Circuit and Method for Fluorescence Lifetime Imaging

Abstract

The detection system for detecting fluorescence lifetime comprises an excitation light source configured to repeatedly generate pulsed excitation light, and a detector (600). The detector (600) includes a single-photon detection circuit (150) for generating a digital pulse upon detection of a photon; a pulse suppression circuit (165) for rejecting detected photons that occur outside each one of a series of measurement time windows, each subsequent measurement time window beginning after the subsequent excitation light pulse has ceased and ending before the next excitation light pulse is generated, each measurement time window having a measurement window duration; and a switched capacitor circuit (160) having an input terminal for receiving a voltage ramp signal that restarts with each new measurement window duration, the switched capacitor circuit configured to iteratively calculate an average voltage based on an exponential moving average function applied to sample voltages recorded over past measurement time windows and determined by the voltage ramp signal in response to photons that have not been detected and rejected, according to the principles of the center-of-mass method, the switched capacitor circuit having a node (172) for outputting the calculated average voltage as a measure of the fluorescence lifetime.

Description

Technical Field

[0001] The present invention relates to the field of capturing fluorescence in the time domain and determining its lifetime.

Background Art

[0002] A well-known method for realizing fluorescence lifetime imaging is in the frequency domain. The light that stimulates the phosphor is sinusoidally modulated, and the received fluorescence is demodulated at the same frequency. The phase change and modulation depth are the main parameters for extracting the lifetime of the phosphor.

[0003] Another method uses the principle of rapid lifetime determination. By using a very fast gated detector to collect the number of photons within two time windows (or bins), the optical response is measured.

[0004] A system based on a single photon detector has been proposed. Either a gated window is used, or the time of an event is recorded using a counter and / or a ring oscillator (time-to-digital conversion), and the data is collected and passed to a DSP processor that processes the events.

Summary of the Invention

[0005] An object of an embodiment of the present invention is to provide a good system and method for determining fluorescence lifetime.

[0006] The above object is achieved by the method and apparatus according to the present invention.

[0007] In a first aspect, the present invention provides a detection system for detecting fluorescence lifetime. The detection system includes an excitation light source configured to repeatedly generate pulsed excitation light and a detector. The detector includes a single photon detection circuit for generating a digital pulse upon detection of a photon and a pulse suppression circuit for rejecting detected photons occurring outside each of a series of measurement time windows, each subsequent measurement time window starting after a subsequent excitation light pulse has ceased and ending before the next excitation light pulse is generated, each measurement time window having a measurement window period, the pulse suppression circuit, and a first switched capacitor circuit having an input terminal for receiving a voltage ramp signal restarted in each new measurement window period, the first switched capacitor circuit being configured to repeatedly calculate an average voltage based on an exponentially weighted moving average function applied to a sample voltage determined by the voltage ramp signal in response to photons recorded, detected, and not rejected over past measurement time windows according to the principle of the centroid method. Measurements are made over subsequent measurement time windows and are always averaged such that the importance of previously acquired samples in the average decreases with the passage of time. The switched capacitor circuit has a first node for outputting the calculated average voltage as an indirect measure of fluorescence lifetime. The calculated average voltage is a voltage and not a time value, and for at least two reasons, namely that a calibration correction and / or a background correction may need to be applied to the calculated average voltage depending on the setup situation in order to obtain the actual fluorescence lifetime, the calculated average voltage is an indirect measure of fluorescence lifetime.

[0008] In an embodiment of the present invention, the exponentially weighted moving average function is implemented based on the ramp signal by sampling the voltage of the ramp signal when a digital pulse corresponding to a detected and not rejected photon is present. The sampled voltage of the ramp signal is used each time to update the exponentially weighted moving average voltage.

[0009] In the detection system according to an embodiment of the present invention, the first switched capacitor circuit may include a first sampling capacitor, a first switch, and a second switch for alternately coupling the first sampling capacitor to a ramp terminal for receiving a ramp signal and a first node. In this way, the ramp signal is sampled, and the sampled voltage is used to update the exponentially weighted average voltage as described above.

[0010] In an embodiment of the present invention, the first switched capacitor circuit may further include a second capacitor configured to be in a charge sharing and redistribution configuration when the second switch is driven to couple the first sampling capacitor to the first node. In an embodiment of the present invention, the first sampling capacitor is at least one order of magnitude smaller, preferably at least two orders of magnitude smaller, than the second capacitor.

[0011] The detection system according to an embodiment of the present invention may further include means for temporarily adding one or more capacitors in parallel with the first sampling capacitor.

[0012] The detection system according to an embodiment of the present invention may further include a second switched capacitor circuit connected in series to the first node of the first switched capacitor circuit and configured to operate at an oscillation rate that does not directly respond to incident photons. The oscillation rate may be determined according to the requirements of the application.

[0013] The second switched capacitor circuit may include a second sampling capacitor, a first node for receiving an average voltage calculated as a measure of fluorescence lifetime by the second sampling capacitor, and a first switch and a second switch for alternately coupling to an output node. In an embodiment of the present invention, the second switched capacitor circuit may further include a fourth capacitor configured to have a charge sharing redistribution configuration when the second switch is driven to couple the second sampling capacitor to the output node. The second sampling capacitor may be at least one order of magnitude smaller, preferably at least two orders of magnitude smaller, than the fourth capacitor.

[0014] A detection system according to an embodiment of the present invention may further include a non-overlapping switch enable circuit for providing a non-overlapping signal for operating the switched capacitor circuit. In an embodiment of the present invention, the non-overlapping switch enable circuit includes a one-shot circuit that generates an output pulse of a predefined duration when triggered, then returns to a stable state and does not generate any further output until triggered again.

[0015] According to an embodiment of the present invention, the single photon detection circuit may include a single photon avalanche diode (SPAD). In such an embodiment, the pulse suppression circuit may include a variable voltage source adapted to lower the voltage across the single photon avalanche diode. In an embodiment of the present invention, the pulse suppression circuit may be adapted to block one of the signals for driving the switched capacitor circuit to prevent the pulse from being taken into account.

[0016] A detection system according to an embodiment of the present invention may further include at least one additional pulse suppression circuit and at least one additional switched capacitor circuit configured to operate in parallel with the pulse suppression circuit and the switched capacitor circuit.

[0017] The detection system according to an embodiment of the present invention may further include a photon counter circuit for counting the number of detected photons. In an embodiment where the detector includes a non-overlapping switch enable circuit for providing a non-overlapping signal for operating the switched capacitor circuit, the photon counter circuit may include a switched capacitor circuit adapted to be operated by a non-overlapping signal for operating the switched capacitor circuit.

[0018] In embodiments including means for temporarily adding one or more capacitors in parallel with the sampling capacitor, these embodiments may each include one or more switches in series with the one or more capacitors, and the one or more switches are opened when the photon counter circuit has counted a predetermined number of detected photons.

[0019] In a second aspect, the present invention provides a fluorescence imaging sensor, comprising: an excitation light source configured to repeatedly generate pulsed excitation light; an array of detectors, each detector including a single photon detection circuit for generating a digital pulse upon detection of a photon, and a pulse suppression circuit for rejecting detected photons occurring outside of each of a series of measurement time windows, each subsequent measurement time window starting after a subsequent excitation light pulse has ceased and ending before the next excitation light pulse is generated, each measurement time window having a measurement window period; a first switched capacitor circuit having an input terminal for receiving a voltage ramp signal restarted in each new measurement window period, the first switched capacitor circuit being configured to repeatedly calculate an average voltage based on an exponentially weighted moving average function applied to a sample voltage determined by the voltage ramp signal in response to photons recorded, detected, and not rejected over past measurement time windows according to the principle of the centroid method, the first switched capacitor circuit having a first node for outputting the calculated average voltage as a measure of fluorescence lifetime; and an array of detectors. Details of the detectors within the array can be as presented in various embodiments of the first aspect of the present invention.

[0020] In a third aspect, the present invention provides a method for determining fluorescence lifetime. The method includes generating a digital pulse upon detection of a photon, rejecting detected photons occurring outside of a measurement time window, calculating an average voltage based on an exponentially weighted moving average function applied to a sample voltage determined by a voltage ramp signal in response to photons recorded, detected, and not rejected over past measurement time windows according to the principle of the centroid method, and outputting the calculated average voltage as a measure of fluorescence lifetime. In embodiments of the present invention, the calculated average voltage may be further averaged.

[0021] The advantages of the embodiments of the present invention are that the timing of photon arrival, which is somewhat available when using a single photon detector such as a SPAD, is fully utilized in lifetime determination thanks to the use of the centroid principle. Furthermore, each of the events detected within the time window is involved in the determination of the lifetime estimate. Since that window can cover most of the available fluorescence, basically, this system can be considered close to optimal. For example, using a single window of 20 ns, lifetimes in the wide range of 200 ps to 15 ns can be detected. This system can operate continuously and asynchronously, obtaining the best lifetime estimate of the number of recent events / detections without the risk of the detector overflowing or, on the other hand, having too few signals. The situation of very low light intensity can also be handled, and the system is mainly shot noise limited. For this reason, a sliding average based on exponential moving average with a configurable memory depth based on the capacitor ratio is used. Older measurements are taken into account less, and newer measurements are considered more. The readout value may be analog, but the ADC does not require a large number of effective bits. Pixel averaging can be complemented, for example, by a second stage of averaging in the DSP to achieve optimal results. The readout of the pixel array is not timing critical and can be done in the background. Furthermore, a second longer window in which the centroid in the time domain is averaged can be operated simultaneously to take into account the background light and / or dark count rate of the single photon detector. A third window makes it possible to discover the lifetimes of multiple phosphors and their relative importance. The aspect of delay in a part of the system can be investigated using windows at different positions. In this way, for example, it is possible to discover the occurrence of fluorescence, which is useful for positioning the window for lifetime estimation closer to its optimal position. The circuit for estimating the centroid in the time domain in a time window including averaging is very small, and multiple instances of these can be included for each pixel for simultaneous operation. Continuous operation of different time windows is also achievable.If classical illumination levels need to be detected simultaneously, photon counters based on similar circuits are also proposed.

[0022] Specific and preferred embodiments of the invention are set out in the appended independent and dependent claims. Features from the dependent claims may be combined with the features of the independent claims and the features of other dependent claims as appropriate and need not be merely as explicitly set out in the claims.

[0023] For the purpose of summarizing the advantages achieved over the prior art and the present invention, specific objects and advantages of the present invention have been described above in this specification. Of course, it should be understood that not all such objects or advantages may be achieved by any particular embodiment of the present invention. Thus, for example, those skilled in the art will recognize that the invention may be practiced or carried out in a manner that realizes or optimizes one advantage or group of advantages taught in this specification without necessarily realizing other objects or advantages that may be taught or suggested in this specification.

[0024] The above and other aspects of the invention will become apparent with reference to the embodiments described below and will be described.

Brief Description of the Drawings

[0025] The present invention will be further described below by way of example with reference to the accompanying drawings.

[0026]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Number

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

[0027] The drawings are merely schematic and not limiting. In the drawings, the sizes of some of the elements are exaggerated for illustrative purposes and may not be drawn to scale. Dimensions and relative dimensions do not necessarily correspond to actual reductions for the implementation of the invention. In different figures, the same reference numerals refer to the same or similar elements.

Best Mode for Carrying Out the Invention

[0028] The present invention will be described with reference to specific drawings with respect to specific embodiments, but the present invention is not limited thereto and is limited only by the claims. Any reference signs in the claims should not be construed as limiting the scope thereof.

[0029] Terms such as first, second, etc. used in this specification are used to distinguish between similar elements and are not necessarily used to describe an order in any temporal, spatial, ranking, or other manner. Such terms are interchangeable under appropriate circumstances, and it should be understood that the embodiments of the present invention described in this specification can operate in an order other than that described or illustrated in this specification.

[0030] Note that the term "comprising" used in the claims should not be construed as being limited to the means recited thereafter. Thus, the scope of the expression "an apparatus comprising means A and B" should not be limited to an apparatus consisting only of components A and B. This expression means that, with respect to the present invention, the most important components of the apparatus are A and B.

[0031] Similarly, note that the term "coupled" should not be construed as being limited to direct connection only. Thus, the scope of the expression "apparatus A coupled to apparatus B" should not be limited to an apparatus or system in which the output part of apparatus A is directly connected to the input part of apparatus B. This expression means that there exists a path between the output part of A and the input part of B, which can be a path including other apparatuses or means.

[0032] Throughout this specification, the mention of "one embodiment" or "an embodiment" means that the particular features, structures, or characteristics described in connection with that embodiment are included in at least one embodiment of the invention. Thus, the appearances of the phrases "in one embodiment" or "in an embodiment" in various places in this specification are not necessarily all referring to the same embodiment, although they may be. Further, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments as would be apparent to one of ordinary skill in the art from this disclosure.

[0033] Similarly, in the description of exemplary embodiments of the invention, for the purpose of rationalizing the disclosure and facilitating understanding of one or more of the various aspects of the invention, it should be understood that the various features of the invention may be grouped in a single embodiment, figure, or description thereof. However, this method of disclosure should not be construed as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as the following claims reflect, aspects of the invention lie in less than all of the features of a single or more of the disclosed embodiments. Thus, the claims following the detailed description are hereby expressly incorporated herein by reference, and each claim stands on its own as a separate embodiment of the invention.

[0034] It should be noted that the use of a particular terminology in describing a particular feature or aspect of the invention is not meant to be limiting such that the terminology encompasses any particular characteristics of the feature or aspect of the invention in question.

[0035] In the description provided herein, numerous specific details are disclosed. However, it will be understood that embodiments of the invention may be practiced without these specific details. In other instances, well-known methods, structures, and techniques have not been shown in detail in order not to obscure the understanding of this description.

[0036] Definitions Fluorescence is the emission of light by a phosphor that has absorbed electromagnetic excitation radiation, such as light. In most cases, the emitted light has a longer wavelength than the absorbed excitation radiation and, therefore, has lower photon energy.

[0037] Fluorescence lifetime is a measure of the time it takes for a phosphor to return to its ground state by emitting photons from its excited state. The lifetime of a phosphor can range from picoseconds to hundreds of nanoseconds.

[0038] Fluorescence lifetime imaging microscopy (FLIM) is a technique for generating images based on the differences in the exponential decay rates of fluorescence from a fluorescent sample. In FLIM, the lifetime rather than the intensity of the phosphor signal is used to create the image. For example, in image-guided surgery and biomedical applications, the lifetime provides additional information about the chemical / biological environment of the phosphor, and many other applications (not necessarily called FLIM) require fluorescence lifetime imaging.

[0039] When a phosphor is irradiated with a pulse of electromagnetic radiation, such as a light pulse, for example, illuminated, to excite its internal state, the electrons will move to a higher energy state for a while. The electrons will return to their original state by emitting photons or heat in an exponential decay process characterized by the fluorescence lifetime of the phosphor. The average time that the phosphor remains in its excited state and emits photons is called the fluorescence lifetime. In FIG. 1, the applied excitation light 100 is a short and intense pulse having a predetermined wavelength from a radiation source such as a laser, and at the instant t of stimulation sis assumed to have a peak. Fluorescence 105 (typically exhibiting a longer wavelength) starts at the instant t of the stimulus when the phosphor is excited s and after the excitation stops, the fluorescence 105 shows an exponential decay with the lifetime τ of the phosphor. The curve of the generated fluorescence 105 is a linear curve with a downward slope of -1 / τ on a logarithmic scale. One of the known measurement techniques for quantifying the lifetime τ is called the center-of-mass method. To apply this method, when the excitation light 100 stops, for example, starting at the first instant t0 and up to the second instant t1, preferably a window period t w can be defined (t w =t1 - t0). The luminescence then follows the following equation:

Equation

Equation

[0040] In an image sensor where the pixels are small and the number of photons derived from the fluorescence scene is small, it is not possible to rely on single-shot measurements such as those proposed in FIG. 1. In reality, during a single window period t w it is very likely that only a few photons or one photon will reach the detector, or no photons will reach the detector at all. At this time, the excitation light pulse 100 is reapplied many times, data is collected during each window period t w and the centroid t cm has to be calculated from all these trigger events in some way, which complicates the measurement very much. To make this feasible, according to an embodiment of the present invention, it will rely on a single-photon detection circuit 150. Such a single-photon detection circuit 150 generates a distinct digital pulse every time a photon is detected. To obtain sufficient information, in the example shown, the excitation pulse 100 is reapplied at a sufficiently low repetition rate because it has a period longer than the measurement window period t w . The repetition rate can be freely selected, but in reality, it can depend on many factors such as, but not limited to, the performance of the laser, the maximum light intensity allowed for eye safety, and / or the fact that the phosphor can be temporarily saturated or even cause a bleaching phenomenon that can damage it. Another factor that can act is the fact that background light also gives photons, and / or the detector used has its own dark current (or dark count rate DCR), which further complicates the measurement. In the embodiment shown in FIG. 3, the light pulses are repeatedly generated at a fixed repetition rate. However, in other embodiments not shown, such a fixed repetition rate is not required and the light pulses may be repeatedly generated with variable interruptions. Similarly, the measurement window periods may or may not have the same length.

[0041] In FIG. 3, curve 101 is the pulsed excitation light 100 that is pulsed at a repetition of every 60 ns, and thus at a repetition rate of 16.667 MHz. The trigger curve 120 shows the trigger / photon events output by the single photon detection 150 circuit that is used when a detection element such as an SPAD detects the presence of photons. This includes light arriving from fluorescence, for example, fluorescence emitted by a phosphor excited by the excitation light pulse 100 in curve 101, light from background illumination, and dark counts from the detector itself. In FIG. 3, six such events have occurred, but the source of any of these six events is unknown. Nevertheless, it can be assumed for now that the background and dark counts are very few in number compared to fluorescence. The rejection curve 114 similarly shows three consecutive time windows having a measurement window period t w = 20 ns that are repeated every 60 ns. When the rejection curve signal has a first value, for example, high, the pulse from the trigger curve 120 is rejected, and when it has a second value, for example, low, the pulse is allowed into the measurement process, giving the selected trigger curve 121. This cycle may be repeated, for example, for tens of microseconds or milliseconds, and always passes the pulse only when the rejection curve 114 has the second value, for example, low.

[0042] Here, in FIG. 4, a voltage ramp signal 122 having a slope m is applied at each new measurement window period t w . This voltage ramp signal 122 is sampled at the instant when the selected trigger occurs in the selected trigger curve 121, and samples

Number

Number

Number

Number

[0043] The averaging length n is a number indicating how deep into the past the memory goes. For example, when n = 99, for each new sample, for 99%, the previous results are reused and only for the new 1% are new samples counted. Older samples become less important gradually and more recent samples become more important. The difference from the "simple moving average" is that to implement the latter, a list of the past 100 items has to be kept and every time a new event occurs, the value of the oldest event has to be deleted and the average of the set of 100 events adjusted to include the new event has to be recalculated. This requires a lot of computation and memory and is difficult to implement for each pixel in an image sensor.

[0044] By taking the exponentially weighted moving average of the time - voltage converted samples, the mass centroid method in the time domain is realized.

[0045] In this particular example, the EMA curve 123 has a voltage level three times lower than the lamp voltage on the voltage ramp signal curve 122 at three events 180, 181, and 182, and as a result, small steps increase at each of these instants (n = 7 in this simulation). In this example, one detection event is passed per window, but it is equally possible, for example, under high light level conditions, that there are more events within the same window, and even so, at low light levels, it can happen that there is one, more than one, or many windows without events. There are no operational problems even if there are multiple events within the same window. If there are a very large number of them and they occur in even more windows, information accumulation can occur and the accuracy can be reduced. Whether this becomes a problem depends on many factors, including the dead time, how many events occur within the same window, how frequently it occurs, the application, and the desired accuracy. In such a situation, reducing the amplitude of the excitation light source can be a solution.

[0046] The voltage ramp signal curve 122 may be a perfect straight line, but it does not have to be, and within its window, it should at least increase or decrease monotonically. This can also start at zero volts, but as long as the voltage between the lamps is within the operating range of the switches of the attached switched capacitor circuit 160 (see below), it can also start at any other suitable voltage level. When the EMA curve 123 reaches convergence, the sampled voltage statistically raises and lowers the EMA voltage and does not always go in the same direction as those at the three events 180, 181, and 181.

[0047] FIG. 5 shows a fluorescence lifetime determination circuit 500 according to an embodiment of the present invention. Circuit 500 includes a single photon detection circuit 150 for detecting the generation of photons and generating a corresponding trigger pulse even when the number is very limited, a pulse suppression circuit 166 for rejecting trigger pulses generated outside the measurement time window, a switched capacitor circuit 160 for performing an exponential moving average function, and a non-overlapping switch enable circuit 155 for generating a signal for driving the switched capacitor circuit 160.

[0048] The single photon detection circuit 150 in the illustrated embodiment implements a single photon avalanche diode (SPAD) detector. The SPAD detector is a semiconductor pn diode junction that is reverse biased beyond its breakdown voltage in a metastable state. The SPAD detector is coupled in series with a resistor R1, and this series connection is coupled between a first node 152 and a second node 153. The first node 152 is biased with a first bias voltage Vbias1, and the second node is biased with a second bias voltage Vbias2, where Vbias2 is higher than Vbias1 to obtain a reverse bias for the SPAD detector. A single photon can initiate breakdown when biased through resistor R1 and can provide a large voltage signal, which is a temporary voltage drop (V input ) as seen in FIG. 5. The voltage across the SPAD detector drops until it is below the breakdown voltage of the SPAD detector, thereby stopping the breakdown. Thereafter, the diode voltage returns to the original voltage by the attached resistor R1, and the system returns to the metastable state again. This drop is detected by a suitable detection circuit such as an inverter or comparator circuit (in the illustrated circuit 150 as an example only, the present invention is not limited thereto, but has an inverter X1 with a trip voltage V trip ), and a trigger pulse V p1 can be generated on node p1. The start of the trigger pulse indicates the moment when a photon is incident on the detector. The trigger curves 120 in FIGS. 3 and 4 are six such trigger pulses V p1is shown. There are many types of SPAD detectors, and more generally SPAD circuits, including active ones (e.g., having active or passive quenching) and types of circuits known in the state of the art, each of which can be used in combination with the present invention. The single photon detection circuit 150 can include any SPAD or any single photon detector. What is common to these circuits is that their output is a digital trigger pulse and its onset indicates the instant of photon absorption. The width of the trigger pulse is not very important and can often vary somewhat.

[0049] In this specification, two exemplary systems for suppressing pulses outside the measurement time window are described. The first, namely the pulse suppression circuit 166 in FIG. 5, performs event rejection by lowering the voltage across the SPAD detector. Here, the pulse suppression circuit 166 includes a variable voltage source 164 coupled, for example, to ground between the first node 152 and the third node 154. When the input signal Reject to the pulse suppression circuit 166, and more particularly to its voltage source 164, is high, the output voltage V1 of the voltage source 164 is adapted, for example, to be raised or set to a higher level such that the first bias voltage Vbias1 is high, and thus the voltage across the SPAD is below its breakdown voltage and avalanche breakdown is prevented. When the input signal Reject to the pulse suppression circuit 166, and more particularly to its voltage source 164, goes low, the output voltage V1 of the voltage source 164 is adapted, for example, to be lowered or set to a low level such that the first bias voltage Vbias1 is low, and thus the voltage across the SPAD is increased again and avalanche breakdown can occur again. The first voltage Vbias1 is often not within the voltage range of the digital power supply of the sensor logic, and thus modulating the bias of the SPAD is not easily done in a fast way (with an accuracy of tens of picoseconds). Vbias1 is, for example, -20V to achieve a bias exceeding the breakdown voltage of the SPAD, and Vbias2 is the trip voltage V of the inverter X1 trip1 or 2 V higher than. Therefore, a more advantageous solution operable in the digital domain is proposed and will be described below with reference to FIG. 6.

[0050] The switched capacitor circuit 160 is used to execute an exponential moving average function as also shown in the bottom graph of FIG. 4. The switched capacitor circuit 160 as shown in FIG. 5 includes two switches XN1, XN2 coupled in series between a lamp terminal 171 which is an input terminal and an EMA terminal 172 which is an output terminal. At a sample node 173 between the two switches XN1, XN2, a first capacitor C1 is coupled to ground. At the output terminal 172, a second capacitor C2 is coupled to ground.

[0051] The switches XN1, XN2 of the switched capacitor circuit 160 are operated such that only one of the switches XN1, XN2 conducts at any instant. The drive signals for operating the switches XN1, XN2 are received from a non-overlapping switch enable circuit 155 which will be described in more detail below.

[0052] In the embodiment shown in FIG. 5, the switches XN1, XN2 are implemented by NMOS pass gates. The gates of these NMOS pass gates are coupled to the non-overlapping switch enable circuit 155 and receive the operating signals from the circuit.

[0053] The non-repeating switch enable circuit 155, in certain embodiments of the present invention, includes an equidistant sequence of inverters coupled in series. In this example, there are four inverters X2, X3, X4, X5, each having output nodes p2, p3, p4, p5. Node p2 is the first switch enable output signal for activating the first switch XN1 of the switched capacitor circuit 160. This node p2 is also an input to the NOR gate X6 which, similarly, has p5 as an input. When p1 goes high, p2 goes low and p5, which was already low, goes high after three inverter delays. This leaves a period of three inverter delays during which both inputs of NOR X6 go low and the NOR gate X6 outputs high at output p7. This is basically a one-shot circuit where the input p1 goes high (for example, for 3 ns) and the output p7 goes high for a predetermined short period (for example, 1 ns) determined by three inverter delays. The output signal of the NOR gate X6 is supplied, for its operation, to the second switch XN2, for example, to the gate of an NMOS pass gate. Many alternative one-shot circuits are possible and are known to those skilled in the art.

[0054] As shown in one but the last graph of FIG. 4, in the execution of the exponentially weighted moving average function, the ramp voltage 122 is used. The externally generated ramp voltage is applied at the Ramp terminal 171 and is connected to the source of the first switch XN1, made by an NMOS pass gate in the illustrated embodiment. The signal V at the drain of the first switch XN1, i.e., at the sample node 173 s awaits an event in accordance with the voltage ramp signal 122 (FIG. 4) as long as the gate of the first switch XN1 is biased high.

[0055] Figure 7 shows this process. The signal corresponding to the passed trigger pulse, for example, the signal on the output node p2 of the inverter X2, is represented by curve 125, and the signal at the output node p7 of the NOR gate X6 is represented by curve 126. Near time = 93 ns, the voltage on node p2 drops, and the first switch XN1 stops conducting. The voltage present at the sample node 173 at that instant is a sample of the lamp voltage applied to the lamp terminal 171 at that instant

Number

Number

Number

Number

[0056] If an output with suitable accuracy is desired, then n should be taken high, i.e., it should be between 100 and 1000, which increases the number of samples required for convergence after initialization of the filter or after a sharp change in its lifetime. Since the time required for the change depends entirely on the number of incoming photons / events and the value of n, it is not known whether this will require a measurement time in microseconds or milliseconds. It is also an option to keep the number n relatively low, e.g., 50, sample the EMA output from outside the pixel, perform additional external calculations and averaging, thereby increasing the accuracy and precision in a second stage, e.g., using DSP digital processing means.

[0057] FIG. 8 shows the operation in which a simulation is performed for 200 samples passing through a measurement time window having a time distribution generated by a random generator taking into account an exponential decay behavior with n = 50 and a lifetime of 2.6 ns. The measurement window period t w is set to 10 ns and the repetition rate is set to 10 MHz. The bar graph 141 shows the histogram of 200 recorded events over a 10 ns window period. The curve 140 is the EMA signal starting with a starting condition of 0 ns (in nanoseconds in this simulation). After 75 samples, a steady state occurs with an average of about 2.4 ns and an accuracy of 7% for the lifetime (for samples from 100 to 200). Using the curve 119 in FIG. 2, the fact that the lifetime is about 1 / 4 of the (10 ns) sampling window can be corrected, bringing the result closer to the original 2.6 ns lifetime used for lifetime distribution sample generation. The accuracy and precision are adapted according to normal statistical rules. For example, increasing n by a factor of 4 gives a factor of 2 improvement in accuracy.

[0058] In practice, the associated circuitry disposed in the pixel should be made as small as possible so that the pixel has the highest possible fill factor. Therefore, it may be recommended to completely omit the first capacitor C1 and rely on the remaining parasitic capacitance of the associated sample node 173. After obtaining an estimated value of the parasitic value C1, C2 may be selected to be n times that value. C2 may then be implemented by one skilled in the art in a manner that best fits the imager pixel.

[0059] The more samples that are averaged, the higher the accuracy. Therefore, it is advantageous to use a large averaging length n. However, even with a small first capacitor C1, the second capacitor C2 will occupy some area of the pixel, and increasing the second capacitor C2 will typically require a larger area in the pixel. In embodiments of the present invention, the second capacitor C2 may be made in the same manner as the capacitor in a dynamic RAM cell, having deep holes in the substrate and providing a large capacitor with a small area.

[0060] In an embodiment of the present invention, two switched capacitor circuits operating with different clocks may be implemented. An example thereof is shown in FIG. 15. The first switched capacitor circuit 160 is operated for each passed event that provides a voltage named EMA on the second capacitor C2 (similar to all the above-described embodiments). On the other hand, the second subsequent switched capacitor circuit 800 is operated at an oscillation rate that may be fixed or settable, rather than in response to incident photons. The structure of the second switched capacitor circuit 800 is similar to the structure of the first switched capacitor circuit and includes two switches XN5 and XN6 coupled in series between the input terminal 172 and the output terminal hybridEMA. At the sample node Vs2 between the two switches XN5 and XN6, a third capacitor C5 functioning as a sampling capacitor is coupled to ground. At the output terminal hybridEMA, a fourth capacitor C6 is coupled to ground. The third switch XN5 and the fourth switch XN6 enable the third capacitor C5 to be alternately coupled to the node 172 for receiving the calculated average voltage as a measure of the fluorescence lifetime and the output node hybridEMA. The fourth capacitor C6 is configured to have a charge sharing redistribution configuration when the fourth switch XN6 is driven to couple the third capacitor C5 to the output node (hybridEMA). In an embodiment of the present invention, the third capacitor C5 may be at least one order of magnitude smaller, preferably at least two orders of magnitude smaller, than the fourth capacitor C6.

[0061] The output signal hybridEMA, which is the voltage on the fourth capacitor C6, is then the output for reading. To generate non-overlapping clock signals for the input CLK1 and CLK2, for example, the frequency f clkA normal oscillator having [the relevant property] can be used. Non-overlapping means that switches XN5 and XN6 do not conduct at the same instant. CLK1 drives switch XN5 and CLK2 drives switch XN6. XN5 and XN6 are, in this example, NMOS pass gates. The voltage on the third capacitor C5 is a sample of the EMA voltage at the instant when CLK1 goes low and switch XN5 stops conducting. CLK2 then goes high and charge sharing occurs between the third capacitor C5 and the fourth capacitor C6. This then gives a classical switched capacitor low-pass filter 800 with a -3 dB corner frequency as follows. [Number]

[0062] The averaging length n1 = C2 / C1 in the first stage and n2 = C6 / C5 in the second stage give an overall averaging length of n1.n2 = (C2.C6) / (C1.C5). Thus, if n1 = n2 = 100, the maximum averaging reaches 10000. Averaging of 10000 samples theoretically gives a relative fluorescence lifetime accuracy of 1%.

[0063] Figure 16 shows a simulation based on a frame rate of 50 fps (20 ms per frame) with n1 = n2 = 100, simulated for 50 frames equal to 1 second. The width of the window period is 20 ns and the repetition frequency is 50 MHz. Thus, the number of cycles per frame is 10 6 is. In this simulation, the clock frequency is f clkIt is selected to be 5 kHz. The lifetime of the photons is selected to be 2 ns and abruptly step - decreases to 1 ns after 20 frames. Curve 900 is the estimated lifetime realized when, on average, photons are received only once in 10,000 cycles. In the case of 1 million cycles in one frame, only 100 photons per frame are available. At frame 20, the lifetime decreases by 1 ns at a time. The response of the estimated lifetime takes some time and decreases to 1 ns in 5 - 10 frames. Within the period of a constant lifetime, an accuracy of 4.7% is achieved over the frames (shown by curve 905 at X = 100). 10 times the photons per cycle, i.e., on average 1000 photons per frame, gives curve 901, with a response that drops in a faster step - like manner in 3 - 5 frames and a more favorable accuracy of 2.7%. 10,000 photons per frame leads to curve 902 with an even more favorable accuracy of 0.5%. 100,000 and 1,000,000 photons per frame have a similar step - like response (curves not shown) and the same low level of accuracy of 0.5%. This setup shows that, thanks to the two - step approach, a dynamic range of at least 4 digits of optical input is supported. Since there is no register overflow and no need to change parameters, this enables operating an imager based on this principle simultaneously at this dynamic range of optical levels at various locations within one view.

[0064] The hybrid EMA averaging principle can be used in any averaging circuit of the present invention.

[0065] The circuit according to an embodiment of the present invention can deviate significantly from what is presented. However, special attention has been paid to keep the transistor count low. Also, since most of the operations are in the digital domain, there are only very limited analog difficulties. When assuming a 1% lifetime accuracy, it is sufficient to have a low-precision ADC (e.g., 8 bits) for the conversion of the EMA voltage. The width and length of the transistors (W / L value) are not important and can be determined by those skilled in the art. Most of the transistors can be of the minimum width and length, but it is necessary to ensure that curves 125 and 126 (Figure 7) do not overlap and preferably reach the full supply voltage, that is, to ensure that the switches of the switched capacitor circuit 160 operate optimally or at least are sufficiently suitable.

[0066] As described above, a solution for a pulse suppression circuit that is more advantageous than what was described with reference to Figure 5 is operable in the digital domain. Figure 6 shows a fluorescence lifetime determination circuit 600 according to an embodiment of the present invention, including a pulse suppression circuit 165 in the digital domain. In addition to this pulse suppression circuit 165, the circuit 600 includes a single photon detection circuit 150, a non-overlapping switch enable circuit 155, and a switched capacitor circuit 160. These latter components are as disclosed with respect to Figure 5 and will not be described in detail here again.

[0067] In the embodiment of FIG. 6, one of the signals driving the switched capacitor circuit 160 is blocked, and in this way, it is possible to prevent the pulses from being taken into account for the execution of the averaging. This is shown in FIG. 6, where the pulse suppression circuit 165 deletes the unwanted pulses on the output node p6 of the NOR gate X6 outside the measurement time window in response to the gate signal Reject, and sends the cleaned-up version of the signal on the node p6 to the node p7 connected to the pass gate XN2 of the switched capacitor circuit 160. In FIG. 4, the signal on the node p6 is given by the curve 120, and the cleaned-up version, i.e., the version having pulses outside the measurement time window, is given by the curve 121. The suppression circuit 165 as depicted in FIG. 6 can also be made in a different way, for example, by including an AND gate (not shown).

[0068] One of the wonderful features of the particular solution presented in FIG. 6 is that, as shown in the detector 900 of FIG. 9, with only five additional transistors and two capacitors, a second EMA2 switched capacitor circuit 162 having its own time window and the same or different voltage ramp signal Ramp2, and a second pulse suppression circuit 167 can be fabricated, and these can operate completely independently (but simultaneously) from the first EMA1 switched capacitor circuit 160 and the pulse suppression circuit 165. In an image sensor, in a low-cost monolithic solution, since the area occupied by the circuit is in a trade-off relationship with the area for the actual light detection region, the area of the additional circuit needs to be kept low. Due to the small required area, the number of centroid measurements operating in parallel can be increased, and at this time, the sampled photons can be permitted or prohibited depending on the dedicated applied window.

[0069] FIG. 10 shows a very useful application for performing the measurement of the two centroids of the input light. When there is background light uncorrelated with the excitation light 100, the following behavior of the received light is obtained.

Number

Number

[0070] In an embodiment of the present invention, the single-photon detection circuit 150 may have an elevated dark count rate (DCR) that has the same effect as background light. Thus, the DCR can be dealt with in the same way using the double mass center approach, as described with respect to FIG. 10. Measuring multiple mass centers can be achieved in parallel using additional circuitry mounted on each pixel in the image sensor, just as described above, or can be achieved successively one after another using the same circuitry, assuming that the conditions in the scene do not change. Also, combinations are conceivable, for example, if the hardware is only provided in parallel for two, first measuring two mass centers in parallel and then measuring another two mass centers.

[0071] FIG. 11 shows another use of the double mass center approach. In lifetime measurements having a short lifetime on the order of nanoseconds, each signal transmission in the setup has a significant delay, whether it is an electrical signal traveling at 60 - 70% of the speed of light or the light itself. To achieve the best accuracy and precision, it is always preferable to place a measurement time window that starts at t0 and does not overly elapse from the maximum instant t of fluorescence 107. Therefore, it is useful to first determine the delay t between the excitation pulse and the actual emission of fluorescence, which is a function of all the involved transmission paths, the optical path, and the overall synchronization principle in the implementation system. A further measurement time window is placed in place, having a start instant t3, which is conveniently selected such that t3 is surely before the start time t of the emission of fluorescence, and having a measurement window period t. In the example shown in FIG. 11, the further measurement time window has an end time t1 and is similar to the other measurement time windows used having a start time t0 and an end time t1. The measured mass center t of the further measurement time window will refer to the instant t beyond the start time t of the emission of fluorescence, and then an estimate can be made to roughly locate t. s from which an overly long time does not elapse. Thus, it is useful to first determine the delay t between the excitation pulse and the actual emission of fluorescence, which is a function of all the involved transmission paths, the optical path, and the overall synchronization principle in the implementation system. A further measurement time window is placed in place, having a start instant t3, which is conveniently selected such that t3 is surely before the start time t of the emission of fluorescence, and having a measurement window period t. In the example shown in FIG. 11, the further measurement time window has an end time t1 and is similar to the other measurement time windows used having a start time t0 and an end time t1. The measured mass center t of the further measurement time window will refer to the instant t beyond the start time t of the emission of fluorescence, and then an estimate can be made to roughly locate t. d between the excitation pulse and the actual emission of fluorescence, which is a function of all the involved transmission paths, the optical path, and the overall synchronization principle in the implementation system. A further measurement time window is placed in place, having a start instant t3, which is conveniently selected such that t3 is surely before the start time t of the emission of fluorescence, and having a measurement window period t. In the example shown in FIG. 11, the further measurement time window has an end time t1 and is similar to the other measurement time windows used having a start time t0 and an end time t1. The measured mass center t of the further measurement time window will refer to the instant t beyond the start time t of the emission of fluorescence, and then an estimate can be made to roughly locate t. s before the start time t of the emission of fluorescence, and having a measurement window period t. In the example shown in FIG. 11, the further measurement time window has an end time t1 and is similar to the other measurement time windows used having a start time t0 and an end time t1. The measured mass center t of the further measurement time window will refer to the instant t beyond the start time t of the emission of fluorescence, and then an estimate can be made to roughly locate t. w3 In the example shown in FIG. 11, the further measurement time window has an end time t1 and is similar to the other measurement time windows used having a start time t0 and an end time t1. The measured mass center t of the further measurement time window will refer to the instant t beyond the start time t of the emission of fluorescence, and then an estimate can be made to roughly locate t. cm3 of the further measurement time window will refer to the instant t beyond the start time t of the emission of fluorescence, and then an estimate can be made to roughly locate t. s of the emission of fluorescence, and then an estimate can be made to roughly locate t. c3 beyond the start time t of the emission of fluorescence, and then an estimate can be made to roughly locate t. s and then an estimate can be made to roughly locate t.

[0072] FIG. 12 shows that the fluorescence lifetime determination circuit 1200 according to an embodiment of the present invention can conveniently include an analog photon counter circuit 168 that provides information about the number of events passing through a measurement window period from the moment a reset pulse is applied on the ResetCount input.

[0073] The photon counter circuit 168 illustrated in FIG. 12 is only one embodiment of possible implementations. This includes, for example, a series connection of pass gate switches X12 and X22, which are NMOS pass gate switches, and has a counter sampling node 175 between both pass gate switches X12 and X22. A sampling capacitor C12 is coupled between ground and the counter sampling node 175. The source side of the first pass gate switch X12 is coupled to a fixed voltage V lower than Vdd (not shown). SX12 The drain side of the second pass gate switch X22 is coupled to the output node Countout. A further capacitor C22 is coupled between ground and the drain of the second pass gate switch X22. The first and second pass gate switches X12 and X22 are actuated by the same signal that also actuates the pass gates XN1 and XN2 in the switched capacitor circuit 160. For example, these signals are supplied to the gates of the pass gate switches X12 and X22.

[0074] When no trigger pulse occurs, the signal at the gate of the pass gate X12 is the signal at node p2, and this signal is high. At this time, the voltage level V at the drain side of the pass gate X12 sc is equal to V SX12 When a trigger pulse occurs, this voltage at the counter sampling node 175 is sampled.

[0075] The signal at node p7 does not overlap with the signal at node p2 and here it goes high, turning on the second pass gate switch X22. Capacitors C12 and C22 are shorted to each other, causing charge sharing and redistribution, whereby the signals at counter sampling node 175 and output node Countout vary towards each other by a value determined by the ratio C22 / C12.

[0076] In this way, the voltage at output node Countout rises in steps each time such an event occurs, but in a saturating manner, with each step of Countout becoming slightly smaller and eventually reaching a saturation voltage close to V SX12 The counter number at saturation depends on the selected capacitor ratio C22 / C12. C12 can be made entirely from its parasitic nodes and keep the circuit small. It is possible to implement counter values from 10 to several hundreds without the need for an overly large capacitor C22.

[0077] Reset switch 700 is coupled between output node Countout and ground to reset output node Countout. Reset switch 700 may be, for example, a transistor, whose gate is coupled to a signal ResetCount applied externally.

[0078] Sometimes, for example, immediately after startup, or when different lifetimes suddenly exist, it is known that the EMA average voltage has not yet converged. In that case, it is possible to accelerate the convergence process by temporarily reducing the number n of the switched capacitor circuit 160. In fact, by changing the value of the capacitor C1, the averaging length n can be modulated in a convenient way. This can be done by externally adding a capacitor to the sample node 173. However, the averaging length n may also be linked to the number of photons passing through the window using the photon counter 168. In FIG. 13, this is shown in such a way that the circuit 169 monitors whether a predetermined number of photons have been counted after the reset pulse on the input ResetCount. Before that, an external capacitor C3 is added to the first capacitor C1 of the switched capacitor circuit 160 by the switch X9. The averaging length n is short, the EMA moves quickly, and gives a rough and quick estimate of the center of mass. Then, since the actuation signal to the gate (node p8) of the switch X9 becomes low, the external capacitor C3 is disconnected, and the averaging length n is increased, starting from a rough and approximate initial estimate and leading to a more accurate measurement. In this way, the curve 140 in FIG. 8 has a rather steep slope and can reach the rough estimate in 10 steps instead of requiring 75 samples to reach it, for example. The circuit in FIG. 13 operates digitally, but of course, more advanced solutions leading to a more flexible increase in the averaging length n may be made by those skilled in the art. For example, the variation of the channel capacitance of the transistor may be utilized, or a dedicated varactor may be used.

[0079] Sometimes, there may be a combination of phosphors that output two lifetimes upon stimulation.

Number

[0080] In this case, there are three unknowns, namely the first lifetime τ1, the second lifetime τ2, and the ratio of the fluorescence intensities emitted at the first instant t0

Number

[0081] In this case, it is necessary to determine three centers of mass. To solve this, it is necessary to strategically select three measurement time windows. As an example, FIG. 14 shows these three measurement time windows, namely the first measurement time window t w1 , the second measurement time window t w4 that is temporally later, and the third measurement time window t w5 with a length that combines the first and second measurement time windows.

[0082] A short lifetime mainly affects the center of mass in the first measurement time window t w1 . A long lifetime mainly affects the center of mass of the second measurement time window t w4 , and the center of mass at t w5 will be greatly affected by each of the three unknowns. In any case, one of ordinary skill in the art can solve three equations based on the three measured centers of mass and determine the three unknowns in the most preferred way. Adding the possibility of having non - negligible background light or DCR (of the single - photon detection circuit 150) will result in adding another unknown variable that can be obtained by additional center - of - mass measurement time windows (not shown) and their mathematical equations (having four unknowns and four equations).

[0083] Any of the systems presented herein, or systems based thereon, may be supplemented by other means known in the state of the art of image sensors. For example, microlenses, color filters may be applied to qualitatively or quantitatively improve the input of light to the single-photon detection circuit. Any means for improving the internal / external quantum efficiency, responsiveness, and detection probability may be applied. A three-dimensional stack may be performed, for example, where the SPAD detection layer may be derived from a wafer / material different from the CMOS circuit wafer. Backside illumination (BSI) may be applied, or current assist may be applied, or silicon-on-insulator (SOI) may be suitable. The proposed embodiments of the present invention can be laid out as pixels for a sensor array, and as a whole constitute an image sensor for acquiring fluorescence lifetime and fluorescence intensity. For each read voltage node, the voltage follower transistor may include a row selection transistor for row-based readout, as is often done in CMOS image sensors. Some signals may be grouped for each of the plurality of pixels, or may be the same for the entire array, such as those defining a window (Reject), a ramp signal, a reset signal, and a signal for determining the averaging length n. In addition to this, standard 3T or 4T image sensor pixels may be added to simultaneously perform standard image sensing. The single-photon detection circuit 150 may include a normal SPAD, but may include any other means for achieving single-photon detection, including an avalanche photodetector (APD) that uses a linear gain technique to operate the diode below breakdown and has a large enough gain to realize digital photon arrival pulses.

[0084] The present invention has been illustrated and described in detail in the drawings and the foregoing description, but such illustrations and descriptions should be regarded as illustrative or explanatory and not restrictive. The foregoing description details specific embodiments of the present invention. However, it will be understood that the present invention may be practiced in many ways, however detailed the foregoing may appear in the text. The present invention is not limited to the disclosed embodiments.

Claims

1. A detection system for detecting fluorescence lifetime, the detection system comprising: an excitation light source configured to repeatedly generate pulsed excitation light; detectors (500, 600, 900, 1200, 1500), a single-photon detection circuit (150) for generating a digital pulse (120) upon detection of a photon; A pulse suppression circuit (165, 166) for rejecting detected photons that occur outside each one of a series of measurement time windows, each subsequent measurement time window beginning after a subsequent excitation light pulse (101) has ceased and ending before the next excitation light pulse (101) is generated, each measurement time window having a measurement window period (t w ), the pulse suppression circuit (165, 166); Each new measurement window period (t w ) having an input terminal (171) for receiving a voltage ramp signal that resumes in the first switched capacitor circuit (160), wherein the first switched capacitor circuit (160) is recorded, detected and not rejected over a past measurement time window according to the principle of the center of mass method. A first switched capacitor circuit (160) having a sample voltage (V s i ) configured to repeatedly calculate an average voltage based on an exponentially weighted moving average function applied to the first switched capacitor circuit, which outputs the calculated average voltage as a measure of the fluorescence lifetime. A first switched capacitor circuit (160) having a first node (172) for the detector (500, 600, 900, 1200, 1500), and a detection system.

2. The detection system according to claim 1, wherein the first switched capacitor circuit (160) includes a first sampling capacitor (C1), and a first switch (XN1) and a second switch (XN2) for alternately coupling the first sampling capacitor to a ramp terminal (171) for receiving a ramp signal and the first node (172).

3. The detection system according to claim 2, wherein the first switched capacitor circuit (160) further includes a second capacitor (C2) configured to be in a charge sharing and redistribution configuration when the second switch (XN2) is driven to couple the first sampling capacitor (C1) to the first node (172).

4. The detection system according to claim 3, wherein the first sampling capacitor (C1) is at least one order of magnitude smaller than the second capacitor (C2).

5. The detection system according to any one of claims 2 to 4, further comprising means for temporarily adding one or more capacitors (C3) in parallel with the first sampling capacitor (C1).

6. The detection system according to any one of claims 1 to 4, further comprising a second switched capacitor circuit (800) connected in series to the first node (172) of the first switched capacitor circuit (160) and configured to operate at an oscillation rate that does not directly respond to incident photons.

7. The detection system according to claim 6, wherein the second switched capacitor circuit (800) includes a second sampling capacitor (C5), and a first switch (XN5) and a second switch (XN6) for alternately coupling the second sampling capacitor to the first node (172) for receiving an average voltage calculated as a measure of the fluorescence lifetime and an output node (hybridEMA).

8. The detection system according to claim 7, wherein the second switched capacitor circuit (800) further includes a fourth capacitor (C6) configured to have a charge sharing and redistribution configuration when the second switch (XN6) is driven to couple the second sampling capacitor (C5) to the output node (hybridEMA).

9. The detection system according to claim 8, wherein the second sampling capacitor (C5) is at least one order of magnitude smaller than the fourth capacitor (C6).

10. The detection system according to any one of claims 1 to 4, further comprising a non-overlapping switch enable circuit (155) for providing a non-overlapping signal for operating the first switched capacitor circuit (160).

11. The detection system according to any one of claims 1 to 4, wherein the single photon detection circuit (150) includes a single photon avalanche detector.

12. The detection system according to claim 11, wherein the pulse suppression circuit (166) includes a variable voltage source (164) adapted to lower the voltage across the single photon avalanche detector.

13. The detection system according to any one of claims 1 to 4, wherein the pulse suppression circuit (165) is adapted to block one of the signals for driving the switched capacitor circuit (160), thus preventing pulses from being taken into account.

14. The detection system according to claim 13, further comprising at least one additional pulse suppression circuit (167) and at least one additional switched capacitor circuit (162) configured to operate in parallel with the pulse suppression circuit (165) and the first switched capacitor circuit (160).

15. The detection system according to any one of claims 1 to 4, further comprising a photon counter circuit (168) for counting the number of detected photons.

16. A non-overlapping switch enable circuit (155) for providing a non-overlapping signal for operating the first switched capacitor circuit (160) and a photon counter circuit (168) for counting the number of detected photons, wherein the photon counter circuit (168) includes a switched capacitor circuit adapted to be operated by the non-overlapping signal for operating the first switched capacitor circuit (160). The detection system according to any one of claims 1 to 4.

17. Further comprising a photon counter circuit (168) for counting the number of detected photons, wherein the means for temporarily adding one or more capacitors (C3) in parallel with the sampling capacitor (C1) each includes one or more switches in series with the one or more capacitors (C3), and the one or more switches are opened when the photon counter circuit (168) has counted a predetermined number of detected photons. The detection system according to claim 5.

18. A fluorescence imaging sensor, An excitation light source configured to repeatedly generate pulsed excitation light, An array of detectors (500, 600, 900, 1200, 1500), each detector A single photon detection circuit (150) for generating a digital pulse (120) upon detection of a photon, A pulse suppression circuit (165, 166) for rejecting detected photons that occur outside each one of a series of measurement time windows, each subsequent measurement time window beginning after a subsequent excitation light pulse (101) has ceased and ending before the next excitation light pulse (101) is generated, each measurement time window having a measurement window period (t w ), the pulse suppression circuit (165, 166); Each new measurement window period (t w ), a first switched capacitor circuit (160) having an input terminal (171) for receiving a voltage ramp signal that resumes in), wherein the first switched capacitor circuit (160) is recorded, detected and not rejected over a past measurement time window according to the principle of the center of mass method. A sample voltage (V determined by a voltage ramp signal in response to photons (121)) s i ), configured to repeatedly calculate an average voltage based on an exponentially weighted moving average function applied to), and the first switched capacitor circuit has a first node (172) for outputting the calculated average voltage as a measure of fluorescence lifetime. A first switched capacitor circuit (160), and an array of detectors (500, 600, 900, 1200, 1500). A fluorescence imaging sensor comprising.

19. A method for determining fluorescence lifetime, Repeatedly generating pulsed excitation light, Generating a digital pulse upon detection of a photon, Rejecting detected photons that occur outside each of a series of measurement time windows, wherein each subsequent measurement time window begins after a subsequent excitation light pulse (101) has stopped and stops before the next excitation light pulse (101) is generated, and each measurement time window has a measurement window period (tw). Rejecting, Using a switched capacitor circuit to receive a voltage ramp signal restarted in each new measurement window period (tw), and based on an exponentially weighted moving average function applied to the sample voltage determined by the voltage ramp signal in response to photons recorded, detected, and not rejected over past measurement time windows according to the principle of the centroid method, repeatedly calculating an average voltage; Outputting the calculated average voltage as a measure of the fluorescence lifetime. A method comprising:

20. The method according to claim 19, further comprising averaging the calculated average voltage.

Images