Electronic particle counting apparatus

Jan. 17, 1956

W. S. PIKE

2,731,202

ELECTRONIC PARTICLE COUNTING APPARATUS

Filed April 3, 1951

<img file="US2731202A_D0001.tif" />

<img file="US2731202A_D0002.tif" />

<img file="US2731202A_D0003.tif" />

INVENTOR

Ulinthrop STike

ATTORNEY

2,731,202

Patented Jan. 17, 1956

United States Patent Office

2,731,202

ELECTRONIC PARTICLE COUNTING APPARATUS

Winthrop Seeley Pike, Princeton, N. J., assignor to Radio Corporation of America, a corporation of Delaware

Application April 3, 1951, Serial No. 219,000

Claims. (Cl. 235—92)

This invention relates to improvements in apparatus for counting uniformly shaped discrete particles spread out- in a field of view. By “particles” is meant not only physical entities having three dimensions, but any discrete areas or spots having an appearance contrasting with that of the general background. While not limited thereto, the present invention finds particular application in systems for determining blood cell count, and it will be described with special reference thereto.

One of the most frequent laboratory tests performed in hospitals is the red blood cell count, for the concentration of such cells in the blood is very greatly changed in certain forms of illness. The usual technique for performing this test is exacting, laborious and time consuming and may, in the case of large hospitals, require the services of a considerable number of skilled technicians. The procedure is as follows: A sample of the patient’s blood is accurately diluted to one part of blood in two hundred parts of diluting solution. The diluting solution used in making a red cell count may be chosen to destroy the white cells, thus leaving only red cells visible in the counting microscope. After agitation to insure thorough mixing, a minute quantity of the diluted blood is placed in a ruled “counting chamber,” Of known volume, under a microscope. The rulings define the volume within which cells are to be counted. The technician then observes the blood within the standard volume of the chamber by means of the microscope and manually counts the cells within that volume, multiplying the result by the appropriate factor to arrive at the number of cells per cubic millimeter in the original sample. In normal blood this figure is about five million and the dilution is such that the technician must count several hundred cells. It will be readily appreciated that the process is tedious, fatiguing and has many opportunities for human error. In practice, a skilled technician can make only about six counts per hour with reasonable accuracy.

Machine counting offers an appreciable saving in time, and therefore money, as well as improved accuracy. Tests with an apparatus of the type disclosed herein indicate that five per cent accuracy may be expected. An accuracy of ten per cent is considered typical for red blood cell counts performed in the conventional manner.

It is a general object of the present invention to provide an improved apparatus for automatically counting uniformly shaped particles, and one that is particularly applicable to the problem of blood cell counting.

A more specific object of the invention is to provide a particle counting apparatus wherein errors due to differences in particle size are minimized.

In accordance with the invention, the foregoing and other related objects and advantages are attained in a system wherein a field of view in which particles appear is scanned with a beam of energy. After the beam has impinged on the viewing field, changes in beam intensity (due to particle interceptions of the beam) are converted io electrical pulses. These pulses are counted. The average pulse width also is determined as a measure of average particle size. This pulse width information then is supplied to the pulse counting section of the system as a correction factor to minimize errors in particle 5 count resulting from variations in particle size.

A more complete understanding of the invention can be had by reference to the following description of illustrative embodiments thereof, when considered in connection with the accompanying drawing, wherein:

Figure 1 is a schematic diagram illustrating generally a particle counting system arranged in accordance with the invention,

Figure 2 shows an alternative form of particle-represehtative-pulse generator suitable for use in a system of 15 the type shown in Figures 1 and 3, and

Figure 3 is a schematic diagram showing circuit details of an apparatus arranged in accordance with the invention.

In the system shown in Fig. 1, the field of view con20 taining particles to be counted'is assumed· to comprise an unruled counting chamber 10 containing a blood sample, prepared: _as previously explained and arranged on the stage 11 of a microscope 12.

A television camera 14 is mounted with its viewing 25 lens 16 facing against the eyepiece 18 of the microscope 12. As usual, the electron beam in the camera will “raster scan” the- field of view presented thereto. The term “raster scanning,” as used herein· and in the appended claims, is intended to mean repeated movement 30 of a beam of energy along a multiline pattern, with each line in the pattern being parallel to and slightly displaced. from the preceding line. Such a scanning pattern is well known in the television art and allied fields.

As the electron beam in the camera 14 scans the field 35 of view presented through the microscope optical system, the usual “video” television signal will be generated. This video signal preferably is supplied to a monitoring television receiver 20, to provide on the viewing screen 22 thereof a picture of the field of view being examined; 40 This, of course, facilitates optimum adjustment of the microscope and camera controls. It has been found that in a television pulse generating system of the type being described, optimum results are obtained with so-called “dark field” illumination.

As is well known, the signals generated in the television camera 14 will include pulses representative of the particles appearing in the field of view of the camera. These pulses are applied to separate channels in the system of Fig. 1. In the first channel, the pulses are 50 counted in a pulse counting circuit 24 to provide on a meter 26 an indication of the number of pulses counted. For example, the metering circuit may comprise a socalled integrating count rate meter, a specific example of which is given hereinafter.

if the size of the particles being counted is constant, then the pulse count indicated by the meter 26 will be quite accurately proportional to the number of particles in the field of view. In practice, however, this is usually not the case. In the case of blood cell counting, the 60 mean diameter of red blood cells shows some variation from person to person. Also, some diseases result in changes in cell diameter. These diameter variations will tend to introduce an error, as will now. be explained.

Each particle in the field will be scanned by more than 65 one line of the television raster and may, thus, be counted several times. Obviously, larger celis will be scanned more times than small cells. Therefore, in an expression relating pulse count to the number of cells in· the field, the cell diameter would appear as- a variable which 70 should be eliminated, Since the cells are round,, and since very few lines of the raster will pass exactly, through the cell center in cutting across cells, most of the pulse

3,781,303 widths will be shorter than the cell diameter. However, probability calculations have indicated that the “most probable diameter” differs from the real diameter by a small and, more important, practically constant amount. This effect would then be small if the shape of the cell remains constant. In practice, the error so produced has proven negligible.

Since the pulse widths produced will be directly proportional to the cell diameters, it is possible to determine the “unknown diameter term” in the expression relating pulse count to particle number. Therefore, the accuracy of the counting process will hinge on having means for measuring the length of the video pulses, and for correlating this pulse width information with the pulse count information in such manner as to correct for the variable particle size factor.

To this end, and in accordance with the invention, the second channel to which the pulses are applied in the apparatus of Fig. 1 comprises a pulse width detecting circuit 28. This circuit may take any one of a number of different forms, one specific example being given hereinafter. In general, the circuit 28 should comprise some type of indicator 30, and a means for adjusting the pulse width to which the circuit will be most sensitive. In other words, the pulse width sensitivity control 32 would be adjusted to obtain a maximum reading on the meter 30. This would indicate that the sensitivity control has been set so that the circuit 28 is responding to the maximum number of pulses, or, conversely that the pulse width sensitivity control 32 has been adjusted to correspond with the average width of the video pulses. Further in accordance with the invention, this pulse width sensitivity control 32 can be ganged to a calibration control 34 for the meter 26. Thus, as the pulse width sensitivity control is adjusted to respond to the maximum number of pulses, the pulse counting circuit sensitivity will be furnished with a correction factor which will minimize the error due to variations in particle size. Without this correction, a total of, say, 5,000 particles of one given average diameter might produce the same reading on the meter 26 as a total of, say, 4,500 cells of different average diameter. The correction factor introduced has the effect of making the pulse count meter 26 indicate the number of pulses that would be produced if all the particles had a diameter equal to their average diameter.

It will be understood that the present invention is not limited to the use of a pulse generator-scanner of the specific type shown in Fig. 1. For example, a photomicrograph or electron micrograph transparency may be made of the area containing the particles to be studied. The beam of energy for scanning the area may be a focussed beam of light, scanned mechanically, or a focussed beam of light generated by a stream of electrons striking the fluorescent surface coating of a cathode ray kinescope tube. An arrangement of this latter type is shown in Fig. 2, and will be discussed hereinafter. The field of view may be either translucent or opaque to the beam of energy, and the residual energy of the beam may be detected either in a transmitted ray or a reflected ray. Where “scanning” is referred to it is, therefore, meant to encompass any one of these equivalents. Further, the particles being counted may appear either as portions of the field of view relatively more opaque to the energizing beam of energy or relatively more transparent thereto. Where the entire field is opaque, the particles may simply have a reflective property different from that of the general background.

The pulse generator-scanner system shown in Fig. 2, comprises a kinescope tube 40, on the fluorescent screen of which is produced a raster used as a flying spot light source. A lens system 42 is provided to image the raster on a transparency 44, which comprises a field of view with particles therein to be counted. Another lens system 46 may be used to focus the light transmitted through the transparency during the scanning process on a photo4 electric cell 48. The photoelectric cell is connected to a voltage source 50 and load impedance 52 in the usual manner, to provide across the load impedance 52 pulses representative of changes in the light reaching the photocell 48.

Turning now to a consideration of a more specific embodiment of the invention, the circuit shown in Fig. 3 includes a pulse generator-scanner 60 such as the television camera 14 of Fig. 1 or the scanning system shown in Fig. 2. Pulses generated by scanning of the viewing field 10 may be amplified and clipped in a conventional amplifier and clipper circuit 62 to provide pulses, A, of relatively uniform amplitude. Pulses from the amplifier and clipper 62 may be supplied to a differentiator circuit 64, the output of which will comprise a positive “pip,” B, corresponding to the leading edge of each particle-representing pulse, and a negative “pip,” C, corresponding to the pulse trailing edge. Following the dif- j ferentiator, a phase inverter amplifier 65 is provided. At the amplifier-inverter output, the positive pips (B) will appear as negative pips, B¹, and the negative pips (C) will appear as positive pips C¹. The reason for this phase inversion will be brought out hereinafter.

From the output of the amplifier-inverter 65, two circuit paths are provided. One comprises a connection from the amplifier 65 to the control grid 68 of a pentode tube 66. The anode 70 of the tube 66 is connected through a meter 72 to an operating voltage source (not shown), designated “B+”. The tube cathode 74 is connected to an adjustable biasing network 76 of conventional form. A pair of resistors 78, 80 connected in series to the voltage source B-j- provides a voltage divider from which adjustable bias voltage can be applied to the tube cathode 74.

The adjustable resistor 80 normally is set so that a cut-off bias will be applied to the tube 66. Therefore, the negative pips corresponding to the leading edges of the original pulses will produce no effect on the tube 66, while the positive pulses may produce plate current flow, depending on the condition of the tube suppressor grid 90, as is explained hereinafter.

The differentiated and inverted pulses also are applied through a coupling diode 82 and an inverter-amplifier 84 to a variable delay line 86. For example, the delay line 86 may be of the type described and claimed in the copending application of H. Kihn, Serial No. 176,324, filed July 28, 1950 and assigned to the assignee of the present invention now Patent No. 2,619,537 issued November 25, 1952. A switch 88 is provided for connecting the delay line 86 to the suppressor grid 90 of the tube 66. With the switch 88 in this position, the suppressor grid 90 is grounded through a resistor 93. Due to the positive cathode voltage, the suppressor grid effectively will be receiving a cut-off bias voltage.

To determine the average diameter of the particles being counted, the switch 88 is placed in the position shown to connect the suppressor grid 90 to the adjustable delay line 86. Also, a switch 92, linked to the switch 88, is opened to remove an adjustable shunt 94 from the circuit of the meter 72. As to any pulse for which the delay of the variable delay line 86 is the same as the pulse width, positive pulses will arrive simultaneously at the control grid 68 and the suppressor grid 90 of the tube 66. This pulse coincidence will allow tube current to flow through the meter 72. Therefore, it can be seen that the adjustment of the delay line 86 which provides maximum reading on the meter 72 will be an adjustment corresponding to the maximum number of pulses of that width, and hence, a measure of the average pulse width. At this point, it should be noted that the adjustable shunt resistor 94 is mechanically linked to the delay line 86.

After this preliminary adjustment has been made, the circuit from the delay line 86 to the suppressor grid 90 is opened by actuating the switch 88. This will connect the suppressor grid 90 to the tube cathode 74, removing

2,731,202 the cut-off bias from the suppressor grid. Simultaneously, the switch 92 will shunt the resistor 94 across the meter 72. All positive pulses arriving at the tube control grid 68 now will cause plate current flow which will be integrated by the meter 72. By virtue of the mechanical connection between the meter-adjust resistor 94 and the adjustable delay line 86, adjustment of the latter automatically will select the value of shunt appropriate for a particular length of pulse. This arrangement automatically eliminates the variable diameter term from the expression relating pulse count, pulse width, and particle number, thereby minimizing the error due to particle size variation.

Tests made with an instrument such as that shown in T Fig. 3 have indicated that an accuracy of plus or minus five per cent is attainable in blood cell counting. Fur, thermore, the instrument shown can be provided with means to indicate the average diameter of red blood cells, as, for example, a calibrated control 98 for the delay line 86. This is a quantity which may well have some clinical value, although it has rarely been used heretofore because of the difficulty of determining it.