Neural signal acquisition and analysis system under Windows95/98 environment

    Abstract: This paper introduces some characteristics of neural discharge signals, and introduces a real-time neural signal acquisition and analysis hardware and software system designed for the characteristics of Windows95/98 operating system. This system provides a basic method for analyzing neural signals. method. Keywords: neural signal, neural signal screening, interrupt, virtual device driver

   

     In neurobiological research, it is often necessary to insert microelectrodes near nerve cells in the brain to record the discharge (i.e., action potential) activity of individual nerve cells to understand the normal functions and abnormal changes of nerve cells. The discharge of nerve cells is a spike-like pulse digital signal with an amplitude of about 100~150mV, a duration of about 1~2ms, and a repetition frequency of up to several hundred Hertz. Nerve cells have irregular basic discharge activity at rest, and will respond after being stimulated, which is manifested by an increase or decrease in their discharge frequency on the basic discharge background, and these increased or decreased parts represent effective neural information. . Since the response of nerve cells to stimulation will appear at a fixed time period after stimulation, while the basic discharge appears randomly, the currently commonly used method of processing neural information is to stimulate the nerve cells multiple times and simultaneously The response results of a pair of each stimulation are superimposed to form the so-called "time histogram after stimulation" and "spike interval histogram", thus revealing the statistical laws of this effective information. According to the needs of research work, we developed a neural signal acquisition and analysis system based on Windows95/98 system. This system can meet the needs of neurobiological research and has certain prospects for promotion.

1 System composition

    The entire system consists of two parts: hardware system and software system. The hardware system mainly includes microelectrode amplifier, window discriminator, data acquisition card and microcomputer. As shown in Figure 1, a glass microelectrode is inserted into the animal's brain to pick up the discharge signal of nerve cells, and the signal is amplified by a microelectrode amplifier. The amplified nerve cell discharge signal is sent to the window discriminator for signal selection and waveform conversion, and then the TTL pulse signal obtained by the conversion, which corresponds to the nerve cell discharge signal, is sent to the data acquisition card. The A/D converter collects this signal into the computer. The computer's task is to analyze and process the discharge signals of nerve cells, store experimental data, and display and output the processing results.

    The software system is a self-programmed signal acquisition and analysis system based on the Windows 95/98 environment. It can produce post-stimulation time histograms and spike time interval histograms of nerve cell discharge signals. In order to make the system work reliably under Windows95/98, we wrote a virtual device driver for the data acquisition card. By operating the hardware through VxD, the program can not only read and write the hardware device stably and reliably, but also meet the requirements of real-time signal processing.

1.1 Design of each hardware in the system

1.1.1 Window discriminator

    Since nerve cells are in volume conductors in the brain, the discharge signals of more than one nerve cell are often picked up by microelectrodes. The function of the window discriminator is to enable the experimenter to set the level window according to the amplitude of the nerve discharge signal, selectively obtain the discharge signal of a certain cell from the discharge of several nerve cells with different amplitudes, and convert the selected signal into TTL pulses enable computers to accurately identify them. The window discriminator is essentially a Schmitt trigger, but an additional circuit is added to meet the needs of neurobiological experiments, such as a window line enhancement generation and adjustment circuit, a stimulation artifact suppression circuit, and a TTL signal output audio circuit. circuit etc. Figure 2 is a schematic diagram of the function of the window discriminator. Any nerve cell discharge signal that enters the screening window can trigger the Schmitt circuit, thereby obtaining a TTL pulse corresponding to the time of the nerve discharge signal at the output end of the instrument. The experimenter can adjust the height of the upper and lower window discrimination lines as needed to select the nerve discharge signals to be collected and analyzed, and discard other unnecessary discharge signals.

1.1.2 Data collection and quantitative

    data collection are completed through a data acquisition card (AC1810 type, Beijing Shuangnuo Technology Co., Ltd.). The data acquisition card has an A/D converter with a maximum sampling rate of 100kHz and a quantization accuracy of 12bits. The card provides 8 channels of double-ended input, each channel has a sample and hold, and can collect 8 signals in parallel. At the same time, the card has a FIFO cache of 1K words. When the FIFO cache is half full, an interrupt signal will be issued.

    In addition to collecting digital electrical signals such as nerve cell discharge, this system also collects other analog electrical signals (such as EEG, ECG, EMG, blood pressure changes, etc.). For convenience, these two input signals with different properties are input into the computer through A/D sampling in the signal acquisition module. In addition, since the spike interval histogram analysis of nerve cell discharge signals requires a time interval resolution of 0.1ms, we selected a sampling rate of 10kHz/channel.

1.1.3 Microcomputer

    Since ordinary PC compatible computers are cheap and support many softwares, we chose the 80x86 series microcomputer to process data and display the processing results.

1.2 Software Programming

    Due to the increasing popularity of Windows95/98 system, we chose Windows95/98 system as the system platform of the software. The software system consists of three parts: VxD, dynamic link library and application program. In the Windows95/98 operating system, applications run at the 3rd privilege level. If the application executes restricted instructions, especially when executing instructions such as task switching and interrupt processing, it will cause the processor to generate an interrupt. As a result, the Windows system will give an error message and may terminate the program. Therefore, operations on the hardware must be performed through the VxD running on the 0th privilege level, and the data analysis and processing parts that have nothing to do with the hardware can be completed by the application program. Windows95/98 is a time-sharing multi-tasking system. In order to retrieve data from the FIFO cache of the A/D converter in a timely manner and avoid losing data due to the thread being in a waiting state, the signal acquisition module adopts the interrupt sampling method to collect data. According to structure, the software system can be divided into two parts: signal acquisition module and data processing module. The signal acquisition module consists of two parts: VxD and data processing module. The signal acquisition module is composed of VxD and DLL, and the data processing module is placed in the application program. The signal acquisition module transmits information to the data processing module through the message mechanism, and the data processing module completes data communication with the signal acquisition module through the Windows API function DeviceIoControl().

1.2.1 Signal acquisition module

    The flow chart of the signal acquisition module is shown in Figure 3. VxD is responsible for performing actual I/O operations and completing interrupt sampling. DLL encapsulates the interface for data exchange between applications and VxD. DLL can make applications and VxD relatively independent, which facilitates future software and hardware upgrades and maintenance. During the initialization process, VxD hooks the interrupt and applies for a buffer area from the system to store the data collected by the interrupt. During the working process, when the operating system responds to the interrupt request issued by the data acquisition card, it calls the interrupt service program in the VxD. The interrupt service program reads the data in the FIFO of the data acquisition card into the VxD buffer area, sends a message to the application program, and then exits the interrupt service program. After receiving the message from VxD, the application reads the data in the VxD buffer area into the temporary data area of ​​the application through the message response function, and then distributes the data collected by the A/D converter to each channel according to the channel source. In the data area of ​​the document class instance corresponding to the channel.

1.2.2 Data processing module

    As mentioned before, in order to reveal the statistical rules of nerve cells' response to stimulation, the collected electrical signals of nerve cells must be processed into post-stimulation time histograms and spike interval histograms. The flow of the processing process is shown in Figure 4. First, According to the TTL pulse signal input by the signal acquisition module, the time corresponding to each nerve cell discharge is calculated. For poststimulus time histograms, these temporal data were converted to time interval data of the time between the moment of activation and the synchronization onset time point. Based on these time interval data, the number of discharges per unit time on the histogram is added up to obtain a function diagram of the change of nerve cell discharge frequency with time, that is, a post-stimulation time histogram based on the results of a single scan experiment is made, thus revealing Statistical laws of nerve cell responses to stimuli. On this basis, we can also further create a post-stimulation time histogram based on the results of multiple scanning experiments, that is, the single-scan post-stimulation time histogram obtained from each scan is superimposed point by point starting from the synchronization starting point, and drawn into a cumulative post-stimulation time histogram to better reveal the statistical rules of nerve cells' response to stimulation. For the spike interval histogram, the time interval between two adjacent discharges is calculated based on the time corresponding to each nerve cell discharge, and then a spike interval event distribution histogram is made according to the time interval of the spike. Revealing changes in neural information encoding. Like the post-stimulus time histogram, this histogram can also be made into an accumulated spike interval histogram of multiple scans.

    The application is written according to the document-view structure. Each channel corresponds to a document class instance, and the data of each channel is stored in the document class instance corresponding to the channel. The data processing methods described above are also placed in the document class. Each document uses multiple views associated with the document to display the results of different methods of data processing on the same data. By calling the method of the document class, the post-stimulus time histogram data and spike interval histogram data are calculated, and then the post-stimulus time histogram and spike interval histogram are drawn in the view class associated with the document, and are drawn by the document The class saves the moment of each discharge as raw data.

2 Application examples

    Figure 5 is an example of actual use of this system. Figure A shows an application example of post-stimulus time histogram. The abscissa of the histogram is time, and the ordinate is the firing frequency of nerve cells. The experiment records the response of a cat cerebellar Purkinje cell to stimulation. It can be seen from the figure that the discharge frequency of the cell increases significantly after being stimulated, that is, it generates an excitatory response to the stimulus, and then its discharge frequency gradually returns to the basic state. Figures B and C show the pre-stimulation and post-stimulation spike interval histograms of the cell respectively, and the analysis windows of the two figures are displayed above Figure A respectively. Comparing the two pictures B and C, it is found that the time interval peak of the cell's discharge shifts to the left after being stimulated, suggesting that the information encoded by it has changed before and after stimulation.

    This system provides a powerful data analysis method for teaching and scientific research in neurobiology and related disciplines (such as physiology, pharmacology, etc.) and has broad application prospects.