Modern multi-parameter monitors generally adopt a modular structure. Each module is an independent microprocessor system that can complete a single function. Multiple modules receive commands from the main controller through the serial port and send out the measurement results. The ECG monitoring module is a key component of it. It picks up the ECG signal under the control of the single-chip microcomputer, amplifies, preprocesses, and transmits it to the main control module.

Ventricular early stroke (PVC) is the most common arrhythmia. Its real-time and correct detection is an important technique in the automatic analysis of ECG. On the basis of ECG monitoring, adding heartbeat and rhythm classification functions, you can carry out arrhythmia monitoring. Ventricular premature beat detection is generally performed in a high-end arrhythmia analysis module. At present, the functions of single-chip microcomputers are continuously enhanced. In the past, single-chip microcomputers mainly used as controllers, after completing their original tasks, there are still many CPU time and control functions left. It is entirely possible to implement some complex algorithms that used to run only on PCs. Use assembly language programming to improve the running speed.

The fuzzy system is a concrete application of fuzzy set theory and fuzzy logic. It simulates the human brain's ability to think and make decisions. It is especially suitable for uncertain systems where mathematical models are difficult to accurately represent. Many information in the medical decision-making system is uncertain, such as a large amount of feature information in the ECG signal, which is suitable for processing with fuzzy inference. The author's laboratory uses fuzzy logic to identify room-based premature Bo, with good results.

The author transplants the above algorithm in the 8051-based ECG module to realize the detection of ventricular premature beats. It is very convenient to carry out fuzzy control or inference on the fuzzy single chip. The key to realizing on digital single chip microcomputer is how to store expression membership function and fuzzy rules, and how to implement inference.

1 Membership function, fuzzy rules and fuzzy reasoning methods

Five morphological parameters, RR interval, R wave width and area, T wave area and peak, are used to describe a QRS wave. The typical morphological characteristics of ventricular premature beats are that the RR interval is small, the R wave width is large and the area is large, the T wave width is large and the peak is large. The membership function determined by statistics and experiments is shown in Figure 1.

In Figure 1, the horizontal axis represents the universe of discourse, and the vertical axis represents the degree of membership. Each input language variable defines four small (S), medium (M), large (L), and extra large (XL) language variable values. The membership function is triangular or semi-trapezoidal.

There are 32 fuzzy rules, some of which are shown in Table 1. Among them, 17 judgments are "PVC", and 15 judgments "may be PVC", indicating that the influence of this parameter is not considered.

Three methods are generally used to realize fuzzy control on the digital single chip: intensity transfer method, direct table look-up method and formula calculation method. The direct table lookup method obtains a fuzzy control table through offline calculation in advance, and stores the control table in the computer memory; during the control process, the corresponding control amount in the table is searched according to the input variables. In general control applications, this method is the fastest. However, there are many input variables (5) in this system, the dimension of fuzzy control table is high, and it is inconvenient to store and check the table. The intensity transfer method is based on the maximum-minimum method of fuzzy control. Each input parameter maps to multiple degrees of membership, and each set of inputs activates multiple rules, which may correspond to different results. Calculate the output strength of the corresponding rules of each combination with the principle of taking the smallest, and then obtain the credibility corresponding to each conclusion according to the principle of maximum membership. In this way, for each input, the output intensity corresponding to each conclusion can be obtained, which is called the output membership degree, and the membership degree with the largest output intensity is taken as the output. Experimental results show that this method is most suitable for this system.

2 The realization of the intensity transfer method on the single chip 8051

2.1 Storage of membership function

Because the digital single-chip microcomputer can only store and process digital information, and the memory capacity of the single-chip microcomputer is limited, it is simply impossible to store continuous curves of all membership functions of the input and output domain of the system. For the triangle membership function, the three-point method is used to store the three vertices of the triangle; for the half-trapezoid on both sides, the three points of the waist and top are also stored. The membership function is stored in ROM, as shown in Table 2 (only RR is listed, and the rest can be deduced by analogy).

2.2 Input blurring

The extracted feature parameters are exact values, and they are compared and combined with the membership function to find the corresponding fuzzy input. The degree of membership ranges from 0 to 1, which can be expressed as 00H to FFH on an 8-bit computer. For this system, each precise input value corresponds to at most two fuzzy input quantities greater than zero, and the remaining fuzzy input quantities are zero. For example: assuming RR = 1.1, it can be seen from Table 2 that it falls on the two intervals of medium (M) and large (L), so the membership for medium (M) and large (L) is:

μm (1.1) = (1.4-1.1) / (1.4-1.0) × FFH = BFH

μl (1.1) = (1.1-1.0) / (1.4-1.0) × FFH = 3FH

Open up an area in RAM to store each fuzzy input? As shown in Table 3.

2.3 Storage of fuzzy rules

The fuzzy rule is expressed as: IF A and B and C and D and E THEN Y (or Z)

The words immediately after "IF" are called antecedents, and the words after "THEN" are called antecedents. First, the input fuzzy values ​​S, M, L, and XL correspond to the numbers 0, 1, 2, and 3, respectively:

RR, RA, RW, TA, TH: S = 0, M = 1, L = 2, XL = 3

Each rule is represented by three bytes. The first four bits of the first byte indicate the fuzzy value of the first antecedent, the lower four bits indicate the fuzzy value of the second antecedent; the upper four bits of the second byte indicate the third antecedent, and the lower four bits indicate the fourth antecedent; The upper four bits of the third byte indicate the fifth antecedent, and the lower four bits indicate the posterior. Among them, F means not considering this predecessor, A means "is PVC", B means "may be PVC". Take the first rule (stored in ROM) as an example, as shown in Table 4.

2.4 Fuzzy reasoning and output

For a set of fuzzy inputs, traverse each fuzzy rule. Taking the first antecedent (3H) of the first rule as the address offset, plus the first address (40H) of the fuzzy input RR stored in the RAM, the membership A1 of XL can be found from the RR storage area in the RAM ; Take the second antecedent of the first rule (3H) as the address offset, and add the first address (44H) of the fuzzy input RA stored in the RAM, then the membership of XL can be found from the RAM storage area in the RAM B1; and so on, the degree of membership C1, D1, E1. According to the intensity transfer method, the minimum value of A1, B1, C1, D1, and E1 is taken as the language change of the post-rule of this rule? "It is the degree of membership of PVC" Y. When all the rules are traversed, we get: "It is PVC The membership degree of "Y1 Y2 ~ Ym", "may be PVC" is the membership degree of Z1 Z2 ~ Zn. Take the maximum value of Y1 Y2 ~ Ym as the membership degree of "Is PVC", and take the maximum value of Z1 Z2 ~ Zn as "" It may be the degree of membership of PVC. Z. If Y <7FH and Z <7FH, the output is "normal"; if Y> Z, the output is "VPC"; if Z> Y, the output may be PVC.

This article discusses the method of fuzzy inference in digital single-chip microcomputer to realize the use of fuzzy logic in ECG monitoring module to detect ventricular premature beats, so that the function of ECG monitoring module can be expanded. The detection results of ventricular premature beats and ECG waveform data are transmitted to the host computer for further analysis of arrhythmia.

Enameled Aluminium Wire

• Raw Material:Copper Rod or Aluminium
• Drawing Rolling :Convert 12.5 mm or 8 mm size of copper rod to size of clients' requirement.
• Annealing:Heat resistance
• Japanning: Painting the copper conductor.
• Finished Products :Copper or aluminum wire
• Detection:check the quality of finished products through the machine.
• Packing :PT25/60/90/200 or others
• Shipping :Container

Enameled Aluminium Wire,Transformer Enameled Aluminum Wire,Polyester Enameled Al Round Wire,Aluminum Wire Winding Magnetic

HENAN HUAYANG ELECTRICAL TECHNOLOGY GROUP CO.,LTD , https://www.huaonwire.com