Chapter 4 - RUNNING THE LTP PROGRAM

4.1 Automatic and Manual Loading of the Protocol (*.pro) File from Disk

If you have already developed custom protocols to run your particular experiment, the last protocol file used will be the one automatically loaded when the program is later restarted. However, if there was never any protocol file saved, then only the integral default values will be initially operating.

Fig. 4.1.1. Protocol File Open Dialog Box. Note that the top line shows the present protocol file.

4.2 Choosing the Basic Protocol

4.2.1  Slow Repetitive Pulse Sweeps

Note that if either the "P0 or P1 Sweeps", or the "Alternating P0 and P1 Sweeps" protocol is chosen, extra Windows are formed for a Train T1 Stimulation Window and a Pulse T1 Stimulation Window.

In normal LTP experiments, a Pulse P0 Sweep normally generates S0 pulse stimulation only, and Pulse P1 Sweep normally generates S1 pulse stimulation only.  Therefore, Slow Repetitive P0 Sweeps produces slow repetitive S0 pulse stimulation, and Alternating P0 and P1 Sweeps produces alternating alternating S0 and S1 pulse stimulation. 

4.2.2  Fast Repetitive LTD Stimulation

In normal LTP experiments, a Pulse P0 Sweep generates S0 pulse stimulation only, and Fast Slow Repetitive P0 Sweeps produces fast repetitive S0 pulse LTD stimulation.  Similarly, a Pulse P1 Sweep generates S1 pulse stimulation only, Fast Slow Repetitive P1 Sweeps produces fast repetitive S1 pulse LTD stimulation.

4.2.3  Enable Train Stimulation

In normal LTP experiments, a Train T0 Sweep generates many (say 100) S0 pulse stimulations, and a eliciting a single Train T0 Sweep would deliver an S0 train stimulation to the preparation.  Similarly, a Train T1 Sweep generates many S1 pulse stimulations, and a eliciting a single Train T1 Sweep would deliver an S1 train stimulation to the preparation.  

As described in Section 4.5, the stimulations in Pulse P0 and P1 Sweeps and Train T0 and T1 Sweeps can be quite varied.  Forinstance, a Train or Pulse Sweep can generate repetitive trains (theta burst) and primed burst stimulation.  In fact, a Pulse Sweep can generate train stimulation, and a Train Sweep can generate single pulse stimulation.

4.2.4  Signal Averaging, Stimulus Artifact Blanking, Digital Filtering

In addition to capturing and analyzing raw sweeps, the LTP Program can also do on- and off-line signal averaging of these sweeps, blank out the stimulus artifacts if required, and  filter the sweeps (Fig 4.2.4).

Single raw sweeps can either be (i) filtered, (ii) stimulus artifact blanked, or (iii) stimulus artifact blanked and then filtered (Fig. 4.2.4a) (but not first filtered and then stimulus artifact blanked).  The insets in Fig. 4.2.4a show a patch-clamp recording of an EPSC from one raw sweep (left trace) showing substantial noise and a large stimulus artifact at the left of the trace, the sweep that has been digitally filtered to reduce the noise (note the large filtered artifact, right top trace), the sweep with the stimulus artifact removed (middle trace), and the stimulus artifact blanked sweep that has then been filtered (right bottom trace).

Alternatively, raw sweeps can first be (i) signal averaged, then this signal averaged sweep can either be (ii) filtered, (iii) stimulus artifact blanked, or (iv) stimulus artifact blanked and then filtered (Fig. 4.2.4b) (but not first filtered and then stimulus artifact blanked).  The insets in Fig. 4.2.4b show EPSCs from two raw sweeps showing substantial noise and a large stimulus artifact (left traces), the signal averaged sweep also with a large stimulus artifact (2^nd trace) obtained from the two raw sweeps, the signal averaged sweep that has been digitally filtered (right top trace), the signal averaged sweep with the stimulus artifact removed (3^rd trace), and the averaged, stimulus artifact blanked sweep that that has been filtered (right bottom trace).

All these traces can be shown on the screen and saved to a data file, and almost all can be plotted on a laserjet printer (raw sweeps cannot be plotted during signal averaging).  Calculations of slopes and peaks are made on the latest processed sweep.  For example, if signal averaging, stimulus artifact blanking and digital filtering are being used, then the averaged, blanked and filtered sweep is the one that will be analyzed. 

Fig. 4.2.4.  Raw sweeps can be signal averaged, stimulus artifact blanked, and/or digitally filtered.   a)  Raw sweeps (with no signal averaging) can be digitally filtered, stimulus artifact blanked, or stimulus artifact blanked first and then filtered.  The insets show the raw sweep (left trace), the  filtered sweep (right top trace), the stimulus artifact blanked sweep (middle trace), and the blanked and filtered sweep (right bottom trace). b) Raw sweeps can also be signal averaged, then filtered, stimulus artifact blanked, or stimulus artifact blanked first and then filtered.  The insets show two raw sweeps (left traces), the signal averaged sweep obtained from the two raw sweeps (2nd trace), the averaged and filtered sweep (right top trace), the averaged and stimulus artifact blanked sweep (3rd trace), and the averaged, blanked and filtered sweep (right bottom  trace).

4.2.4.1  Signal Averaging

If Signal Averaging is chosen each sweep is first acquired and plotted in Grey, then the average of the sweeps is then plotted in LightBlue. The setting for the number of sweeps to average is set by the NumSweepsAvg field (see Fig. 1.4.1C; the Num Sweeps Avg field = 4 and is just below the SlowPulsePer field in the Pulse S0 Stimulation Window).

You can also signal average during fast repetitive LTD stimulation which is set by the Num Sweeps Avg field (also see Fig. 1.4.1C; this Num Sweeps Avg field = 20 and is just below the TotNumSweeps field in the LTD section of the Pulse S0 Stimulation Window).

4.2.4.2  Stimulus Artifact Blanking

Fig. 4.9.1b,c and Fig 4.2.4.2 show the effect of stimulus artifact blanking.  If Stimulus Artifact Blanking is chosen, stimulus artifact blanking is enabled, and a S0 or S1 Blank field appears on the Detection page (see Fig. 4.9.1b,c).  Beginning at the time of the start of the S0 or S1 pulse to the time set in the S0 or  S1 Blank field (1 ms in Fig. 4.9.1), the stimulus artifact data in the ADsweep array is (sometimes) set to the Average of the 1 data point value just before blanking begins, and one point value just after blanking ceases (see Fig. 4.9.1c and Fig. 4.2.4.2a).  If the time in the S0 or S1 Blank field is set too short, then the stimulus artifact is not entirely blanked, if it is set too long (but not rediculously long) generally there aren't any problems as long as the peak amplitude is not clipped.

LTP22A only took the Average of the point just before blanking begins and the point just after blanking ceases (see Fig. 4.9.1c and Fig. 4.2.4.2a).  However, in LTP24, I also added the capability of taking the Slope between the point just before blanking begins and the point just after blanking ceases (Fig. 4.2.4.2b), and also taking the point just before blanking begins and then Hold that voltage until to the point just after blanking ceases (Fig. 4.2.4.2c).  Setting the Average, Slope or Hold vlanking method is accomplished by entering A, S or H into the Average/Slope/Hold field next to Blank (Fig. 4.2.4.2).  

Stimulus artifact blanking is useful for determining the peak amplitudes of individual EPSPs when the stimulus artifacts are riding on top of the previous EPSP (eg Fig. 4.9.1b), or when trying to determine the area or peak amplitude of a whole train which could be seriously distorted by the stimulus artifact (eg Fig. 4.9.2).

Stimulus artifact blanking has also turned out to be useful when trying to fit exponential curves to the decay phase of closely spaced EPSCs when the artifact for the next EPSC occurs during this decay phase.  This approach has been used for exponential curve fitting of synaptic trains by the Synaptosoft MiniAnalysis Program.  It's interesting how removal of stimulus artifacts allows EPSPs and EPSCs to be analyzed as spontaneous events.  The Slope method has proved particularly useful for fitting exponential curves to synaptic potentials by having a smoother decay phase; in fact it is almost impossible to see where the stimulus artifact was.  

The Hold method has proved useful for blanking stimulus artifacts where the point just after blanking ceases varies widely (such as blanking the stimulation in extracellular CA3 recording where the antidromic spike occurs right after the stimulus artifact).

a Average
b Slope
c Hold

4.2.4.3  Digital Filtering

Digital filtering is done using a Gaussian digital filter (Colquhoun D, and Sigworth, FJ,  Fitting and statistical analysis of single channel records.  In B. Sakmann and E. Neher editors.  Single Channel Recording.  Plenum Press, London, 191-263, 1983).  

If Filtering is chosen, each sweep is first acquired and plotted in Grey, and then digitally filtered at an appropriate frequency and plotted in LightBlue. The setting for digital filtering frequency is set by the Filter cutoff frequency field (-3 dB) in the Pulse Detection Window (Figure 4.2.4.3).  Note in Fig 4.2.4.3 and in the filtered traces in Fig. 4.2.4 how filtering the stimulus artifact can really distort the waveform around the stimulus artifact, and that blanking the stimulus artifact removes the distortion around the artifact.

Digital Filtering can be very useful for reanalyzing Peak Amplitude because the algorithm to measure peak amplitude merely picks out the most positive or negative ADsample. Therefore, if there is a lot of noise, the Peak Amplitude measurement will be artificially increased by the extraneous noise. Appropriate levels of digital filtering can reduce this noise and give a more accurate Peak Amplitude measurement.

Digital filtering shouldn’t be necessary for analyzing the Slope since the slope is a linear regression line (least squares fit) through the data points. Interestingly, digital filtering can reduce the sweep-to-sweep jitter in the slope data points, possibly because the digital filtering effectively extends the time over which the slope is calculated by including sample points before and after the time at which the slope is calculated.

When digital filtering is on, the amplitude/slope calculation analysis is done on the filtered ADsweeps only.

4.2.5  Setting Which Sweeps to Save to disk

4.3  Set the Number of AD Channels

The LTP Program can acquire data on one channel (AD0 or AD1), or on two channels (AD0 + AD1, see Fig. 4.8.1b)

Then set the AD data acquisition channels to AD0, AD1, or both AD0 and AD1 in the AD Channels Used Dialog Box (Fig. 4.3.1) by pressing the SPACE bar on the check boxes.

4.4 Set the Data Acquisition Values

4.5 Choosing Pulse/Train Stimulation Protocols and Setting Stimulation Values

4.5.1  Choosing the Stimulation Protocol

This will bring up the Extracellular and Intracellular Stimulation Protocol Dialog Box (Fig. 4.5.1). The stimulations that can be controlled by this dialog box can include: S0 and S1 extracellular stimulation (organized in terms of pulses, trains and dual trains), Digital Outputs (DO0, DO1 and DO2) and IntraCellular (IC) Analog Output stimulation organized as epochs (e.g. up to 6 sequential pulses, PulseA to PulseE). See Fig. 4.5.1 for almost all the possible types of stimulation output shown (although to use these all at once would be rediculous).

where in this case Sweep?? S0 stimulation would produce one or more trains and S1 stimulation would produce one or more pulses.  (Note: For this version of the LTP program (LTP24), there must be at least one S0 or 1 S1 pulse in each sweep (APPENDIX B.2).

where in this case Sweep?? would be producing Intracellular epoch-like stimulation and Rm stimulation; there would be no Digital stimulation.

So, for example, with the default start-up (no protocol file loaded), Fig. 3.1.1 shows train stimulation where TrainSweep T0 has S0=Trains, S1=None, and Intracellular, Digital and Rm are Off (see also the Stimulus Protocol Dialog Box, Fig. 4.5.1). The common 1 sec, 100 Hz train stimulation on S0 would be produced if TrainSweep T0 was elicited by Ctl-F8.

In another example with default start-up (no protocol file loaded), Fig. 3.1.2 shows paired-pulse stimulation and testing for Rm/Rs resistance where PulseSweep P0 has S0=Pulses, S1=None, Intracellular=Off, Digital=Off and Rm=On (see also the Stimulus Protocol Dialog Box, Fig. 4.5.1). This would produce paired pulse S0 stimulation also with the intracellular test pulse for Rm/Rs measurement when the PulseSweep P0 train was elicited by pressing either Ctl-F5 or Ctl-F6.

4.5.2  Useful Examples of Stimulation

Full stimulation capability.  Fig. 4.5.2 shows the full capability of P0 Sweep stimulation (but T0, T1 and P1 Sweep stimulation have the same capability). S0  has DUAL TRAINS, S1 has TRAINS, and Intracellular, Digital and Rm stimulations are all On in the Stimulus Protocol Dialog Box.

Fig. 4.5.2.  Complex stimulation that the LTP Program can generate.  This sweep generates simultaneous extracellular primed burst stimulation (S0, top trace, yellow, see “S0 DUAL TRAINS” fields), extracellular theta burst (repetitive train) stimulation (S1, second trace down, magenta, see S1 TRAINS fields), and intracellular analog stimulation (IC, white bottom trace, see “ICELL” fields) (with depolarization occurring during the S0 primed burst stimulation, hyperpolarization occurring during the S1 theta burst stimulation, and a depolarizing Rs/Rm pulse at the end of the trace to measure cell resistance and patch electrode series resistance cell resistance(see “Rm,RS” fields)).  It also generates digital Sync pulse (DO0 and DO1, white traces) and long Pulse output (D2, white trace) (see "S"  and "P" “DigOut” fields).

Fig. 4.5.3 shows that pathway independence can be tested by comparing heterosynaptic paired pulse stimulation with one pulse on S0 and one pulse on S1 with the effects of homosynaptic paired pulse stimulation (see Fig. 3.1.2). Fig. 4.5.3 is produced by Pulses on both S0 and S1, and Intracellular, Digital and Rm stimulations are Off in the Stimulus Protocol Dialog Box. 

Theta burst stimulation.  Fig. 4.5.4 shows a theta burst stimulation capable of inducing LTP. Only S0=Trains is on in the Stimulation Protocol Dialog Box, and just set the NumTrains greater than 1 to get repetitive train stimulation.

Fig 4.5.5 shows Dual Train stimulation on S0. Only S0=DualTrains is on in the Stimulation Protocol Dialog Box. The three dual trains is set by setting the NumTrainLoops=3.

Primed burst stimulation.  Fig. 4.5.6 shows primed burst stimulation capable of inducing LTP. Only S0=DualTrains is on in the Stimulation Protocol Dialog Box, only one three dual train is produced(by setting NumTrainLoops=1), and the first train has only one pulse (NumPulses=1).

Fig. 4.5.7 shows how intracellular stimulation can be coincident with extracellular pulse and train stimulation. In this example S0=DualTrains, and Intracellular and Rm stimulation are On in the Stimulus Protocol Dialog Box. In this example an intracellular depolarization occurs during the one priming S0 pulse, and hyperpolarization occurs during the S0 train. Intracellular voltage changes could have interesting effects on the ability of extracellular pulses and trains to induce LTP or LTD.

Single Sweep LTD Stimulation.  Fig. 4.5.8 shows a way of inducing LTD by generating a single, 3 minute long, 5Hz, 900 S0 pulse LTD stimulation sweep. Using a single long sweep with many pulses to get LTD stimulation is a way of ‘working around’ a limitation of LTP24 that allows usually only 1-2 Hz repetitive short stimulation sweeps on most Pentium and 486 computers. Note, however, that in LTP24 the number of S0 or S1 stimulus pulses per sweep is limited to 1000.

After you have chosen your basic Stimulation Protocol in the Stimulation Protocol Dialog Box (Fig. 4.5.1), then you have to set the Stimulation Field Values in TrainSweeps T0 and T1, and PulseSweeps P0 and P1. Experiment.

It is important to realize that there is a reciprocal relationship between the TrainSweep or PulseSweep ADinterval (see Fig. 3.1.6) and the maximum Sweep Time (see upper left field in Fig. 4.5.8) that is allowed. For instance, if we assume that a maximum of 100,000 samples/sweep (on an 16 MB computer) can be obtained, then if the PulseSweep ADinterval is set to 100 usec, then the maximum PulseSweep time will be 10 sec. To extend the maximum PulseSweep time to 20 sec, the PulseSweep ADinterval must be increased to 200 usec. This is because the program always outputs an analog output voltage every 100 usec, but it may acquire a sample only every 100, 200 up to 50000 usec.

Therefore, it is important to understand that what you see in the Train and Pulse Stimulation graphs is what stimulus pulses you will generate. If you want to generate a repetitive stimulation and not all of it will ‘fit’ in the stimulation graphs, increase the ADinterval in the Miscellaneous Window. If not all of your stimulus pulses ‘fit’ in the stimulation graphs, no error message will be produced to tell you that. It is up to you to see that this doesn’t occur.

Finally, it is important to realize that extracellular S0 stimulation on the Digidata 1200 (AnalogOut0) and Labmaster (DA0) boards can also be either Monophasic or Biphasic. This is set by the Monophasic/Biphasic Extracellular Stimluation Dialog Box (Fig. 4.5.9). Monophasic is the default condition, so that when the S0 PulseVolt is set to +5.0 volts and the PulseDur is set to 0.1 msec (see Fig. 4.5.8), then AnalogOut0 or DA0 produces a 0 to +5.0 volt pulse lasting 0.1 msec  before returning to 0 volts. However, if S0 is set to Biphasic, then when the S0 PulseVolt is set to +5.0 volts and the PulseDur is set to 0.1 msec, then AnalogOut0 or DA0 produces a 0 to +5.0 volt pulse lasting 0.1 msec which is followed by a -5.0 volts also lasting 0.1 msec before returning to 0 volts. When the PulseVolt is negative, a biphasic pulse would be first negative and then positive. The true S0 stimulus duration in the Biphasic mode is twice that stated in the PulseDur. Note that even though the Monophasic/Biphasic dialog box shows that S1 can be biphasic, this capability is not present in LTP24.

4.6 Choosing the Waveform Calculations To Do

After setting the sweep stimulations, you have to decide which calculations you want to do. To do this call up the Amplitude/Slope Calculations To Do Dialog Box (Fig. 4.6.1 when using channel AD0 or AD1, or Fig. 4.6.2 when using both AD0 and AD1). On the right the dialog box shows which calculations can be done: DC Baseline before of each Stimulus Pulse, the Peak Amplitude, Peak Latency, Area, Duration, 10-90% Rise Time, 10-90% Decay Time, Coastline, PopSpike Amplitude, PopSpike Latency, Slope and/or Average Amplitude of synaptic potentials produced by S0 and S1 stimulation, and if the Rm pulse is On, cell resistance (Rm) and patch pipette series resistance (Rs).

Fig. 4.6.1. Dialog Box for Amp/Slope Calculations To Do for the single Channel AD0 (there is a similar dialog box for AD1)

Fig. 4.6.2. Dialog Box for Amp/Slope Calculations To Do for 2 AD channels.

The top line in the dialog box (under the title) shows where this calculations can be plotted or displayed. MainPg determines whether none, one or two Calculation graph will be presented in the Main ViewPage Window. For example, a Slope calculation checked (‘X’ means On) produces a Slope Calculation graph in the Main Page Window (see Fig. 3.1.4). More than two MainPg calculations checked will not be accepted.

CalcPg0 and CalcPg1 determine which calculation graphs will appear on these pages, for example, Fig. 3.1.5 shows is the result when PeakAmp and Slope of CalcPg0 are checked. None to four calculations per CalcPg can be checked. More than four calculations will not be accepted. If no calculations are checked on a CalcPg, the CalcPg will not appear.

The CalcLine determines which calculation values from the last sweep will be presented on the Calculation Line. None to three calculations can be checked. More than three calculations will not be accepted.

If any of the locations (e.g. MainPg, CalcPg0, CalcPg1 or CalcLine) appear on the row of the 'Calculations To Do' (e.g. DC Baseline, Peak Amplitude, Peak Latency, Area, Duration, Rise Time, Decay Time, Coastline, PopSpike Amplitude, PopSpike Latency, Slope, Avgerage Amplitude, Rs and/or Rm), that calculated value will also be saved to the Amplitude/Calculation file (*.AMP) (see Section 4.11, Fig 4.8.1c and Section 4.14.3). Otherwise the value will be set to 0 in that file.

Version 2 of the 'LTP' program (LTP24) can now analyze all S0- and S1-evoked synaptic responses in all channels in both Pulse and Train sweeps (see Section 4.8).

4.7 Setting the Calculation Detection Criteria

The fields that set the ranges for detecting the various calculations are set in the Pulse Waveform Detection Window (Fig. 3.1.3). However, only if at least one of the MainPg, CalcPg0, CalcPg1 or CalcLine check boxes is chosen for a particular waveform calculation row in the Amp/Slope Calculations To Do dialog box (Figs. 4.6.1 and 4.6.2) will the particular calculation be plotted and saved to the Amplitude/Calculation file as a non-zero value.

4.7.1. DC Baseline

If DC Baseline, Peak Amplitude, Peak Latency, Area, Duration, Rise Time, Decay Time, Average Amplitude, or Slope (for Low% -> High% Peak Amplitude) is chosen, then the DC Baseline value will be calculated.

time fields shown in the Pulse Detection Window in Fig. 4.7.2 set the pre-stimulus pulse baseline to be between these two ‘Baseline’ time values, and both are relative to the stimulus pulse.

4.7.2 Peak Amplitude and Peak Latency

and is shown by the PkAmp solid line of the Pulse Detection Window in Fig. 4.7.2. The first ‘Peak’ time value must be before the Peak Amplitude, and the second ‘Peak’ time value must be after the Peak Amplitude, and both are relative to the stimulus pulse.

The A/P/N field determines whether the peak will be Automatically (A) determined to be positive or negative, forced to be Positive (P), or forced to be Negative (N). The normal value is A, automatic. Automatic calculates whether the average of the points between the Peak: ___ to ___ms after pulse time fields is more positive than the baseline average value. If so, the peak is positive; otherwise the peak is negative.

Peak Latency is the time between the occurrence of stimulus pulse and the peak.

4.7.3 Slope

The Slope is calculated by taking all the waveform voltage/current points from the slope beginning time point to the slope end time point, and using these points to calculate a linear regression line (least squares fit) through the data.

to bring up the Slope Calculation Methods dialog box with the following choices (Fig. 4.7.3).

Fig. 4.7.3. Slope Calculation Method Dialog Box.

where the first % field is the Low% field and the second % field is the High% field.

Both methods have their advantages. If the latency between the stimulus pulse and the slope shifts with time, the Low% -> High% Peak Amplitude method is best. However, when the EPSP/EPSC amplitude approaches 0, the Low% -> High% method begins to calculate slopes made of noise and therefore gives spurious result. In contrast, the Begin -> End Times method continues to accurately measure the slope when the EPSP/EPSC amplitude approaches 0. In general, because the latency of the EPSP/EPSC normally does not change much, and because the amplitude of an EPSP/EPSC can often go to 0, the , the Begin -> End Times is usually the method of choice and is the default value.

Slope detections in our group are usually of 0.6 to 2.0 msec duration. When sampling every 100 usec, this is 7 to 21 AD samples, respectively. Obviously the longer the slope duration the better, provided the slope still remains on the ‘straight’ part of the EPSP/EPSC. On-line signal averaging also decreases slope error measurement.

I have been particularly concerned that the AD sample-to-sample time jitter (on the order of 2-4 usec, APPENDIX B.4) could affect the slope measurement. However, this jitter is essentially random, and I think that the random sample-to-sample jitter would essentially add white noise to a sloped voltage or current acquired with no sample-to-sample jitter. Because a linear regression (least squares fit) line is calculated for these points, the sample-to-sample jitter should have very little effect. By comparison, it is obvious when analyzing/reanalyzing data that any 50/60 Hz line noise in the signal clobbers an accurate slope measurement.

Finally, check that the slope calculation is working correctly. Now that the basic acquisition, stimulation and amplitude and slope calculations have been performed it is important to double check that the slope calculations are working correctly. To check that the slope calculated by the LTP24 program are correct, input a ramp or triangle waveform from a waveform generator. See if the LTP24 program gives the correct slope calculations compared to what you know is being output (which can be independently checked with an oscilloscope). Make sure your waveform generator has very low noise. (See Section 2.9.1.)

4.7.4 Area

time fields shown by the solid horizontal Area line of the Pulse Detection Window in Fig. 4.7.4. Just as with the Peak Amplitude measurement, the A/P/N field determines whether the peak will be Automatically (A) determined to be positive or negative, forced to be Positive (P), or forced to be Negative (N).

Fig. 4.7.4. Detection of Area of the peak more negative than the pre-pulse baseline. The Area is measured in the range of 2 and 18 ms after the stimulus pulse (solid horizontal line). However, because the waveform goes positive at 16 ms, the area is only measured between 2 and 16 ms after the pulse.

Notice, that when the waveform goes to the opposite polarity of the peak, those values are not calculated in the area (for example, in Fig. 4.7.4, when the waveform goes positive after 16 ms, the area is only between the first ‘Peak’ time field and up to 16 ms after the pulse.

4.7.5 Duration

time fields shown by the dotted horizontal peak line in the Pulse Detection Window (Fig. 4.7.5). Just as with the Peak Amplitude measurement, the A/P/N field determines whether the peak will be Automatically (A) determined to be positive or negative, forced to be Positive (P), or forced to be Negative (N).

Duration can measure the duration of bursts (multiple spikes or peaks), and is therefore particularly useful in epilepsy studies for measuring the duration of epileptiform bursts and electrographic seizures.

4.7.6 10-90% Rise Time and 10-90% Decay Time

time fields shown by the dotted horizontal peak line in the Pulse Detection Window in Fig. 4.7.6. Just as with the Peak Amplitude measurement, the A/P/N field determines whether the peak will be Automatically (A) determined to be positive or negative, forced to be Positive (P), or forced to be Negative (N).

4.7.7 Coastline

time fields and is shown to occur between the left and right brackets on the waveform (Fig. 4.7.7). (Alternatively, if Peak Amplitude is also being calculated, Peak: ___to ___ms after pulse A/P/N:_ indicates the time fields). The Coastline is measured in mV or pA. Coastline does not depend upon DC baseline.

For example, the coastline of an EPSP of 1 mV amplitude would be 2mV.  Coastline is indicative of, and sensitive to, the addition of extra population spikes in an epileptiform burst and can therefore be useful in epilepsy studies.

However, coastline will also increase with extraneous high frequency noise and you should be sure that sufficent external analog (Section 2.7) or internal digital filtering (Section 4.2.4.3) is applied to the waveform when the Coastline measurement is made.

4.7.8 PopSpike Amplitude and PopSpike Latency

The PopSpike Amplitude is calculated as the amplitude from the popspike peak to the intersection with an interpolated tangent dotted line drawn between the pre-popspike peak to the post-popspike peak (shown be the solid vertical line in Fig. 4.7.8). PopSpike Latency is calculated as the time between the occurrence of stimulus pulse and the popspike peak. PopSpike Amplitude and PopSpike Latency do not depend upon DC baseline or Peak Amplitude.

Next, you must set the time range the popspike will be detected over by the "__to__ms after pulse" fields above. (If Peak Amplitude it to be calculated then use the time fields in the "Peak:__to__ms after pulse A/P/N:_" .)  This time range is shown between the left and right bracket located on the waveform in Fig. 4.7.8.

4.7.9 Average Amplitude

Fig. 4.7.9. Detection of Average Amplitude between 23 and 27 ms after the stimulus pulse (solid line) relative to the pre-pulse baseline (dotted line).

4.7.10 Cell Resistance (Rm)

Cell resistance (Rm) detection occurs automatically. The Rm measurement is the difference between the averaged PreRmBaseline and the averaged RmPulse (see Fig. 4.7.11.2). That part of the Rm measurement taken during the RmPulse is an average taken between 50% and 90% of the RmPulse. That part of the Rm measurement taken during the PreRmBaseline is taken between 10% and 90% of that period, or a maximum duration of the time equal to 50-90% of the RmPulse.

Note that the Cell Resistance, Rm, is actually the amplitude of the voltage or current deflection during the current injection pulse when stimulating intracellularly. It is measured in mV or pA and is therefore only an indirect measurement of Cell Resistance that can be obtained by 

        Rm = V_Pulse / I_SteadyState - Rs

Correction to the calculation of input resistance measurement Rm in October, 2006.  In the previous versions the on-line manual of the LTP Program, the cell input resistance Rm was incorrectly said to be calcuclated using  

            Rm = V_Pulse / I_SteadyState 

and is now correctly calculated using  

            Rm = V_Pulse / I_SteadyState - Rs           

 where V_pulse is the amplitude of the RsRm voltage clamp test pulse relative to baseline of the preceding epoch, I_SteadyState is the amplitude of the current measured between the baseline and 50% and 90% of the pulse when the current has reached steady state, and Rs is the patch electrode series resistance (see Section 4.7.11).  

Because Rs is usually much less than Rm,  Rm ~ V_Pulse / I_SteadyState  still is roughly true, but since Rs values are typically 5% to 10% of Rm values, previous calculation of Rm values (before October, 2006) were roughly 5% to 10% too high.  Since Rm = V_Pulse / I_SteadySstate - Rs  is the theoretically correct function, it should be used in the Rm calculation when Rs is measured.  Note, however, that since Rs is usually slightly overestimated, Rm now will be slightly too low!

In the LTP Program, Rs is almost always measured during patch clamp voltage clamping (assuming you click the AnalysesToDo Rs check box), but is not measured during patch clamp current clamping, and therefore Rs = 0 in this case, and Rm = V_Pulse / I_SteadyState.  Furthermore, during whole cell single electrode voltage clamping, where series resistance is theoretically zero, you would not measure Rs and it would therefore be set to Rs = 0, and Rm = V_Pulse / I_SteadyState.  For intracellular current clamping using a bridge circuit, Rs would also not be measured and  would therefore be set to Rs = 0, and Rm = V_Pulse / I_SteadyState.

4.7.11 Patch Electrode Series Resistance (Rs)

The patch electrode series resistance (Rs) can either be taken as the 1) peak of the capacitative transient, 2) the extrapolated peak fitted by a single exponential, or 3) the extrapolated peak fitted by a double exponential.

Unfortunately, the exponential fitting is still a work in progress. The code appears to be fine, but in practice we have found that this double exponential fit (which is a better fit for a hippocampal cell than a single exponential) can be variable due to the relatively slow AD sampling interval of 100 usec in LTP24, and therefore may not be dependable. Whether the fit is dependable depends to a certain extent on what is recorded. You have to test out the dependability for yourself. Future versions of the LTP program will have to have substantially higher sampling frequency of 10-20 usec in order for Rs exponential fitting to work well.

Instead, we find that the most accurate determination of Rs is to measure an unfiltered Rs peak on the on-line oscilloscope at various times during the experiment, and use the Rs peak measurement made by the LTP program to detect if there is a change in true Rs. This assumes that you have cancelled out the pipette capacitance.

Note that for all three ways to measure the patch electrode series resistance (Rs) is actually the peak or fitted amplitude of the capacitative transient and is measured in pA. Rs must be manually converted to resistance by Rs = V_Pulse / I_Peak.

Fig. 4.7.11.1. Series Resistance Calculation Method Dialog Box.

If the Peak Rs detection method is chosen, no fields need to be entered. For this value to be even proportionally related to the true Rs value, make sure you have canceled out the pipette capacitance.

For a discussion on the correct measurement of Series Resistance using exponential curve fitting see Ogden and Stanfield (Patch clamp techniques for single channel and whole-cell recording, In: Microelectrode Techniques, The Plymouth Workshop Handbook, Second Edition, Ed. D. Ogden, The Company of Biologists Ltd., Cambridge, 1994).

4.8  Analyzing All S0- and S1-Evoked Postsynaptic Responses in Both AD Channels in a Sweep

Beginning with version 2.22A, the LTP Program has been capable of acquiring data from 2 channels and analysing all S0- and S1 evoked synaptic responses on both acquisition channels.  

Fig. 4.8.1 illustrates this with one acquisition channel extracellularly recording the CA1 dendritic layer of the hippocampal slice, and the other acquisition channel extracellularly recording the CA1 cell body layer (panels AD0 and AD1 respectively in Fig. 4.8.1b).  The stimulation during the sweep consists of paired-pulse S0 stimulation (yellow trace) followed by paired-pulse S1 stimulation (magenta trace) (Fig. 4.8.1a).  Both S0 and S1 stimulations evoke fEPSPs in the CA1 dendritic layer (AD0) and fEPSPs with overriding population spikes in the CA1 cell body layer (AD1).  All S0-evoked synaptic responses are analysed and shown by yellow lines in AD0 and AD1, and all S1-evoked synaptic responses are analysed and shown by magenta lines.  The values of the peak amplitude calculations for AD0 are shown in PkAmp0 (yellow crosses for S0 evoked responses, and magenta squares for S1-evoked responses).  The values of the population spike amplitude calculations for AD1 are shown in PSamp1 (yellow crosses for S0 evoked responses, and magenta squares for S1-evoked responses).

a    b    c "Amp_Filename" "0611R000.AMP" "Analysis_Reanalysis" "#"  "Filename" 0 "06110214.P0"    1 "06110214.P0" 2 "06110214.P0" 3 "06110214.P0" 4 "06110214.P0" 5 "06110214.P0" 6 "06110214.P0" 7 "06110214.P0" "Offline_Reanalysis"  "TimeOfDay"   "15:49:50.1" "15:49:50.1" "15:49:50.1" "15:49:50.1" "15:49:50.1" "15:49:50.1" "15:49:50.1" "15:49:50.1" ""    "" "Time_sec" 0.0100 0.0600 0.1400 0.1900 0.0100 0.0600 0.1400 0.1900 "" "AD" 0 0 0 0 1 1 1 1 "" "Sx"  "S0" "S0" "S1" "S1" "S0" "S0" "S1" "S1" "" "Pul#" 0 1 0 1 0 1 0 1 "PkAmp" "mV_mV"  -1.758 -2.578 -1.958 -2.617 0.000 0.000 0.000 0.000 "PSamp" "mV_mV" 0.000 0.000 0.000 0.000 -0.857 -3.657 -1.032 -8.218

Fig. 4.8.1.  Two channel acquisition and analysis all S0- and S1-evoked synaptic responses in both acquisition channels.   a)  S0 paired-pulse stimulation (yellow top trace) and S1 paired-pulse stimulation (magenta bottom trace).   b) Top two panels show calculated values for the peak amplitude of the synaptic responses in AD0 (PkAmp0), and the pop-spike amplitude from the responses in AD1 (PSamp1).  The bottom two panels show channel AD0 (a recording of fEPSPs from the CA1 dendritic layer), and channel AD1 (a recording of population spikes from the CA1 cell body layer). The peak and pop-spike amplitude calculations of paired S0-evoked synaptic responses are shown in yellow, and of paired S1-evoked synaptic responses in magenta.   c) Program output.  A simplified Amplitude events list file (‘AMP’ file) from the analysis of the sweep in (b). The ‘Time sec’ field shows the time of the stimulus pulse from when analysis starts (in this case simply the beginning of this sweep). Channel AD0 is calculated first, S0-evoked EPSPs are calculated first, and pulse number is calculated sequentially. When the PkAmp values (for AD1) or the PSamp values (for AD0) are not calculated, they are set to 0.

When the LTP Program performs an on- or off-line analysis of one or more acquisition sweeps, it outputs an Amplitude events list file (the ‘AMP’ file) such as the simplified one shown in Fig. 4.8.1c for the analysis of the sweep in Fig. 4.8.1b.  “TimeOfDay” shows the time the sweep began.  “Time sec” shows the time of the stimulus pulse from when analysis starts (in this case simply the beginning of this sweep since this was the only sweep analyzed).  “AD” shows the AD channel from which the synaptic response was obtained (channel AD0 is calculated first).  “Sx” shows whether S0 or S1 stimulation was used to evoke the synaptic response (S0-evoked responses are calculated first).  “Pul#” shows the number of the S0 or S1 pulse that evokes the synaptic response (pulse number is calculated sequentially).  “PkAmp” shows peak amplitude values of the S0- and S1-evoked fEPSPs in channel AD0.  “PSamp” shows the population spike amplitude values of the S0- and S1-evoked responses in channel AD1.  When the PkAmp or PSamp values are not calculated (e.g. PkAmp values for AD1 and PSamp values for AD0), they are set to 0.  The AMP file can be loaded into a spreadsheet program and the columns can be sorted as desired.

4.9  Special Analyses of Trains

Sometimes the experimenter is interested in examining each postsynaptic response evoked by a stimulus pulse in a train, in which case the baseline and synaptic response of each pulse is analyzed (as with the ‘two pulse trains’ in Fig. 4.8.1).  Alternatively, the synaptic responses evoked by train stimulation can be treated in a special manner by the LTP Program as described below.

First, use the AmpFile menu (Fig. 3.4.3) to call up the Analysis of Pulses in Trains dialog box (Fig. 4.9).  For extracellular electrode stimulation S0 and S1, for Sweeps T0, T1, P0 and P1, it allows four choices to be made: 1) Analyze every pulse in train (e.g. as in Fig. 4.8.1), Analyze every pulse but use the baseline of the 1st pulse as the baseline of all the pulses (Section 4.9.1), use only the first pulse in the train, thereby analyzing the whole train (Section 4.9.2), and finally, use the baseline of the first pulse in the train with the response of the last pulse in the train (Section 4.9.3).

Fig. 4.9.  Analysis of Pulses in Trains dialog box.

. 

4.9.1  Analyze every pulse in train, but use the baseline of the first pulse as the baseline for each pulse

First, the synaptic responses to each train pulse may be analyzed relative to the baseline of the first pulse in the train.  Fig. 4.9.1 shows the peak amplitude measurement of each fEPSP in a 100 pulse train relative to the prestimulus baseline of the first pulse.  In this example, stimulus artifacts have been automatically removed (Fig 4.9.1c) because this allows peak amplitude measurement that is not contaminated by stimulus artifacts occurring near the fEPSP peak.

a    b    c

Fig. 4.9.1. Analysis of the peak amplitude of each mossy fiber-CA3 fEPSP in a 100 pulse train relative to the prestimulus baseline of the first pulse.  Stimulus artifacts have been automatically removed.   a)  Top panel: Calculated peak amplitude values (PkAmp0). Bottom panel: Recording of the whole train showing superimposed peak amplitude calculations.   b) Sweep time  magnification showing the first 9 fEPSPs of the train. The light gray traces show the waveform with stimulus artifacts present, the dark gray traces show the waveform with the stimulus artifacts removed.   c) Amplitude and sweep time trace magnification showing automatic (1 ms here) blanking of the stimulus artifact (light gray trace) riding on the top of the second fEPSP.

4.9.2  Analyze whole train by analyzing only first pulse in train but detecting whole train

Second, trains can also be analyzed as a single entity.  If the baseline and response of only the first train pulse is used, all stimulus artifacts are blocked, and the time of measurement is set sufficiently long to encompass the whole train, then the synaptic response of the entire train will be measured (Fig. 4.9.2).  With this analysis, the peak amplitude of the largest EPSP in the train and the area of the  synaptic response of the entire train can be obtained.  Removal of stimulus artifacts is necessary to permit accurate calculation of area and peak amplitude.

Fig. 4.9.2.  Analysis of trains as a single entity by using the baseline and response of only the first pulse in the train, but with the time of measurement set to a sufficiently long duration after the first stimulus pulse (1.5 to 50 ms here) to encompass the whole train.  This measures the peak amplitude of the largest CA1 fEPSP in the train and the area of the entire synaptic response of the train.  Accurate calculation of area requires removal of stimulus artifacts.  Top two panels: Calculated peak amplitude (PkAmp0) and area (Area0) values for each train.  Bottom panel: Recording of the entire 4 train sweep.

4.9.3  Analyze train using baseline of the first pulse and response of the last pulse

Third, trains can also be analyzed by using the baseline of the first pulse and response of the last pulse (Fig. 4.9.3).  This measures the response at the end of the train regardless of the length of the train. In addition to measuring the last synaptic response of a train, this is useful for measuring responses that only occur after the train has ended, such as post-inhibitory rebound.

Fig. 4.9.3.  Analysis of a train using baseline of the first pulse and response of the last pulse.

4.10  AutoReset Timebase (AutoZoom)

In the default condition when AutoReset ADsweep timebases is set to On, the beginning ADsweep time returns to 0 every time a new ADsweep is acquired or loaded, and the ADsweep duration is sent to the full time available every tune a new ADsweep us acquired or loaded.

However if AutoReset ADsweep timebases is set to Off, the beginning ADsweep time and the ADsweep duration remain at the zoomed in values every time a new ADsweep is acquired or loaded. Fig. 4.7.9 shows the loading of an ADsweep with AutoReset ADsweep timebases set to Off. In this case the start of the ADsweep graph begins 140 msec after the acquisition has begun, and the ADsweep graph duration is only 100 msec of the total 400 msec acquisition time (sweep in Fig. 4.7.11.2).  This is a useful way of closely looking at an EPSC when also using the Rm/Rs pulse.

4.11  Setting Whether to Save Files

Note that the number of AD sweep files that can be handled is 10000 in LTP24. We have found (at least on Pentium II computers with Windows 98 / DOS 7.0) that with more than 1000 ADsweep files per directory, saving and reloading slows down.

4.12  Saving the Protocol (*.pro) File to Disk

To open the Protocol File SaveAs Dialog Box (Fig. 4.12.1). If the protocol file is to be saved under a different name, and then enter the filename - any filename ending in pro, such as "sigavg.pro".

Fig. 4.12.1. Protocol File SaveAs Dialog Box. Note that the top line shows the protocol file loaded.

Updating the protocol file is very useful in resuming the experiment where you left off if a crash or power outage has occurred (see Section 4.17).

4.13  Run the Experiment

There can be some ADsweep-to-sweep jitter during rapid LTD stimulation (see APPENDIX B.6). Measure this sweep-to-sweep jitter on your computer, and if the jitter is too large, do your LTD stimulation with an auxiliary stimulator.

4.14 Printing AMP Graphs, ADsweep Graphs and AMP Text Files After an Experiment

Printing out AMP graphs, ADsweep graphs, and AMP text files is usually done at the end of an experiment. However, it can also be done on-line during the experiment if a printer is directly connected to your data acquisition computer (or the printer is connected to a networked computer, see Section 7.1).

and then entering a number in the field between 1 and 10 (Fig. 4.14). See Fig. 4.14.2.4 for ADsweep graphs with lines of of 1,3,5 and 7 dot widths.

4.14.1 AMP Graphs

An example of printing Amp/Slope graphs to a LaserJet printer from a *.LJ0 graph file is shown in Fig. 4.14.1.

Fig. 4.14.1. (Next Page) An example of printing Amplitude/Calculation graphs to a LaserJet printer from a *.LJ0 graph file. (To print Figs. 4.14.1 and 4.14.2 in this manual I used IMSI Graphics Converter Gold to convert the *.LJ file to a BitMap (*.BMP) file.

4.14.2 ADsweep Graphs

and, if using Channel AD0 or AD1, choose from 1 to 9 one channel charts per page with various x/y aspect ratios from the LaserJet Chart Layout Dialog Box (Fig. 4.14.2.1).  If using Channels AD0 and AD1, choose from 1 to 6 two channel charts per page from the LaserJet Chart Layout Dialog Box (Fig. 4.14.2.2).

Fig. 4.14.2.1. One AD Channel LaserJet Chart Layout Dialog Box.

Fig. 4.14.2.2.  Two AD Channels LaserJet Chart Layout Dialog Box.

An example of printing one ADsweep graph per page of a patch clamp experiment to a LaserJet printer from a *.LJ graph file is shown in Fig. 4.14.2.3. An example of printing 4 ADsweep graphs of extracellularly recorded field EPSPs to a LaserJet printer from a *.LJ graph file is shown in Fig. 4.14.2.4. The linewidth of each sequential graph as been changed from 1 to 3 to 5 to 7 dots (300 dots/inch).

Fig. 4.14.2.3. Printing one ADsweep graph per page of a patch clamp recorded current injection pulse (left) and an EPSC (middle) to a LaserJet printer from a *.LJ graph file.

Fig. 4.14.2.4. Printing four ADsweep graphs per page of extracellularly recorded field EPSPs to a LaserJet printer from a *.LJ graph file. The linewidth of each sequential graph as been changed from 1 to 3 (default) to 5 to 7 dots (300 dots/inch).

4.14.3 AMP ASCII Text File Structure

However, usually the AMP text files are directly loaded into a spreadsheet program such as Excel, or plotting program such as Sigma Plot.

An example of the structure of an AMP text file is shown in Fig. 4.14.3.

"Amp_Filename" "0611R000.AMP" "Analysis_or_Reanalysis" "Offline_Reanalysis" "" "" "" "#"  "Filename" "TimeOfDay" "Time_min"  "Time_sec" 0 "06110214.P0"  "15:49:50.1" 0.000167 0.0100 1 "06110214.P0" "15:49:50.1" 0.001000 0.0600 2 "06110214.P0" "15:49:50.1" 0.002333 0.1400 3 "06110214.P0" "15:49:50.1" 0.003167 0.1900 4 "06110214.P0" "15:49:50.1" 0.000167 0.0100 5 "06110214.P0" "15:49:50.1" 0.001000 0.0600 6 "06110214.P0" "15:49:50.1" 0.002333 0.1400 7 "06110214.P0" "15:49:50.1" 0.003167 0.1900 8 "06110215.P0" "15:50:20.1" 0.500167 30.0100 9 "06110215.P0" "15:50:20.1" 0.501000 30.0600 10 "06110215.P0" "15:50:20.1" 0.502333 30.1400 11 "06110215.P0" "15:50:20.1" 0.503167 30.1900 12 "06110215.P0" "15:50:20.1" 0.500167 30.0100 13 "06110215.P0" "15:50:20.1" 0.501000 30.0600 14 "06110215.P0" "15:50:20.1" 0.502333 30.1400 15 "06110215.P0" "15:50:20.1" 0.503167 30.1900 "" "" "" "DC" "PkAmp" "PkLat" "Area"   "Dur" "AD"  "Sx" "Pul#" "mV_mV" "mV_mV" "ms" "mV_mV*ms" "ms" 0 "S0" 0 0.000 -1.758 0.0 0.000 0.000 0 "S0" 1 0.000 -2.578 0.0 0.000 0.000 0 "S1" 0 0.000 -1.958 0.0 0.000 0.000 0 "S1" 1 0.000 -2.617 0.0 0.000 0.000 1 "S0" 0 0.000 0.000 0.0 0.000 0.000 1 "S0" 1 0.000 0.000 0.0 0.000 0.000 1 "S1" 0 0.000 0.000 0.0 0.000 0.000 1 "S1" 1 0.000 0.000 0.0 0.000 0.000 0 "S0" 0 0.000 -1.632 0.0 0.000 0.000 0 "S0" 1 0.000 -2.658 0.0 0.000 0.000 0 "S1" 0 0.000 -1.832 0.0 0.000 0.000 0 "S1" 1 0.000 -2.717 0.0 0.000 0.000 1 "S0" 0 0.000 0.000 0.0 0.000 0.000 1 "S0" 1 0.000 0.000 0.0 0.000 0.000 1 "S1" 0 0.000 0.000 0.0 0.000 0.000 1 "S1" 1 0.000 0.000 0.0 0.000 0.000

"RisTm" "DecTm" "CoastLn" "PSamp" "PSlat" "Slope" "AvgAmp"   "Rs"   "Rm" "ms" "ms" "mV_mV" "mV_mV" "ms" "mV_mV/ms" "mV_mV" "mV_mV" "mV_mV" 0.000 0.000 0.000 0.000 0.0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 -0.857 0.0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 -3.657 0.0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 -1.032 0.0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 -8.218 0.0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 -0.915 0.0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 -3.723 0.0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 -1.158 0.0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 -8.339 0.0 0.000 0.000 0.000 0.000

Fig. 4.14.3.  The first two sweeps (taken 30 sec apart) of an Amplitude/Calculation ASCII text file.  This is the full 22 columns wide AMP file for the simplified AMP file shown in Fig. 4.8.1c, but it had to be separated into two sections to fit on this page. Only PkAmp and PSamp were calculated, the other values were not calculated and were set to 0.

4.14.4 LTP222A/LTP230D/LTP24  ADsweep ASCII Text File Structure

Although you normally do not want to directly examine the ADsweep ASCII text files, the header and the first data point of an ADsweep file for LTP version 2.30D is shown in Fig. 4.14.4.1. This will be helpful if you want to write a custom program to do different analyses of the ADsweep files than is available in LTP24.

Fig. 4.14.4.1.  The new header for all ADsweeps obtained with LTP222A, LTP230D and Ltp24.  This head contains all stimulation information (including S0- and S1-evoked, Rm/Rs, and epoch analog and digital stimulation).  The numbers on the left are just header line numbers and are not included in the actual header.

4.14.5   LTP114J ADsweep ASCII Text File Structure

The older header for all ADsweeps obtained with LTP114J and earlier LTP programs is shown in Fig. 4.14.4.2.  LTP24 will still analyze these LTP114J ADsweep data files.  Generally there is no reason to currently program for this header, but some program such as Tim Benke's Non-Stationary Fluctuation Analysis already have.

Fig. 4.14.4.2.  The old header for all ADsweeps obtained with LTP114J and earlier LTP programs.  LTP24 will still analyze these LTP114J ADsweep data files, but not all stimulation information is in them.  Only the time of the first S0 pulse in Pulse Sweeps P0 and AP0, the time of the first S1 pulse in Pulse Sweeps P1 and AP1, and the time of the RmRs pulse is contained in this header.   The numbers on the left are just header line numbers and are not included in the actual header.

4.15  Zip Data Files at the End of an Experiment?

4.15.1  There are no longer strong reasons to zip data files

In earlier versions of this manual (e.g. version 1.14J, Sept, 1999), I suggested zipping up your data files at the end of an experiment (at least to put on 1.44MB floppies and 100MB Iomega 'Zip' disks, although not necessarily to Write-Once/Read-Many CD-R's).  This was primarily because the many ASCII ADsweep files saved in an experiment was an inefficient use of disk space (at least a 10 fold decrease in saving efficiency).

However, the data saving situation has changed substantially since Sept., 1999.  First, hard disk size has increased by over 5 times - online data storage with 10-30 GB hard disks means that storage is almost unlimited.  Second, the cost of permenant CD-R storage is now almost negligable (approximately $1 for a 650MB disk).  Third, the LTP24 program will now reanalyze data directly from the read-only CD-R's (while writing the new *.AMP and *.LJ files to the hard disk).  If these CD.R's contained zipped files, they would first have to be copied to the hard disk, unzipped, and then reanalyzed.  Fourth, LTP24 automatically creates a new data folder at the beginning of each experiment, and it is easy to create additional directories 'on-the-fly' during the experiment.  This makes data organization much easier.

Therefore, I suggest you save your data on your 10+ GB hard disk, and one other place (such as Iomega Zip disks, or the hard disk in another computer), and then write a couple of CD-R's when you have ca 600 MBs of data.  The key is to always have your data saved in at least two places.

4.15.2 Zipping Files with PKZIP

However, you may still want to zip your files using a file compression program such as ZipMagic (www.zipmagic.com), WinZip (www.winzip.com), or PKZIP (www.pkware.com).

See the PKZIP manual for more details or type PKZIP at the command line to get an abbreviated explanation of what else to do. PKZIP is one of the standard file compression program for Microsoft/Intel PCs. I have used PKZIP for years without a single problem.

4.16 Working Around Program Bugs

Programmers have a saying that the only program without any bugs is the one that's no longer being used. I've tried to stamp out all the bugs, but I know there are several that are still there. Unfortunately, you'll find more.

All the bug I know about can be 'worked around' (we do), or else the LTP program simply cannot be used with that particlular computer / AD board combination (see APPENDIX A.1 and APPENDIX A.2).

See also Section 4.17.

4.17 Recovering from a Crash

This program is designed to recover 'gracefully' from a crash or a power failure. (Whether your preparation recovers from a power failure is another story). When data is written to a file, the file is opened, immediately written to, and then immediately closed. Files are not left open until the program is exited. Therefore, files can (hopefully) only be corrupted with a power outage or crash if that occurs in that short time while they are written to. And then, only the latest file (such as the last ADsweep file) could be corrupted. All the other ADsweep files will be OK, and all the Amplitude and LaserJet files can be reconstructed from the remaining ADsweep files. This is in contrast to several data acquisition programs (which shall remain nameless) which opened data files at the beginning of the experiment and only closed then at the end of the experiment. If a crash or power failure occurred during the experiment, all data was lost!

If the program crashes and you are in the middle of an experiment, at least write down verbatim any error messages (include all correct punctuation and upper/lower cases) and any other possibly useful information that might indicate when the program crashed).

After a crash or power failure, restart the program in the start-up directory as normal.  The LTP24 program will automatically load the last protocol file saved. Therefore, it is very useful to save the most recent field and dialog box values every time you change them by resaving the present protocol file (by using File -> Save as described in Section 4.12). Remember, these will be the values in the program when the program is restarted after a crash.

When LTP24 restarts it checks the last ADsweep files (*.T0, *.T1, *.P0, *.P1, *.AP0, *.AP1), Amplitude files (*.AMP) and LaserJet files (*.LJ?) saved. When saving the next ADsweep, Amplitude or LaserJet file, it will check to see if it already exists, and if so it will not write over this file, but will instead increment the file number and try writing again until a file number is found where the file doesn't exist, and then it will be saved.

Then try rerunning the program. See also Section 4.16, Working Around Program Bugs.

Go to LTP Manual - Chapter 5

Return to the LTP Manual (Table of Contents)

Return to the LTP Program Home Page