This invention relates to a microscope focus control system for the fast measurement and control of image focus in microscopy.
The ability of an optical system to recognise or maintain a particular focus position with respect to a series of objects at different distances is often a very important requirement, and many ways of doing so have been developed over the years. In a microscope, images are viewed at a very high magnification, of perhaps 100 or more, which means that very small changes in object distance of the micrometre scale resulting from mechanical instabilities in the optical system can be sufficient to significantly change the focus.
With the increasing prevalence of electronic methods of image acquisition, the possibilities of processing such images to automatically determine and set what is likely to be the “best” focus are being exploited, but processing of signals from an array or imaging detector can impose a significant computational overhead and slow the speed at which the focus can be adjusted to compensate for the movement. Furthermore, the design of such systems may require them to be physically integrated within the design of the microscope itself
In accordance with the invention there is provided a microscope focus control system for monitoring changes in the position of a microscope objective comprising an illumination source emitting any radiation used in microscopy, for example visible light, ultra-violet or infra-red radiation, and a polarisation-sensitive detector positioned to receive a reflected beam from an interface having a reference refractive index, such as a glass cover slip placed over a specimen, and configured to detect polarisation components of the reflected beam, wherein an elongate aperture is disposed proximal the illumination source to generate a rectangular beam to impinge on the interface and a second aperture is located at a position corresponding to an unfocused region of the reflected beam from the interface so as to direct unfocused reflected light onto the polarisation-sensitive detector, such that changes in polarisation components of the reflected beam are used as a measure of change in position of the interface relative to an objective.
The reflected beam is preferably generated by Fresnel reflection from the interface.
Changes in polarisation components, and thus the s and p components, in the reflected beam may be used to generate signals to control movement of microscope equipment and thereby to maintain the interface and objective at fixed distance with respect to each other, and so to restore focus when movement has occurred away from a focused position. Typically this is performed by using a z-axis drive and associated controller to move the objective.
The position corresponding to the unfocused region where the second aperture is located is preferably further from the polarisation-sensitive detector than a position corresponding to a focused region of the reflected light, i.e. the focal point of the reflected light.
An illumination beam from the illumination source may be polarised to control relative contributions of s and p components within the illumination source.
The illumination source is preferably a laser diode and may generate wavelengths in the infra-red.
The interface is preferably an uncoated glass coverslip as is used to cover a microscope sample mounted on a glass slide.
If desired, a magnifying lens may be disposed between a focal point of the reflected beam and the polarisation-sensitive detector so as to increase magnification and so increase sensitivity.
The polarisation-sensitive detector is preferably locatable proximal an imaging sensor associated with a microscope. Alternatively a dichroic mirror or other optical splitter element may be positioned between the interface and the polarisation-sensitive detector so as to allow reflected light to be separated from imaging light resulting from a microscope imaging source and thus allowing for separation of the polarisation-sensitive detector and microscope image plane.
The illumination source may be of a different wavelength to an imaging source associated with a microscope with which the system is to be used.
The system may further comprise a feedback control system or controller responsive to changes in polarisation components of the reflected beam to maintain a constant distance between an objective and the interface.
The system may further comprise an actuator, such as a z-axis drive, moveable responsive to changes in polarisation components of the reflected beam to maintain a constant distance between an objective and the interface so as to restore original distance and focus position between objective and interface.
The microscope focus control system as aforesaid is suitable for retrofitting to a microscope system.
The invention also lies in a microscope system incorporating a microscope focus control system as aforesaid.
The invention will now be described by way of example and with reference to the accompanying drawings in which:
The underlying difference between two major imaging applications of a camera and a microscope is shown in
The extent of the out-of-focus blur depends on the optical aperture of the system, which is a measure of the range of angles over which the system lens can collect light from an object. In the application of a conventional camera system 10 as is shown in
However, in the field of microscopy, the situation is very different. For a microscope as shown in
The invention enables focus monitoring and control for a microscope and is based on the angular polarisation dependence of the reflection from a refractive index interface, known as Fresnel reflection. The general principle of Fresnel reflection will now be explained.
As shown in
n
1 sin θi=n2 sin θt Equation 1
However, some fraction of incident light 30 will instead be reflected as beam 40 at an angle θr, where
θi=−θi Equation 2
and it is this light 40 undertaking Fresnel reflection which is utilised in the present invention.
If the angle of incidence θi, is large, the Fresnel reflected light 40 shows a substantial dependence on the polarisation state of the incident light. Whether explicitly polarised or not, light can be viewed as a mixture of two mutually perpendicular polarisation components, termed s when the electric vector is at right angles to the refractive index interface, and p when the electric vector is parallel to it. “Unpolarised” light can be viewed as having equal s and p components.
For Fresnel reflection at incident angles θi of less than about 10°, the reflected intensity r is essentially polarisation independent when it approximates to
r
tot
2=[(n1−n2)/(n1+n2)]2 Equation 3
where n1 and n2 are the refractive indices on either side of interface 32. In the typical case of an air-uncoated glass interface such as encountered when a specimen is mounted on a slide and a glass slip cover is placed over the specimen, the refractive index of air n1 is 1 and the glass slip value n2 is approximately 1.5, which means that about 4% of the light is reflected.
The reflection becomes progressively more polarisation-dependent as the angle of incidence θi increases, and the following relations for the reflected intensities r s and r p of the s and p components apply:
r
s
2=sin2(θi−θt)/sin2(θi+θt)] Equation 4
and
r
p
2=tan2(θi−θt)/tan2(θi+θt)] Equation 5
where θi and θt are the incident and transmitted angles at interface 32 as shown in
These relations are effectively identical for small angles of incidence, because of the increasing equivalence of the sine and tangent functions, but for larger angles the difference may be substantial, as shown in
By using a polarisation-sensitive detector to measure the relative intensities of the s and p components of beam 40 that is reflected from an air:glass interface 32, which in this case will be the coverslip between the objective and the specimen, information can be gained about the incident angle θi of incident light 30. Such a detector typically consists of a polarising beam splitter that separates and directs the s and p components of Fresnel light 40 to a pair of sensors, typically photodiodes, that allow the individual intensities of the s and p components to be independently and rapidly measured. Whilst a Fresnel reflection is particularly preferred, any other form of reflection with sufficient polarisation dependence could be utilised for focus control, for example specular reflections from a mirrored metallic surface exhibit polarisation dependence as the incident light goes off axis.
Such a polarisation-sensitive detector can be used as the basis for a microscope focus control system as shown in
Although objective 50 typically has many elements with optical surfaces within it, these are generally coated with anti-reflective coatings. As shown by the ray paths in
If the distance between objective 50 and coverslip 54 changes, then the focus position of reflected beam 40 also changes, and this in turn changes the amplitude and the s:p polarisation ratio of the signal reaching polarisation-sensitive detector 66.
Accurate focus stability and control is important when imaging at high magnifications, such as for a microscope, as very small changes in object distance can substantially degrade the focus. Where an object moves to a slightly different distance from the objective, for example by vibration, the object's image will also move to a different distance from the tube lens, but this difference is much greater as it is given by the difference in object position multiplied by the square of the magnification. This is as a result of the longitudinal magnification of the microscope.
To direct light into objective 50 and illuminate sample 52, 45° mirror 62 is used to reflect collimated beam 60 into space 64 between objective 50 and tube lens 56. The most common application is the excitation of fluorescence within the specimen where the fluorescence emission will be at longer wavelengths than the wavelength of beam 60. Mirror 62 is chosen to have properties such that these longer wavelengths are transmitted rather than reflected.
To detect the changes in focus position using Fresnel reflections, a microscope focus control system comprising polarisation-sensitive detector 66, illumination source 70, slit 72, aperture 106 and additional lenses and mirrors as required are incorporated into the microscope structure as shown in
To ensure mirror 76 does not obstruct detection pathway 78 for reflected beam 40, mirror 76 is selected to reflect only a small proportion of the reflected light 40 resulting from illumination source 70, so that mirror 76 is largely transparent for detection. The illumination efficiency will be correspondingly reduced, but can be compensated for by using a more powerful beam, such as that from a laser diode.
For clarity,
In a practical system, polarisation-sensitive detector 66 cannot share the same physical space as imaging sensor 82, which also normally needs to be located at image plane 58. These two components can be located side-by side, as shown in
Some microscopes offer a more extended collimated space between objective 50 and tube lens 56 which instead allows mirror 84 to be located in the space between mirror 62 and tube lens 56, although then a second tube lens is needed between mirror 84 and imaging sensor 82′ in order to focus beam 86 onto the sensor 82′.
There is a potential problem in respect of discriminating between what is s and what is p at polarisation-sensitive detector 66 compared with at the reflection interface if illumination is with a circular beam. This is because for light arriving at and then leaving coverslip 54 at some angle to the optical axis as shown in
It is to address this potential problem, narrow rectangular slit 72 is incorporated within illumination light source 70 to create a beam in the form of a narrow rectangular sheet 98 instead of a circular beam, with the limits of beam 98 shown by the boxed outline in
When the distance between objective 50 and coverslip 54 changes, then the focus position of reflected beam 40 also changes, and this changes the s:p polarisation ratio of the signal received by sensors 68 within polarisation-sensitive detector 66.
If beam 40 were fully in focus at aperture 112, then there would initially be no change in the signal received by detector 66 if the distance between objective 50 and coverslip 54 were to change. Then beyond some change threshold, the detected signal would be reduced, but that by itself would give no information on its polarity. Therefore, instead beam 40 is configured to be out of focus at aperture 112 by ensuring focus point 104 is not coincident with aperture 112, which ensures that a change in the distance between objective 50 and coverslip 54 either increases or decreases the focus and so gives the polarity of the focus shift. The out-of-focus region selected for the location of diaphragm 106 can in principle be on either side of image plane 58, but in practice the nature of the specimen and the optical properties of the objective with respect to out-of-focus light may favour one side or the other. Also the distance of the detector assembly, and in particular diaphragm 106 and aperture 112, in respect of its distance from tube lens 56, can be chosen to give an offset between the distance into the specimen to be monitored and stabilised, and the distance to the reflection point at the air:glass interface of coverslip 54.
For
As the reflected Fresnel beam 40 goes out of focus at aperture 112, the more peripheral regions of beam 40, which are relatively deficient in the p polarisation component as shown in
Reflections from other surfaces in the optical system, for example within objective 50, will be at close to normal incidence. Not only are these other surfaces likely to be coated to reduce such reflections, but also all these reflections will have essentially equal s and p contributions and their signal components will cancel. Conversely, the reflections from uncoated coverslip 54 occur at much higher angles, and so their s and p contributions can be significantly different.
One way of measuring the s:p polarisation ratio is to divide the s component signal into the p component signal but this either requires calculations in software or specialist analogue circuitry. Alternatively, and as preferred for the present embodiments, two signals representing the s and p components of the reflected beam can be made effectively identical when the sample is at focus, then any changes in focus can be quantified by subtracting one of the signals from the other to give a small residual signal used to control a feedback system, resulting in a faster and more accurate way to measure a small change in the ratio of these components.
Thus at a particular focus position, the relative gains in feedback control signals generated in response to the s and p signals within reflected beam 40 are adjusted so that the gains are made the same and a feedback baseline is created such that when one gain signal is subtracted from the other, there is a zero difference signal. Any shift in focus will generate a residual feedback signal, which will vary according to the degree of focus shift and change in the s and p polarisation components. This feedback signal can be used as a measure of the change in focus, or it can be used in an electromechanical feedback loop to restore the focus to its original position. In this type of feedback control system, the residual signal is sent to a controller that moves objective 50 in such a direction as to drive the residual signal back towards zero, and hence to restore the focus. Alternatively, deliberate offset signals can be introduced into the feedback loop, to move the objective to other focus positions, in order to acquire a “z stack” of images at different focus positions through the sample.
Thus any change in focus causes an immediate change in the signal recorded by polarisation-sensitive detector 66, and this change is immediately converted by controller 108 into a feedback signal to adjust z-axis drive 110 and the position of objective 50 relative to coverslip 54 in order to restore the focus. Controller 108 and z-axis drive 110 have been omitted from
The illumination level of light 70 is typically set so that the detected s and p signals are sufficiently above the system background noise to provide fast and sensitive operation, while at the same time being no higher than necessary to achieve this, in order to minimise undesirable effects such as the risk of overloading polarisation-sensitive detector 66, or of illumination signal 74 breaking into the imaging pathway. This can be achieved by automatically setting the illumination level to achieve this optimum condition when the system is switched on, and then subsequently controlling it with respect to this reference level.
Microscopes are fundamentally high magnification devices and when a shift occurs between the relative position of the objective and the sample, for example by mechanical vibrations affecting a sample stage, a change in the relative position will result in a change in the image position by the change in relative position multiplied by the square of the magnification. Thus a small change in distance at objective 50 causes a large shift in the focus position of the image. It is for this reason that accurate focus stability and control is particularly important when imaging at high magnifications, as very small changes in object distance can substantially degrade the focus.
Considering a specific example of a microscope with a ×40 objective of NA (numerical aperture) 1.4, the angular range of a beam focused at image plane 58 is the inverse sine of the NA of the objective, divided by its magnification, hence plus and minus 2.0°. The focal length of tube lens 56 that produces the image at image plane 58 is typically between 165 and 200 mm, and
This gives us a choice of sizes and positions for diaphragm 106 and aperture 112, with two alternatives 106, 112 and 114, 116 shown as examples in
For aperture 112 at the position shown in
An alternative approach to increase sensitivity is to introduce further magnification into the system, as shown in
This is explained further in more detail with reference to
For comparison,
In
The introduction of lens 118 as shown in
If desired, illumination beam 74 generating the Fresnel reflection 40 can be polarised to control the relative contributions of s and p components within the beam. If the beam is polarised at 45° to the detected s and p directions, then its s and p contributions will be effectively equal, but as the angle is rotated either side of this value, either the s or the p states are progressively favoured. Polarisation is of assistance because the detection system measures the relative values of the s and p states, and for detection it is preferable for the comparison to be made on s and p values in the reflected beam that are nearly equal. Polarisation of the illumination beam is achieved by rotating the polarisation angle of illumination beam 74 using a polariser located between illumination source 70 and mirror 76.
For a given shift in focus, the absolute value of the difference will depend on the intensity of the illumination. If it is desired to quantify the difference between the s and p contributions of the coverslip reflected beam 40 that reaches polarisation-sensitive detector 66 in respect of an absolute positional change, rather than just to use it as a feedback control signal to keep the focus constant, then the intensity must be kept constant as well.
One option for keeping the illumination intensity constant, is to have a further sensor that measures the illumination intensity at source 70, and which then stabilises it by an electronic feedback loop. As an alternative, the illumination can be controlled so that the sum of the s and p detected signal components remains constant which should have a broadly similar effect.
Another modification for use with the system if desired is using electronic background subtraction to remove any spurious signals resulting from any background light that may reach polarisation-sensitive detector 66, or of any background signals within electronics associated with detector 66. This is achieved by switching or pulsing the illumination on and off at a relatively high frequency compared with the measurement bandwidth, which in this case means using a frequency of typically a few kilohertz. The detection of each of the s and p signals is itself made differentially, so that the level of each when the illumination is off is subtracted from that when the illumination is on. These signals can then easily be filtered in order to remove the switching frequency, so that they are then effectively continuous as seen by the subsequent detection electronics. This will allow use of a lower illumination intensity for a given level of performance, and hence a correspondingly lower risk of illumination beam 74 interfering with image sensor(s) 82 in any way.
The focus monitoring and control described requires just a single polarisation-sensitive detector monitoring changes in s and p components from a single out-of-focus image. The change in polarisation composition of the detected signal is used as a measure of any change in the position of the objective relative to the refractive index interface, and thus any change in the focus. Any change in polarisation composition of the detected signal is used in a feedback control system to maintain a constant distance between the objective and the refractive index interface, such as the glass coverslip. This assists with speed of detection and correction of focus changes as less processing is required to quantify the changes in focus and the adjustments required to restore focus. The system also allows for sensitivity to be adjusted if desired by introducing a magnifying lens.
Number | Date | Country | Kind |
---|---|---|---|
2212300.4 | Aug 2022 | GB | national |