Normal_v00/c01_MM.EDF 0 1 +0.0Stim_wrist30.2Right Abductor Pollicis Brevis 6.5wrist Notice that the program detects whether the signal is an ordinary signal or an "Annotation". The output of the program can be redirected to any other file for further processing. 4.4. Normal_v00_R.gz ===================== A simple interface of the data contained in the sample with R (http://cran.r-project.org/) Once in "R" you can access to "Normal_v00_R.gz" in this way (it is not necessary to decompress the file): > load("Normal_v00_R.gz") > dim(s) [1] 1000 7 50 A three dimension matrix (1000 values x 7 traces x 50 patients) is loaded at variable "s". The seven traces are (in this order): 1. Motor Median, stimulus wrist 2. Motor Median, stimulus elbow 3. Antidromic second finger Sensory Median, stimulus wrist 4. Antidromic second finger Sensory Median, stimulus elbow 5. Antidromic fifth finger Sensory Ulnar, stimulus wrist 6. Antidromic first finger Sensory Radial, stimulus wrist 7. Antidromic first finger Sensory Median, stimulus wrist Then you can try: > plot(-s[,3,17],type="l") And you plot the third curve of patient 17. Notice that the time scale has to be adjusted to a time increment of 0.05 ms for motor and to a time increment of 0.02 ms for sensory traces. This is an example > plot(c(0:999)*0.05, -s[,1,17],type="l") > plot(c(0:999)*0.02, -s[,3,17],type="l") The first command plots the first trace of patient 17 (a motor response), the second command plots the third trace of patient 17 (a sensory response). You can relate the traces with the information contained in the text file in the next way: > d <- read.table("data_v00.txt", header = T) > names(d) [1] "file" "rec" "key" "sex" "age" "height" "weight" "BMI" [9] "dist1" "per.I" "per.II" "per.V" "temp" "lat1" "lat2" "lat3" [17] "lat4" "amp1" "amp2" "area" "dist2" "vel" "dif" > summary(d[d$key==1,]$sex) F M 31 19 The table has been read to variable "d" 5. Additional information ========================= You can get more tools to handle EDF files in the next sites: - http://www.neurotraces.com/edf (tutorials, examples...) - http://www.hsr.nl.edf (the definition of EDF and EDF+, tools, commercial products...) - http://www.librasch.org (a library that handles EDF+ together with the interface with a lot of programming languages) In order to contact on this and other related issues please write to Jesus Olivan Email: jolivan@neurotraces.com (any suggestion will be welcomed)