Quantitative particle identification digital autoradiography

Abstract

Methods and apparatus are disclosed for concurrent digital autoradiography of alpha-particle events and positron events using a spatially resolving charged particle detector and a gamma-ray detector. Alpha particles and positrons are detected by the charged particle detector. Positrons are identified based on coincidence with gamma-ray events. A positron-emission autoradiograph is formed based on positron positions. Alpha particles, both coincident and anticoincident with positrons, are identified based on energy deposition patterns. An alpha-emission autoradiograph is formed based on alpha-particle positions and energies. Separation of coincident alpha and positron events recorded by the charged particle detector improves position detection accuracy. Validation of positron imaging for alpha-emitter distribution or dosage, based on correlation of respective autoradiographs, is described. Variations for particle identification based on isotropy, energy, coincidence, or anticoincidence are presented. Disclosed techniques are applicable to alpha-emitting radionuclides with complex decay chains.

Description

QUANTITATIVE PARTICLE IDENTIFICATION DIGITAL AUTORADIOGRAPHYCROSS REFERENCE TO RELATED APPLICATIONS

[0001] This application claims benefit of U.S. Provisional Patent Application No. 63 / 651,184, filed May 23, 2024, which is incorporated in its entirety herein by reference.ACKNOWLEDGMENT OF GOVERNMENT SUPPORT

[0002] This invention was made with government support under 75N91019D00024 awarded by the National Institutes of Health. The government has certain rights in the invention.BACKGROUND

[0003] Theragnostic radiation therapies incorporate two radioisotopes, one for treatment and one for imaging. Alpha particle radiation is often effective for cancer treatment, however the very short range of alpha particles in tissue renders direct imaging of the alpha particles difficult. Positron emission tomography (PET) is well established, and the combination of alpha and positron emitters for theragnostic treatment has been recognized. However, the accuracy with which positron imaging reflects the distribution of alpha-emitting nuclei or the resultant energy dose distribution in tissue remains an open question. Conventional approaches to comparing alpha-emitter distribution with positron emission images can rely on separate experiments to determine alpha-emitter and positron-emitter distributions in different samples. Controlling for variation between samples and between experimental configurations can be difficult. Accordingly, there remains a need for improved techniques to generate or correlate alpha and positron images.SUMMARY

[0004] In brief, examples of the disclosed technologies perform concurrent acquisition of alpha-emission and positron-emission events, using a common sample and a common spatially resolving charged particle detector to detect alpha particles and positrons emitted from the sample. A second gamma-ray detector can be used to detect positron annihilation events. Positrons can be positively identified based on coincidence with gamma-ray events from the second detector, and an autoradiograph of a positron-emitting radionuclide can be formed based on positions of the detected positron events. Alpha particles, both those coincident and anticoincident with positrons, can be identified based on energy deposition patterns. An autoradiograph of an alpha-emitting radionuclide can be formed based on positions of the detected alpha-particle events.

[0005] In some examples, coincident positrons and alpha particles in the charged particle detector can be separated and distinguished. Separation of coincident alpha and positronevents recorded by the charged particle detector improves position detection accuracy of both particles.

[0006] In further examples, the disclosed technologies can be applied to a planar biological sample infused with both an alpha-emitting radionuclide and a positron-emitting radionuclide. Respective autoradiographs of the two radionuclides can be correlated to validate (or determine the accuracy of) positron imaging as an indicator of alpha-emitter or a dosage distribution.

[0007] The foregoing and other objects, features, and advantages of the invention will become more apparent from the following detailed description, which proceeds with reference to the accompanying figures.BRIEF DESCRIPTION OF THE DRAWINGS

[0008] FIG. 1 is a flowchart of a first example method according to the disclosed technologies.

[0009] FIG. 2 is a hybrid diagram of an example apparatus according to the disclosed technologies.

[0010] FIG. 3 is a diagram illustrating use of a spatially resolving charged particle detector according to the disclosed technologies.

[0011] FIG. 4 is a diagram illustrating use of a gamma-ray detector according to the disclosed technologies.

[0012] FIG. 5 is a flowchart of a second example method according to the disclosed technologies.

[0013] FIG. 6 is a flowchart of an example method for identifying particles according to the disclosed technologies.

[0014] FIG. 7 is a flowchart of an example method for determining isotropy of an energy distribution pattern according to the disclosed technologies.

[0015] FIGS. 8A-8B are images illustrating exemplary spatial signatures of beta particles and alpha particles respectively, as can be found with the disclosed technologies.

[0016] FIG. 9 is a flowchart of a third example method according to the disclosed technologies.

[0017] FIG. 10A - 10B illustrate energy spectra of alpha particles arising from radioactive decay of an alpha-emitting radionuclide which can be used in some examples of the disclosed technologies.

[0018] FIG. 11 A - 1 ID illustrate energy spectra of beta particles and gamma rays arising from radioactive decay of a positron-emitting radionuclide which can be used in some examples of the disclosed technologies.

[0019] FIG. 12 is a diagram illustrating coincidence processing of events from a charged particle detector and a gamma-ray detector according to the disclosed technologies.

[0020] FIG. 13 is a set of example autoradiographs illustrating application of the disclosed technologies.

[0021] FIG. 14 is a set of images illustrating a first example application of the disclosed technologies to an animal sample.

[0022] FIG. 15 is a set of images illustrating a second example application of the disclosed technologies to an animal sample.

[0023] FIG. 16 is a chart showing correlation between positron emitter density and alpha dosage density for the experiment of FIG. 15.

[0024] FIGS. 17B-17D are images illustrating an example of spectrally resolved imaging of a beta source (whose emission energy spectrum is shown in FIG. 17A), according to the disclosed technologies.

[0025] FIG. 18 is a diagram schematically depicting a computing environment suitable for implementation of disclosed technologies.DETAILED DESCRIPTION Introduction

[0026] Targeted alpha-particle therapy is increasingly used for cancer treatment. An alphaemitting radionuclide can be selectively delivered in a desired vicinity, but can migrate due to biological processes. In order to monitor therapy, it is desirable to measure delivered dosage distribution. However, common alpha particles have short range in solids or liquids (typically 10 - 100 pm) and can be difficult to detect in situ in practical applications.

[0027] Positrons are easier to image, even deep within a subject, through detection of gamma rays produced in electron-positron annihilation. Annihilation events typically occur within 1-3 mm of a host positron-emitting nucleus. An alternative is to image gamma rays emitted alongside alpha decay, as proxies for the alpha particles themselves. However, attempts with single-photon emission computed tomography (SPECT) cameras have not been able to provide spatial accuracy of dose distributions. The spatial resolution achievable with positron imaging is better suited to dose distribution studies.

[0028] Accordingly, a theragnostic treatment can combine an alpha-emitting radionuclide with a positron-emitting radionuclide, both infused into a same tissue volume. The alphaemitter can deliver localized radiation dosage, while the positron-emitter can be imaged.

[0029] However, a number of effects can lead to variations between the distribution of the two radionuclides, and there is a potential that a positron image may not accurately reflect the alpha dosage distribution. It is desirable to validate the positron image as a proxy for the alpha dosage distribution.

[0030] In one approach, a first sample can be infused with the alpha emitter and a second sample can be infused with the positron emitter. The samples can be imaged on respective analyzers designed specifically for alpha-particle and positron imaging. However, differences between samples, sample preparation, and analyzers can introduce additional variations between the resulting autoradiographs.

[0031] Examples of the disclosed technologies overcome these issues by concurrently acquiring alpha-particle events and positron events in a single spatially resolving charged particle detector from a single planar sample. Differences between samples and sample preparation are eliminated, and analyzer differences can be minimized.

[0032] Accurate determination of position is desirable for both positrons and alpha particles. Disclosed detectors can provide positional accuracy commensurate with pixel size, in the range 20 - 50 pm. Positional accuracy can be improved by separating coincident events reported by the charged particle detector and independently analyzing each region in which charge is deposited.

[0033] Accurate identification of both alpha particles and positrons is desirable. Disclosed examples apply various characteristics to distinguish alpha particles from other energy depositions, notably isotropy or energy. In some examples, alpha particles and positrons may be accompanied by beta particles, either from other radioactive decay channels or as energy secondary electrons ejected by passage of alpha particles through matter. For positive identification of positrons, disclosed examples utilize coincidence between a candidate beta particle and gamma ray events detected by a second detector incorporating, e.g., a scintillator. However, gamma ray events can also arise from multiple sources: some from positron annihilation events, and others from other decay channels. To distinguish annihilation gamma rays from other gamma rays, some examples apply an energy filter to gamma ray events.

[0034] Accurate determination of timing is desirable for coincidence detection with tight coincidence windows for high signal-to-noise ratio. Modern electronics packages can provide timestamp accuracy in the 0.5 - 10 ns range (often 1 - 2 ns) for both the charged particledetector and the gamma ray detector. Disclosed examples implement stream processing of event timestamp data to detect coincidences and identify positron events.

[0035] The distributions of identified alpha particles and positrons can be used to form respective autoradiographs. Correlation between the two autoradiographs can validate the use of positron imaging as a proxy for alpha dosage. In examples, a calibration factor can be developed to convert positron image density to alpha dose density, on an areal or volumetric basis.Terminology

[0036] To facilitate review of the various embodiments, some terms are explained below as used in this disclosure. Occasionally, and where clear from the context, a term may also be used in a different meaning.

[0037] An “array” of physical objects (such as sensors, or pixels of a semiconductor sensor) or data objects (such as pixels of an image) is a collection of the objects that have a defined spatial relationship. In some examples, the spatial relationships are regular, meaning that at least two pairs of the objects have substantially similar relative spacing and / or relative orientation. Pixel arrays can be rectangular, with pixels arranged in rows and columns. A rectangular pixel array can have length and width which can be denominated in spatial units (e.g. 2 cm x 2 cm) or in units of pixels (e.g. a 256 x 256 pixel array). In other examples, an array can have defined but irregular spatial relationships.

[0038] “Calibration” refers to a procedure or data for converting a raw measurement into a desired physical unit. In examples, a calibration can be applied to charge collected at one pixel of a charged particle detector to determine an amount of energy deposited by a radioactive decay particle at that pixel; a calibration can be applied to the collective charge or energy measured at multiple pixels to determine an initial energy of a corresponding radioactive decay particle; or a calibration can be applied to a photomultiplier output signal to determine an amount of gamma-ray energy deposited in a scintillator to which the photomultiplier is coupled. A calibration can use a single multiplicative factor or a nonlinear function for the conversion. A nonlinear function can be implemented as a table lookup with interpolation, as a piecewise linear function having multiple linear segments, or as a formula to be evaluated.

[0039] A “clock signal” (or simply “clock”) is a periodic oscillatory signal having a fixed frequency. Counting periods of the clock signal can allow time to be measured, e.g. for timetagging event data with a digital timestamp. A “clock source” is a circuit operable to generate a clock signal. A clock signal having period T can be represented as a continuouslyincreasing phase cp(t), where t represents time, such that cp(t+T) - cp(t) = 2TI. TWO clock signals are “synchronized” if their respective phases cpl(t), cp2(t) have a linear relationship, e.g. cpl = A- cp2+B, where A and B are constants. The fixed frequency of a clock signal does not preclude it being turned ON or OFF. Illustratively, a clock signal can be activated for the duration of data collection in an autoradiography procedure. Clock signals can be gated, e.g. to perform autoradiography successively over multiple similar samples, which can be advantageous to gather sufficient data with radionuclides having low activity or short halflife.

[0040] “Coincidence” refers to a property of two events being simultaneous to within a predetermined time interval dubbed a “coincidence window”. The events can be energy deposition by two distinct particles in a same detector (e.g. an alpha particle and a beta particle in a charged particle detector) or different detectors (e.g. a positron track detected in the charged particle detector and an annihilation gamma ray detected in a gamma-ray detector). Measured times can have some error due to particle or signal transit times, or due to statistical fluctuations. For example, a positron can deposit energy within a charged particle detector over some time duration which can end with annihilation of the positron and a proximate electron. The annihilation results in a pair of back-to-back gamma rays, one or both of which can travel to a scintillator sensor. A gamma ray can be absorbed or scattered in the scintillator, leading to emission of photons which travel to a light transducer in a photomultiplier. Further time is spent amplifying, shaping, and digitizing the ensuing electrical signal before a digital timestamp is attached to the event. Each time-of-flight or signal processing delay can also have variability. The coincidence window and statistical variations in detected event times are both associated with the timing resolution of the detectors. The coincidence time window can be set to accept simultaneously occurring events with some probability, such as 99%. A charged particle detector can have similar delays and variability. Time offsets can be set to account for time-of-flight and signal processing times, and the coincidence window can be set to account for variability. Ultimately, a positron track in the charged particle detector and a subsequent scintillation signal in the gamma-ray detector can desirably be determined to be coincident. Coincident events can often be associated with a same radioactive decay event, but other coincident events can be random coincidences. Studies described herein utilize coincidence windows between 10 - 200 ns, often 20 - 50 ns or 40 - 80 ns.

[0041] “Coincidence processing” is a process of analyzing detected events to identify coincident events. In examples, digital timestamps of two events can be compared. If the timedifference between the two timestamps (e.g. after application of appropriate time offsets) is less than or equal to a predetermined coincidence window, the events are coincident, otherwise the events are not coincident.

[0042] “Correlate” refers to an act of determining whether or to what extent a magnitude of one variable is associated with a magnitude of another variable. If some measure of the correlation is above a threshold, then the measured magnitudes (or distribution) of either variable can be used to predict the amplitudes (or distribution) of the other variable. Applied to autoradiographs of two radionuclides, intensities of two autoradiographs can be compared on a pixel-by -pixel or region-by -region basis to determine whether the spatial distribution of one radionuclide can predict the spatial distribution of the other radionuclide. In some examples, one or both autoradiographs can be preprocessed as part of the correlation act. To illustrate, the physics of alpha particle propagation and positron propagation can differ, leading to differing sharpness of the alpha and positron autoradiographs. Thus, it can be advantageous to apply sharpening to a more diffuse autoradiograph or to apply blurring to a sharper autoradiograph, in order to improve the measure of correlation between the two autoradiographs.

[0043] A “detector” is a subsystem operable to generate an output signal when energy (e.g. from a radioactive decay event) is deposited within an active region (e.g. a sensor) within the detector. Generally, a detector can include a sensor, a signal acquisition device, some analog signal processing electronics, a digitizer, and some digital signal processing electronics. As an illustration, a gamma-ray detector can include: a scintillator sensor which can generate scintillation light in response to energy deposition by a gamma ray; a photomultiplier incorporating a photoelectric signal acquisition device and electron amplification circuitry for analog signal processing; analog electronics for pulse shaping; a digitizer; and digital circuitry to apply a digital timestamp to digitized event data and form and output a digital record of event data. Filtering and calibration can optionally be performed in the analog or digital domain, or further downstream by a computer receiving the digital data packet. As another illustration, a spatially resolving charged particle detector can include: a semiconductor sensor in which electrons and holes can be excited into a conduction band in response to local energy deposition by a charged particle; an array of pixel electrodes on at least one surface of the semiconductor sensor, as the signal acquisition device; amplifier and comparator circuitry for analog signal processing; digitizers for each pixel; and digital circuitry to apply digital timestamps to digitized pixel data and output a digital packet of event data (dubbed “pixel event record” for a single pixel, or “deposition event record” formultiple pixels). Filtering and calibration can be performed in the analog or digital domain, or further downstream by a computer receiving the event data. The pixel event record can provide pixel coordinates at which charge (representing deposited energy) above a threshold was collected. Such a detector can distinguish energy deposited at one position from energy deposited at another position and is said to be “spatially resolving”. The pixel coordinates and amplitudes (e.g. charge or energy) in the event record(s) enable a position and energy to be assigned to a given sensed radioactive decay particle. In some examples, a charged particle detector can be implemented as two chips: a first chip can include the semiconductor sensor and attached pixel electrodes; a second chip (dubbed “readout chip” or simply “readout”) can include the digitizer and digital circuitry. Analog signal processing can be distributed between the two chips.

[0044] Two areas are “disjoint” if they neither overlap nor touch. Applied to pixel arrays, two regions can be regarded as disjoint if the shortest path from one region to the other passes through at least a predetermined threshold number of pixels (which are not part of either region). In some examples, the threshold can be one pixel but, in other examples, the threshold can be two, three, or up to 10 pixels, to address statistical fluctuations in contributing pixels towards the edges of an energy deposition pattern.

[0045] “Efficiency” is a quantitative measure of any one or more effects that detract from ideal measurements of radioactive decay events. Geometric efficiency reflects the arrangement of sample and detectors. To illustrate, the geometric efficiency of a planar charged particle sensor can be about 50%, because the other half of randomly emitted charged particles can be emitted in directions away from the charged particle sensor, due to the generally isotropic emission of radioactive decay particles. The geometric efficiency of a gamma-ray detector can be further reduced due to its finite depth - some emitted gamma rays may pass through a scintillator sensor without any interaction or energy deposition. Detector efficiency reflects other event loss in a detector. Measurements of alpha particle events can be less than ideal because of energy absorption between the emitting nucleus and the sensor, and further because of saturation at one or more pixels having large amounts of energy deposition. Because of typically higher energy deposition density, alpha particle events are more susceptible to pixel saturation than beta particle events. Some of these effects can be compensated through calibration, but others may not be amenable to compensation. Some gamma ray events can exhibit Compton scattering, and only a fraction of incident gamma ray energy can lead to scintillation within a sensor. Accordingly, the corresponding calibrated energy can be too low to satisfy an energy filter e.g. for a positron annihilation gamma ray.

[0046] “Energy” is a quantitative measure of the ability to do work, and can be measured in Joules (J), electron Volts (eV), or similar units. An emitted radioactive decay particle can be in motion, having kinetic energy about (l / 2)mv2for alpha and beta particles in a non- relativistic approximation, where m and v are mass and velocity, or about hc / Z for a gamma ray, where h is Planck’s constant, c is the speed of light, and A, is the wavelength of the gamma ray. Particles can interact with atoms in a sensor, doing work to eject electrons or other secondary emission, stimulate photon emission, or heat the sensor. The work done by the incident particle reduces its kinetic energy until none is left (or, in the case of a positron, the positron undergoes annihilation), and some effects in the sensor (e.g. conduction band charge carriers or emitted photons) can be measured. Measured sensor effects can be proportional to the kinetic energy of the incident particle. Accordingly, measured signal amplitudes (e.g. charge collected in a semiconductor sensor, or photomultiplier pulse height in a gamma detector) can be used to determine, or as a proxy for, energy deposited by the incident particle or the kinetic energy of the incident particle. In a semiconductor sensor, energy deposition can be distributed among multiple pixels. The distribution of collected charge or deposited energy over multiple pixels is dubbed an “energy deposition pattern”. Properties of the energy distribution pattern, such as size, shape, isotropy, or total amplitude, can be used to characterize a given event, determining e.g. a particle type, position, and energy associated with that event. As described herein, energy can be determined from signal amplitude(s) by applying one or more calibrations.

[0047] “Filter” refers to an act of retaining desirable event or pixel data and discarding other event or pixel data. Illustratively, filtering can be performed based on energy (or another amplitude measure) or on time (e.g. according to a timestamp and a coincidence window).

[0048] “Identify” refers to an act of assigning an event (e.g. some detected energy deposition in a detector) to a particular class of event. As one illustration, a gamma ray can be identified as a positron annihilation gamma ray based on one or more characteristics such as: coincidence with detection of a charged particle (particularly, a beta particle), or energy within a first predetermined window. As another illustration, a charged particle can be identified as an alpha particle based on one or more characteristics such as: isotropic energy deposition pattern, or energy within a second predetermined window. As a further illustration, a charged particle can be identified as a beta particle based on one or more characteristics such as: anisotropic energy deposition pattern, energy within a third predetermined window, or (for positrons) coincidence with a detected gamma ray having energy within the first predetermined window.

[0049] An “image” is a two-dimensional representation of a parameter value over a region of interest of a sample. An image can be represented as intensities of respective pixels in a pixel array, with pixel intensity being set proportional to the parameter value at each pixel location. In some examples, the imaged parameter can be density or activity of a radioactive nuclide within a sample, and the image is dubbed an “autoradiograph”. Density can be reported proportional to a count of decay events in a region corresponding to an image pixel. Activity can be reported proportional to initial energies of particular radioactive particles (e.g. alpha particles and / or beta particles, whose energy can be absorbed near the decayed nucleus) in a region corresponding to an image pixel. In some examples, image pixels can be in 1:1 correspondence to pixels of an associated charged particle detector, but this is not a requirement and, in other examples, the image pixel array can be scaled, rotated, clipped, or otherwise transformed relative to the detector pixel array. Data presented in images can be filtered e.g. based on particle energy.

[0050] “Infuse” refers to a process whereby a material is introduced into a biological sample and allowed to disperse, e.g. by fluid transport, diffusion, or metabolic processes. Of interest herein is infusion of materials containing molecules tagged with one or more radionuclides. The material can be introduced by injection, surgical implantation, intravenously, nasally, or orally. The material can be introduced in one or more administrations. In examples where a sample is infused with two or more radionuclides, same or different materials or administrations can be used for the several radionuclides.

[0051] “Isotropy” refers to the absence of angular variation in a two-dimensional distribution such as an energy deposition pattern. Various measures can be used to quantify isotropy. One technique described in the context of FIG. 7 is based on occupancy of contributing pixels within a bounding area of an energy deposition event. Other techniques include fitting an energy deposition pattern to a two-dimensional function such as a circle or ellipse. To illustrate, the distribution can be fitted to an ellipse or ellipse-shaped function, and M = 1 - e can be used as a measure of isotropy M, where e is the eccentricity of the ellipse. To illustrate, a circle has eccentricity e = 0 and isotropy M = 1, while a straight line segment can have e = 1 and M = 0. Alternatively, amplitudes of an energy deposition pattern can be fitted to a Gaussian, cosine, or other peaked distribution over a two-dimensional area. Still further, a parameter such as standard deviation or full- width at half-maximum (FWHM) can be evaluated in different directions, the percentage variation p over directions can be calculated, and M = 100 - p can be used as a measure of isotropy.

[0052] Often, an object can have a “major surface”, which is a surface of the object whose area is not substantially exceeded by any other surface of the substrate. A “planar” object is an object having parallel major surfaces, and the distance between those surfaces is the “thickness” of the object. A “transverse” plane is parallel to a major surface, and “transverse extent” is a maximum dimension of a major surface in its own plane.

[0053] A “particle” refers to a product of radioactive decay or positron annihilation. Thus, alpha particles, beta particles (including electrons and positrons) and gamma rays can be particles. An alpha particle is a bound combination of two neutrons and two protons, similar to a4He nucleus. The term beta particle encompasses electrons (e-) and positrons (e+), which are antiparticles. A gamma ray is a photon emitted during nuclear decay or electron-positron annihilation. Gamma rays commonly have energies above 100 keV. A particle need not have mass; gamma rays or neutrinos are examples of massless particles. The term particle excludes secondary electrons, secondary photons, or other objects arising from processes other than radioactive decay or positron annihilation.

[0054] A “photomultiplier” generates an amplified electrical signal from an optical input. Thus, a photomultiplier incorporates a light transducer device and a charge amplification device. Some common photomultipliers can be embodied in a vacuum tube (dubbed a “photomultiplier tube” or “PMT”), as a silicon photomultiplier (“SiPM”) or a microchannel plate (“MCP”). A “light transducer” or “photodetector” is a device which converts an optical signal into an electrical signal, e.g. through the photoelectric effect. To illustrate, the first dynode in a PMT can serve as a light transducer, or an avalanche photodiode can provide both light transducer and amplification in a SiPM. Common SiPMs incorporate an array of avalanche photodiodes and can operate at <100 V. In some examples, a Hamamatsu R13089- 100 PMT (Hamamatsu Photonics, Hamamatsu, Japan) has been used.

[0055] A “pixel event record” is a digitized data packet outputted by a detector, which contains energy deposition information for one pixel of a sensor. A “deposition event record” is a digitized data packet which contains energy deposition information for multiple sensor pixels that report energy deposition in coincidence. Pixel event records and deposition event records can contain amplitude, time, and position information.

[0056] “Positron annihilation” is a process whereby a positron encounters an electron and both are replaced by a pair of gamma rays. The gamma rays have about 511 keV each, measured in a reference frame of the center-of-mass of the electron-positron system.

[0057] A “power supply” is a device operable to provide electrical energy required to operate one or more other devices. Some detector components such as photomultipliers andsemiconductor sensors can require a bias voltage ranging from tens to thousands of volts. Other electronics can require voltages usually in a range 1 - 30 V for amplifiers and other analog components or 0.5 - 5 V for digital logic, analog to digital converters, or clock sources. Computers can variously be powered by wall power or wall adapters generating up to about 20 V.

[0058] An “event” is a detection of energy deposition by a detector, which can correspond to one or more particles. Some events can be associated with particular classes of particles. Thus, a “gamma ray event”, a “positron event”, or an “alpha particle event” can respectively refer to events identified with gamma rays, positrons (including a coincident annihilation gamma), or alpha particles.

[0059] A “radionuclide” is a nuclear species (or chain of such species) that exhibits radioactivity. In some instances, a parent nucleus decays into a stable daughter nucleus and the radionuclide is the single species of the parent nucleus. In other instances, a parent nucleus decays into a daughter nucleus which is also radioactive, and the decay chain can continue over multiple species until one or more stable species are reached. In such case, as used herein, the parent nuclide refers collectively to all radioactive species in the decay chain.

[0060] A “readout chip” (or simply “readout”) is a component of a detector forming and transmitting digital data packets containing measurements associated with detected particles. A readout chip can be implemented using digital electronics and can perform functions such as digitization, filtering, calibration, other digital signal processing, or transmitting data over a network or bus to a data receiver.

[0061] Referring to a sample, a “region of interest” (ROI) is a portion of a sample over which a radionuclide is mapped. Referring to a deposition event, a region of interest is a portion of a sensor area or volume in which energy deposition is being analyzed. The term ROI does not refer to any human interest.

[0062] A “sample” is a quantity of material upon which autoradiography can be performed. In some examples, the sample can be a portion of tissue extracted from a living or recently living organism. A sample can include organ tissue. Some samples can be obtained by biopsy.

[0063] A “scintillator” is a class of sensor which emits light when a detected particle, such as a gamma ray, deposits energy within the sensor volume. In some examples, a CeBr3 crystal scintillator (Berkeley Nucleonics, San Rafael, CA) has been used.

[0064] A “sensor” is a component of a detector in which a particle can interact with the sensor material to generate charges or photons which can be sensed or measured. The sensorin a gamma-ray detector can be a scintillator which can be coupled to a photomultiplier. The sensor in a charged particle detector can be a planar semiconductor sensor to which electrodes are attached, e.g. in a pixel array. With bias voltage applied, signals developed on the pixel electrodes can be used to measure charge, arrival time, or position of an incident particle.

[0065] “Signal processing” refers to operations performed on electrical waveforms or digital data representing detected particles. Electrical waveforms (e.g. pulses) can be analog signals having continuously variable amplitude, and processing of these signals can be regarded as being in the “analog domain”. Analog signal processing can include amplification, pulse shaping, filtering, gating, summing, or comparison with thresholds. Signal processing can include digitization, e.g. by an analog-to-digital converter to convert an analog signal into a digital signal having a finite number (usually two) of amplitude levels. Processing of digital signals can be regarded as being performed in the “digital domain”, and can include filtering, calibration, summing, time-tagging, measurement (e.g. of position or energy), classification (e.g. alphas vs beta particles), coincidence processing, or comparison with thresholds. Such signal processing can be performed independently on individual decay events, and can be customized for respective types of particles.

[0066] A “spatial signature” is a set of values for one or more characteristics of the energy deposition from a particle detected in a spatially resolving particle detector. Example characteristics can include size or extent, isotropy or shape, amplitude, or amount or distribution of energy deposited. The spatial signature can be derived from analysis of an energy deposition pattern. Based on spatial signature, a given particle or detector signal can be classified as e.g. an alpha particle or a beta particle.

[0067] A “subject” can be a living human, live or sacrificed animal, or plant from whom or from which a sample is extracted. Some human or animal subjects can be radiotherapy patients.

[0068] In the context of radionuclide infusion, “tagging” refers to an act of adding a radioactive atom to a molecule, or replacing an existing atom on a molecule with a radioactive atom. In varying examples, a tagged molecule can be a radiotherapy drug; a biologically active molecule such as a protein, antibody, or drug; a molecule participating in biological processes such as glucose; or an inert molecule. In some instances, tagging can be performed on an ionic compound in solution, in which case an original ion can be replaced with an ion of a radioactive atom.

[0069] “Time- tagging” refers to an operation of assigning a time or timestamp to a detected event. The time tag (i.e., “timestamp”) can be a digital data item attached to or embedded within a data record representing the time-tagged event. Timestamps can be derived from clock signals similar to 211, 221 of FIG. 2.

[0070] The terms “top,” “bottom,” “up,” “down,” “above,” “below,” “horizontal,” “vertical,” and the like are used for convenience, with respect to a common configuration in which a planar sample is oriented horizontally and arranged above a parallel planar semiconductor sensor. One of ordinary skill will understand from this disclosure that a choice of actual orientation can be varied without departing from the scope of the disclosed technologies.First Example Method

[0071] FIG. 1 is a flowchart 100 of a first example method according to the disclosed technologies. In this method, alpha, beta, and gamma particles are detected and two images are formed: one of alpha particle events (“alpha image”) and one of positron events (“positron image”). These images can be autoradiographs of a sample. Advantageously, separation of coincident alpha and beta signals improves the accuracy with which the position of each can be determined.

[0072] At process block 110, spatially resolved first signals of radioactive decay products are acquired from a first detector. These radioactive decay products can include beta particles, some (first) alpha particles coincident with respective beta particles, and (second) other alpha particles anti-coincident with the beta particles. These species are illustrated schematically in a Venn diagram 128: circle 114 represents beta particle events; circle 112 represents alpha particle events; intersection 124 of circles 112, 114 represents coincident alpha, beta events; area 122, which is the portion of circle 112 excluding intersection 124, represents alpha particle events which are anti-coincident with beta particle events; and area 126, which is the portion of circle 114 excluding intersection 124, represents beta particle events which are anti-coincident with alpha particle events.

[0073] At process block 120, the first signals of the first alpha particle events are separated from the first signals of respective beta particles with which each first alpha particle is coincident. Illustratively, regions 124 corresponds to coincident first signals of alpha particles and beta particles, which are separated from each other at block 120.

[0074] At process block 130, an alpha image is formed based on the first signals of the first alpha particles (within area 124) and the first signals of the second alpha particles (withinarea 122). That is, the first signals of alpha particles both coincident and anti-coincident with beta particles can be used to form the alpha image.

[0075] At process block 140, second signals representing positron- annihilation gamma rays can be acquired from a second detector. Gamma rays are represented as circle 132 in another Venn diagram 158. Circle 116 can represent all particles identified as beta particles (e.g. electrons and positrons), or all particles tested for coincidence with gamma rays (in some examples, also including alpha particles or other energy deposition events observed in the charged particle detector).

[0076] At process block 150, a positron image is formed based on the first signals of the beta particles (e.g. within area 154 of Venn diagram 158; these beta particles are included within circle 114 of Venn diagram 128) which are coincident with the second signals.

[0077] Numerous variations and extensions can be implemented within scope of the disclosed technologies. In examples, each first signal can contain amplitude information for multiple pixels associated with an event, indicative of charge collected or energy deposited at each pixel. The energy deposition patterns can be analyzed to determine measures of isotropy for respective events. The first alpha particles (e.g. within 124) or the second alpha particles (e.g. within 122) can be identified based on respective measures of isotropy exceeding a predetermined threshold. The energy deposition patterns can also be used to identify some of the first signals as beta particles. In some examples, a detected first signal can be identified as a beta particle based on its respective measure of isotropy being below a second threshold, or based on determining that an energy measure (derived from the energy deposition pattern) is within a predetermined range. In some instances, energy deposition reported in the first signals can be simultaneously present in disjoint regions, each region defined by a respective group of pixels. Thus, the separating at block 120 can be performed by separating disjoint regions of coincident energy deposition in the first signals, and processing each disjoint region independently to determine whether the energy deposition in the respective region can be identified as an alpha particle or a beta particle. In this way, for events of set 124, respective regions can be identified containing one of the first alpha particles and a corresponding coincident beta particle.

[0078] In further examples, a respective alpha particle energy can be determined from the first signals for various ones of the first alpha particles and the second alpha particles. The first signals can be pixel event records or deposition even records from the first detector, and the energy of a given alpha particle can be obtained by summing respective energies for each of a plurality of pixels. The readout for each of the summed pixels can include an amplitude,e.g. proportional to an amount of charge collected at that pixel, and a calibration can be applied to convert the amplitude into the respective energy for that pixel. The calibration can be nonlinear, e.g. to compensate for saturation.

[0079] The first method can extend to applying first and second clock signals, derived from a common clock source, to the first and second detectors respectively. Based on the first clock signal, the first signals can include first digital timestamps, e.g. included with each pixel event record or deposition event record. The method can extend to producing the second signals by digitizing analog signals derived by amplifying (e.g. with a photomultiplier) signals from a scintillator, and applying second digital timestamps based on the second clock to the digitized signals. The digital timestamps of the second signals (e.g. gamma-ray events) can be compared with the digital timestamps of the first signals (some of which are beta particle events, e.g. circle 114) to determine coincidences, which can be identified as positron events.

[0080] The method can extend to analyzing correlation between intensities of the alpha and positron images at corresponding spatial positions of the images. The spatial positions can be respective image pixels, or can be multi-pixel regions within the images.

[0081] The first method, and any of its extensions or variations, can be performed by, or under control of, one or more hardware processors configured to execute instructions stored in computer-readable media.Example Apparatus

[0082] FIG. 2 is a schematic diagram 200 of an example apparatus. This apparatus can be used to perform simultaneous independent autoradiography of two radionuclides in a sample. Dashed lines are used to show entities that can be part of the environment of the depicted apparatus.

[0083] Charged particle detector 210 and gamma-ray detector 220 are positioned proximate to a sample volume in which sample 202 can be placed. While detectors 210, 220 are shown on opposite sides of sample 202 in FIG. 2, this is not a requirement. In other examples, detector 210 can be placed between sample 202 and detector 220, and other configurations can also be used. Desirably, gamma-ray detector 220 can be situated to optimize solid angle coverage over at least a region of interest of sample 202, improving detection efficiency for positron events. In examples, sample 202 and detector 210 are both generally planar, and can be placed adjacent and parallel to each other.

[0084] Clock source 230 can be coupled to provide synchronized clock signals 211, 221 to charged particle detector 210 and gamma-ray detector 220 respectively. Computer 240 can becoupled to detectors 210, 220 and clock source 230 as shown. Additional items not shown in FIG. 2 can include electrical components, such as power supplies or instrumentation, or mechanical components, such as support structures.

[0085] Computer 240 can be configured to execute program logic 242, which can be embodied as computer-executable instructions stored on computer-readable media. Through execution of program logic 242, computer 240 can perform various operations represented as flowcharts 214, 224 within logic 242.

[0086] As indicated by arrow 212, event data from charged particle detector 210 can be transmitted to computer 240 and processed according to left-hand flowchart 214. Detector 210 can be a spatially resolving detector as described herein, and the event data can be in the form of a pixel event record for each contributing pixel, or a deposition event record including all coincident contributing pixels. The event data can include digital timestamps according to clock 211.

[0087] At process block 213, first signals SI can be acquired from detector 210. In some instances, multiple particles can be detected in coincidence in different areas of detector 210. That is, pixels in region Ra of detector 210 can detect deposited energy at times {Ta} and pixels in region Rb can detect deposited energy at times {Tb}, where times {Ta} and { Tb } meet a predetermined coincidence criterion. Illustratively, midpoints Ta, Tb of {Ta}, {Tb} can be the same, or within a predetermined coincidence window ATc, e.g. \Ta — Tb | < ATc. In some instances, multiple particles can be reported within one deposition event. At block 223, the first signals of coincident particles can be separated. To illustrate, region Ra and region Rb can be separated by a gap of at least a predetermined number of pixels having deposited energy or collected charge below a threshold, and the regions can be separated based on identifying that regions Ra, Rb are disjoint. Occasionally, but less frequently, disjoint regions of three or more particles can be identified and separated.

[0088] Whether coincident or not, the spatial energy deposition patterns of the first signals can be analyzed at block 233. Those first signals having the characteristics of a detected alpha particle can be identified as alpha particles at block 243. Illustratively, for a pair of coincident particles, the energy deposition pattern in region Ra can be identified as an alpha particle al, while the energy deposition pattern in region Rb can be identified as a beta particle. Such a coincident pair would fall within area 124 of Venn diagram 128. Other types of coincidences can also occur, for example alpha and gamma, two alphas, or an alpha and two betas, withrelative rates depending on activities of various nuclei along the decay chain of an alphaemitting radionuclide in sample 202.

[0089] The analysis and identification at blocks 233, 243 can also be applied to first signals that are not coincidences. Some of these first signals can have spatial energy deposition patterns characteristic of alpha particles, which would fall within area 122 of Venn diagram 128, and can also be identified as alpha particles a2. These alpha particles a2 are anticoincident with any detected beta particles. The non-coincident first signals can also include beta particles of area 126 and, sometimes, gamma rays or other stray signals.

[0090] At block 253, an image II can be formed using the positions of both coincident alpha particles (al) and anti-coincident alpha particles (a2). Image II can be output at block 263, e.g. to output device 250. Non-limiting examples of output device 250 include a display, disk storage, or printer.

[0091] Event data from gamma ray detector 220 can also be transmitted to computer 240, as indicated by arrow 222, and can be processed according to right-hand flowchart 224. Event data can be in the form of data packets with a digital timestamp according to clock 221. Raw or calibrated energy data can also be included in the data packets.

[0092] At process block 215, second signals S2 can be acquired from detector 220. At process block 225, coincidences between a given signal S2 and a corresponding signal SI can be determined based on their respective digital timestamps. Dotted arrow 217 shows the dependence of block 225 on data acquired at block 213. At block 245, coincidences from block 225 can be identified as positron events. Then, at block 255, positron image 12 can be formed based on positions of the positron events in charged particle detector 210. At block 265, image 12 can be outputted to device 250.

[0093] Numerous variations and extensions can be implemented within scope of the disclosed technologies. In some examples, charged particle detector 210 can include a planar semiconductor sensor, the area of which can be subdivided into a rectangular array of pixels, each of which can have a bias voltage applied between an anode and a cathode. Radioactive decay products incident on the sensor can lose energy as they traverse one or more pixels. The deposited energy can stimulate secondary emission and excitation of carriers into a conduction band, leading to charge collection on the pixel electrodes. A readout chip can be coupled to the sensor and can be configured to provide digital packets indicating an arrival time and a signal amplitude for each pixel. In varying examples, the arrival time can be common to multiple pixels in a single deposition event, or provided individually for each pixel. The signal amplitude can be proportional to the amount of charge collected, or can becalibrated into energy units as described herein. The arrival time can be based on counting cycles of clock 211 provided to detector 210. Each pixel of detector 210 can have length and width dimensions, measured along row and column axes of the pixel array, in a range 10 pm to 1 mm, 20 pm to 100 pm, or about 50 pm. The pixels can form an array having length and width dimensions independently configurable from 16 to 4096 pixels, 64 to 1024 pixels, or about 256 pixels in each direction. The pixels can form an array having length and width dimensions in a range 0.2 cm to 20 cm, 0.5 cm to 5 cm, about 1 to 2 cm, or about 2 to 4 cm. Multiple detector chips (such as Timepix3) can be arrayed to increase the overall dimensions of detector 210, as may be desirable for image comparisons on large organs or other large samples.

[0094] Gamma ray detector 220 can include a scintillator, a bias voltage supply, a photomultiplier, and a digitizer. The photomultiplier can be coupled to receive photons from the scintillator and a bias voltage from the bias voltage supply, and can be configured to output electronic signals corresponding to the received photons. The digitizer can be configured to receive the electronic signals from the photomultiplier and clock 221 from clock source 230 and can be configured to generate timestamped digital packets 222 corresponding to the received photons. One detector 220 can include multiple scintillator crystals. Multiple detectors 220, e.g. with distinct photomultipliers and digitizers, can be parallelized.

[0095] Non-limiting scintillator materials include BGO, Lil, Nal, ZnS, CeBr3, a mixed cerium halide, a cerium-doped halide, or a plastic scintillator. Example photomultipliers include a semiconductor photomultiplier, a photomultiplier tube, or a microchannel plate. The bias voltage to a photomultiplier tube can be in the range 1 to 5 kV, while the bias voltage to semiconductor or microchannel plate photomultipliers can be in a range 20 to 500 V, or about 100 V. The digitizer can be configured to apply an energy or amplitude filter to the received signals to generate timestamped digital packets 222 for events having energy within a predetermined range, e.g. characteristic of positron-annihilation gamma rays, while discarding events having energy outside the predetermined range. Alternatively, the energy or amplitude filter can be applied in computer 240 at block 225.

[0096] Separating signals at block 223 can be based on disjoint groups of pixels (e.g. regions Ra, Rb) having amplitudes above a first threshold separated by at least a predetermined number of other pixels having amplitudes below the first threshold. The alpha particle identification at block 243 can be based on determination of shape or size of the respective energy distribution pattern at block 233.

[0097] In some examples, pixels of image II can have intensities representative of total energy deposited at the corresponding position of detector 210. Thus, an alpha particle determined to have energy Ei and a position of Xi, Yi (as image pixel indices) can be used to update the image intensity A(Xi, Yi) according to A(Xi, Yi) A(Xi, Yi) + Ei. That is, image Il can be based on both positions and energies of the various first and second alpha particles. In other examples, the pixels of image 12 can have intensities representative of alpha particle counts rather than energy. That is, the illustrative alpha particle can contribute to image intensity A(Xi, Yi) according to A(Xi, Yi) A(Xi, Yi) + 1.

[0098] To determine Ei for each alpha particle, a calibration can be applied to the charge collected on each pixel to obtain an energy measure, which can be summed over all pixels within the energy distribution pattern of the alpha event. The calibration can account for energy losses outside the semiconductor sensor, e.g. in a mylar film separating sample 202 from detector 210. Additionally, pixels can sometimes saturate, meaning that doubling energy deposition within a pixel can lead to less than 2x increase in the collected charge. To account for saturation, a nonlinear calibration can be stored and applied to the pixel signal amplitudes in order to determine the energy Ei. Additional calibration can be performed to account for alpha energy loss within sample 202. Experiments with varying sample thicknesses can be used to determine tissue absorption. Monte Carlo simulations can also be used.

[0099] Coordinates Xi, Yi of a given alpha particle can be determined as a centroid of the respective energy deposition pattern. Similarly, position coordinates of a positron event can be determined as a centroid of an energy deposition pattern of the corresponding first signal (e.g. a beta particle in area 114 that is coincident with an annihilation gamma event detected by detector 220). Imaging the centroid of an alpha energy deposition event can more accurately reflect the position of a source radionuclide, as compared to imaging the entire energy deposition pattern.

[0100] In further examples, logic 242 can be extended to perform analysis of correlation between images II, 12; and to output a measure of the correlation. Such output can be used to determine whether or to what level of accuracy positron image 12 can be used to establish the distribution of the alpha-emitting radionuclide shown in alpha image II.Example Charged Particle Detector

[0101] FIG. 3 is a diagram 300 illustrating use of a spatially resolving charged particle detector according to the disclosed technologies. Certain aspects of structure, environment, and output of a charged particle detector are described with reference to diagram 300.Charged particle detector 310 is shown in dashed outline and can be similar to detector 210 ofFIG. 2. For convenience of illustration, the layer stack of detector 310 is drawn with gaps between layers. However, in a practical device, adjacent layers of detector 310 can be in physical or electrical contact (electrical contacts not shown).

[0102] Semiconductor sensor 350 is shown sandwiched between a common electrode 352 and an array 354 of pixel electrodes. Pixel electrodes are coupled to readout chip 360. Bias power supply 340 provides bias voltage between common electrode 352 and (through readout 360) pixel electrodes 354, developing an electric field in the volume of sensor 350.Deposition of energy by a charged particle at any point within sensor 350 can excite charge carriers, e.g. electrons and holes, into respective conduction bands, which can be swept to electrodes 352, 354 under influence of the biasing electric fields. Thus, energy deposition leads to charge collection at nearest one or more pixel electrodes 354.

[0103] Readout 360 can perform analog signal processing, digitization, time-tagging, and digital signal processing of charge signals received from pixel electrodes 354. In some examples, each pixel has an independent signal processing path: a charge pulse is amplified and compared with a threshold, the amplified signal rises and crosses the threshold at time tl. The amplified charge pulse relaxes slowly and eventually drops below the threshold at time t2. The time above threshold St = t2 - tl can be related to the total amount of charge in the charge signal received at electrode 354, which in turn can be related to the amount of energy deposited e.g. by a radioactive decay particle. Additionally, readout 360 can receive a clock signal from clock source 330 and can time-tag the event according to time tl.

[0104] An output record can be generated for each charge pulse recorded by each pixel. Readout 360 can include an output controller which collects, formats, and outputs pixel output records 371-379, e.g. as a temporally ordered list 370. Records can be transmitted (arrow 363) over a bus or network connection to a storage device or computer.

[0105] As shown, output list 370 contains a list of N pixel event records, 1, 2, ... i, ... N, ordered according to timestamp as indicated by time arrow 365. Each record 375 can include at least three fields: Ti, indicating digital timestamp of the ithrecord; Pi, identifying the pixel reported in the ithrecord; and Ei, indicating a measure of the energy deposition for the ithrecord. In varying examples, the Ei field can report raw or amplified charge received at an instant pixel electrode 354, time-over-threshold, or a calibrated pixel energy value.Regardless of which form Ei takes, the Ei value can be used to determine the energy deposited at that pixel’s transverse position within the semiconductor sensor.

[0106] In examples, detector 310 can be implemented using a commercial product such as Timepix3 (CERN Medipix Collaboration, Geneva, Switzerland; Advacam, Prague, CzechRepublic; or Amsterdam Scientific Instruments, Amsterdam, Netherlands). In trials with representative biological samples, Timepix3 detectors 310 were found to provide substantially linear dosimetry (with deviations from linearity under 10%) up to 600 Becquerel (Bq) for samples infused with alpha-emitting223Ra and up to 1.5 kBq for samples infused with positron-emitting18F. Commonly encountered tissue samples with 10 - 100 Bq radioactivity are well within range of such detectors. Turning to time resolution, detected radioactivity decay events aggregate charge pulses over numerous pixels, and were found to have time resolution between 5 - 10 ns (often 7 - 8 ns) full-width half-maximum (FWHM).

[0107] Turning to sample 302, a generally planar sample 302 can be placed adjacent and parallel to sensor 350. Optionally, a thin film (e.g. 12 pm thick mylar) can be used as a protective barrier between sample 302 and detector 310, to protect detector 310 from radioactive contamination by sample 302. In some examples, Melinex S polyester mylar film (Tekra, LLC, New Berlin, WI) has been used. Radioactive decay particles emitted in sample 302 can travel (arrows 306) to semiconductor sensor 350 and lose energy as they interact with the material in sensor 306. While radioactive decay particles can be isotropically emitted over 4n solid angle, charged particles have relatively high rates of linear energy transfer (LET), and particle detection rates drop off quickly with increasing transverse distance from a point source (e.g. in sample 302). The full-width half maximum of the energy deposition pattern of a radioactive point source is commonly between 0.01 and 1.0 mm. In one experiment, a full-width half-maximum of 87 pm was measured for a 10 pm diameter alpha emission source on a detector with 55 pm pixel pitch. A map of energy deposition detected within sensor 350 is a convolution of (i) the transverse distribution of radioactive sources within sample 302 (ii) with a point source kernel. Subject to a small amount of blurring, the map of energy deposition is a close facsimile of the source distribution, a principle commonly used in autoradiography. FIG. 14 shows example alpha and positron images that follow the structure of a tissue sample.

[0108] Numerous variations and extensions can be implemented within scope of the disclosed technologies. Multiple charged particle detectors can be provided, e.g. to accommodate larger samples 302 or to increase geometric efficiency of charged particle detection. To illustrate, sample 302 can be sandwiched between two detectors 310.

[0109] Records 371-379 can contain additional or different fields. For example, two or more of time-over-threshold, total collected charge, or calibrated pixel energy can be included in each record 375. Output data can be event organized rather than pixel organized. To illustrate, pixel event data can be gathered within a coincidence window and can be presentedas respective charged particle deposition events at respective times Ta, Tb, etc. To illustrate, charged particle event data can be presented as {Ta, (Pl, El), ... (Pm, Em)}, {Tb, (Pl, El), ... (Pn, En)}, ... .

[0110] Pixel energy calibration, e.g. from charge or time-over-threshold to energy, can be provided by a vendor or determined in a separate calibration experiment with a known source. In varying examples, pixel calibration can be applied within readout 360 or downstream in a computer (similar to computer 240 of FIG. 2).

[0111] Pixel event data can be filtered before output. To illustrate, only pixel event data having Ei greater than or equal to an output threshold value can be included in outputted list 370. This output threshold value can be different from the threshold value used to determine e.g. tl, t2 as described above.Example Gamma-Ray Detector

[0112] FIG. 4 is a diagram 400 illustrating use of a gamma-ray detector according to the disclosed technologies. Certain aspects of structure, environment, and output of a gamma-ray detector are described with reference to diagram 400. Gamma-ray detector 420 is shown in dashed outline and can be similar to detector 220 of FIG. 2.

[0113] As shown, scintillator sensor 450 is positioned facing sample 402, and is optically coupled to photomultiplier 455, which in turn is electrically coupled to electronics module 460. Deposition of energy by a gamma ray at any point within sensor 450 can excite atoms in sensor 450, subsequent de-excitation of which can generate optical photons. Optical photons reaching photomultiplier 455 can generate initial electrons, e.g. through the photoelectric effect, which can be amplified into a shower of electrons. Bias power supply 440 provides bias voltage to photomultiplier 455 to support electron amplification. Thus, energy deposition leads to an electron pulse outputted from photomultiplier 455 to electronics module 460.

[0114] Electronics module 460 can perform analog signal processing, digitization, timetagging, and digital signal processing of pulses received from photomultiplier 455. The amplitude of each pulse can be proportional to the energy deposited by an incident gamma ray in scintillator 450. Module 460 can receive a clock signal from clock source 430 and can time-tag each pulse according to a time measured along its rising edge.

[0115] An output record can be generated for each pulse outputted by photomultiplier 455, for at least those pulses exceeding a threshold amplitude. Module 460 can include an output controller which formats and outputs pixel output records 471-479, e.g. as a temporallyordered list 470. Records can be transmitted (arrow 463) over a bus or network connection to a storage device or computer.

[0116] As shown, output list 470 contains a list of N events, 1, 2, ... i, ... N, ordered according to timestamp as indicated by time arrow 465. Each record 475 can include at least two fields: Ti, indicating digital timestamp of the ithrecord; and Ei, indicating a measure of the energy deposition for the ithrecord.

[0117] In examples, module 460 can incorporate a scintillator readout such as model DT5725 from Caen S.p.A. (Viareggio, Italy) can be used.

[0118] Turning to sample 402, a generally planar sample 402 can be placed proximate to sensor 450. Gamma rays emitted in sample 402 can travel (arrows 406) into scintillator sensor 450 and lose energy as they interact with the material in sensor 406. Gamma rays can be isotropically emitted over 4TT solid angle. Having relatively low rates of linear energy transfer (LET), particles from a given point within sample 402 can be detected throughout sensor 402. For practical sample sizes of interest, the position at which scintillation photons are generated in sensor 402 does not provide much information about the location at which the associated gamma ray was emitted. However, the time at which the photomultiplier pulse is outputted to module 460 is within a few nanoseconds of the gamma ray emission time. Thus, the event timestamp Ti can be used to detect coincidences between gamma events outputted by gamma-ray detector 420 and charged particle events outputted by detector 310 of FIG. 3. In particular, coincidences can be used to confirm beta events from detector 310 as positron events, and the position of those coincident beta events can allow generation of an accurate autoradiograph of a positron-emitting radionuclide in sample 302 or 402.

[0119] Numerous variations and extensions can be implemented within scope of the disclosed technologies. Multiple gamma-ray detectors can be provided, e.g. to increase geometric efficiency of gamma ray detection. To illustrate, sample 402 can be positioned between two or more detectors 420. Detectors 420 can be placed behind detector(s) 410 (as seen from sample 402), or behind each other, to increase the volume in which gamma rays can be detected.

[0120] Records 471-479 can contain additional or different fields. For example, with multiple scintillators 450 or multiple photomultipliers 455, each record 475 can include an identifier of the reporting scintillator 450 or photomultiplier 455.

[0121] Gamma ray energy calibration can be provided by a vendor or determined in a separate calibration experiment with a known source. In varying examples, energy calibrationcan be applied within readout 460 or downstream in a computer (similar to computer 240 of FIG. 2).

[0122] Gamma ray event data can be filtered before being output. To illustrate, only gamma ray event data having Ei within a window predetermined window encompassing 511 keV can be included in outputted list 470.Second Example Method

[0123] FIG. 5 is a flowchart 500 of a second example method according to the disclosed technologies. In this method, pixel event data, which can be similar to 470 of FIG. 4, is analyzed and used to generate autoradiographs similar to images II, 12 described in context of FIG. 2. This method can be performed online, e.g. as the data is being collected, or offline, e.g. after a data acquisition session is completed.

[0124] At process block 510, pixel list data (470, 212) can be received, e.g. from a charged particle detector or a storage device. At block 520, pixel event records having timestamps within a coincidence window can be gathered into a record of an energy deposition event for a given “event time” associated with the coincidence window. The associated pixels are dubbed “contributing pixels”. In some instances, an energy deposition event contains only pixels near to each other, indicating that only one charged particle was detected at the instant time. In these instances, determination of particle energy and position can be straightforward. In other instances, the energy deposition event contains pixels of two or more groups of pixels distributed over the pixel array, which can indicate energy deposition contributions from two or more particles, dubbed “contributing particles”. In such instances, additional operations can be performed in order to accurately determine energies and positions of the contributing particles.

[0125] At block 530, a decision can be made whether the energy deposition event contains multiple regions of interest. To illustrate, the contributing pixels can be scanned, and neighboring pixels can be gathered into groups. Because energy deposition patterns of alpha and beta particles usually have multiple contributing pixels, groups having fewer than a first threshold number of pixels can be discarded, and the number of groups remaining can be a count of energy deposition regions of interest at the event time. If there is only one such region, then the method can follow the N branch from block 530 directly to block 550 for particle identification. Notably, any alpha particles detected in this way are anti-coincident with beta particles (area 122 of FIG. 1).

[0126] However, if multiple regions of interest are found, the method can follow the Y branch from block 530 to block 535. At block 535, the multiple regions can be separated forindividual processing, which can commence at block 540. At block 550, particle identification can be performed on an instant region. That is, where multiple regions were found, each region can be processed in turn to identify a respective particle. In some instances, these regions may include a beta particle and a coincident alpha particle (area 124). Block 550 is also invoked for anti-coincident particles, which can be alpha particles (area 122) or beta particles (126).

[0127] At decision block 560, the method branches according to the result of particle identification at block 550. For alpha particles, the method follows the a branch from block 560 to block 562, where energy E and position (X, Y) of the alpha particle can be determined. Then, at block 572, this alpha particle information can be saved to an alpha-emitting radionuclide image (II). That is, an alpha-emission autoradiograph can be incremented at position (X, Y) by an amount proportional to E. The method proceeds to decision block 580.

[0128] Returning to decision block 560, the method follows the e+ branch in the case of positrons, from block 560 to block 564, where position (X, Y) of the positron can be determined. Then, at block 574, this positron information can be saved to a positron-emitting radionuclide image (12). That is, a positron-emission autoradiograph can be incremented at position (X, Y) by a count of 1. The method proceeds to decision block 580.

[0129] In further instances, block 550 can determine that a given region corresponds to neither an alpha particle nor a positron. To illustrate, the given region can be a beta particle track without any coincident gamma ray event, as can arise from (i) a beta decay within the alpha-emitting radionuclide’s decay change (as discussed in context of FIG. 10), (ii) a positron event with undetected annihilation gamma rays, or (iii) a cosmic ray, gamma ray, or other noise. In such instances, the method can follow the X branch from block 560 directly to block 580.

[0130] Thus, all paths from block 560 eventually reach decision block 580. If a current deposition event had multiple regions and some regions remain to be processed, the method follows the Y branch from block 580 back to block 540 to process the next region. Otherwise, the method follows the N branch from block 580 to block 590. If any events remain to be processed in the incoming data stream (block 510), the method can follow the Y branch from block 590 back to block 520 to process additional events. Otherwise, the method can follow the N branch from block 590 to block 599 and stop.

[0131] Numerous variations and extensions can be implemented within scope of the disclosed technologies. In some examples, multiple alpha-emitting nuclei can be imaged separately, e.g. based on differing alpha particle energies. That is, at block 572, each alphaparticle can be accumulated into a respective image according to its energy E. Other variations are described in context of FIGS. 6-7 or elsewhere herein.Example Particle Identification

[0132] FIG. 6 is a flowchart 600 of an example method for identifying particles detected by a charged particle detector. This method can be used to implement block 550 of FIG. 5. In this method, shape of an energy distribution pattern, total deposited energy, and gamma coincidence can be used to distinguish and identify alpha particles and positrons.

[0133] At optional block 651 , energy calibration can be applied to pixel event records if not already applied. At block 653, a determination can be made whether an instant energy deposition pattern is isotropic. If so, the method can follow the Y branch from block 653 to return block 663, where the method can return an identification of alpha particle for the instant energy deposition pattern.

[0134] If not isotropic, the method can follow the N branch from block 653 to block 655, where a determination is made whether the total amount of energy deposited is within a predetermined range. An upper bound can be set based on the positron energy endpoint for a given radionuclide, e.g. about 905 keV for89Zr. A lower bound can be set based on a noise floor, such as in a range 10-100 keV. If the total energy is within range, the method can follow the Y branch from block 655 to block 657. Otherwise, the method can follow the N branch from block 655 to return block 665, where the method can return an identification of “X”, signifying an event not identified as either alpha or positron.

[0135] At block 657, coincidence can be tested with a stream of gamma events. If a coincident gamma is found, the method can follow the Y branch from block 657 to return block 667, where the method can return an identification of positron for the instant energy deposition pattern. Otherwise, the method can follow the N branch from block 657 to return block 665, described above.

[0136] Numerous variations and extensions can be implemented within scope of the disclosed technologies. In examples, alpha detection can be based on total energy, in addition or alternatively to isotropy; or the alpha detection can distinguish different nuclei based on energy. Determination of total alpha particle energy can include calibration to compensate for energy losses outside the charged particle detector, as discussed in context of FIG. 10 or elsewhere herein. Optionally, alpha particle identification can be conditioned on anticoincidence with a 511 keV gamma ray.

[0137] In further examples, energy test at block 655 can be omitted, and positron identification can be made solely based on coincidence, or based on a combination ofcoincidence and anisotropy. Alternatively, positron identification can be made based on energy and coincidence without consideration of anisotropy.Example Isotropy Determination

[0138] FIG. 7 is a flowchart 700 of an example method for determining isotropy of an energy distribution pattern according to the disclosed technologies. This method can be used to implement decision block 653.

[0139] At block 751, a determination of pixel extent can be made. To illustrate, contributing pixels can be found to span rows R1-R2 and columns C1-C2 of a pixel array. The pixel coordinates Rl, R2, Cl, C2 define a bounding box of an energy deposition pattern. However, for a linear track, the spread AR = R2-R1 and AC = C2-C1 can differ significantly. Accordingly, at block 753, the size of a bounding square can be computed. To illustrate, the area A of the bounding square can be computed in units of square pixels as A = max(AR, AC)2. At block 755, a target pixel count T can be determined as a fraction of A. Because a circle occupies about n / 4 = 78.54% of its bounding square, and a linear beta track may include pixel event records for 10-20% of pixels in the bounding square, the target pixel count T can be set between these values, e.g. in a range 20-75% of A, 30-50% of A, or about 40% of A.

[0140] At block 757, the actual number of pixels P having energy above a threshold can be determined. Then, at block 759, P and T can be compared. If P > T, then the method can follow the Y branch from block 759 to return block 761, terminating with a result “isotropic”. Otherwise, the method can follow the N branch from block 759 to return block 763, terminating with a result “anisotropic”.

[0141] FIG. 8 A is an image 801 illustrating exemplary spatial signatures 811 of beta particles emitted by a 1 kBq18F positron source. Signatures 811 are seen to comprise generally piecewise linear segments. FIG. 8B is an image 802 illustrating exemplary spatial signatures 821 of alpha particles emitted by a 100 Bq 227THsource; these tracks are seen to be generally isotropic. Images 801-802 are aggregated over an exposure time of 60 s, and the lower left comers of both images 801-802 show spatial pileup of numerous individual tracks. However, detectors according to the disclosed technology can discriminate particle decay times with <100 ns resolution, and individual particle tracks can be discriminated temporally even for MBq sources.

[0142] Numerous variations and extensions can be implemented within scope of the disclosed technologies. In examples, a smoothing filter can be applied to an energy distribution pattern before, or as part of block 757. Alternative techniques can also be used,such as fitting the pattern of contributing pixels to a two-dimensional shape (e.g. a circle), or fitting intensities of reporting pixels to a function over such a 2-D shape, such as a Gaussian, cosine, or another peaked function.Third Example Method

[0143] FIG. 9 is a flowchart 900 of a second example method. In this method, two autoradiographs of distinct radionuclides are concurrently obtained from one sample. Correlation between the autoradiographs are used to establish that an image of one radionuclide can serve to inform spatial distribution of the other.

[0144] A biological sample can be infused with a positron-emitting radionuclide and an alpha-emitting radionuclide. At process block 910, this biological sample is mounted proximate to a spatially resolving charged particle detector and a gamma-ray detector. As one example, an arrangement similar to that of sample 202 and detectors 210, 220 can be used. An objective of the method can be to determine correlation between respective images of the two radionuclides.

[0145] At block 920, spatial signatures of first signals received from the charged particle detector (similar to 220) can be analyzed to identify alpha particles. Example characteristics of alpha particles include generally isotropic energy distribution and total energy in a window commensurate with known alpha particle emission energy (or, energies) of the alpha-emitting radionuclides. One or more such characteristics can be used to distinguish first signals of alpha particles from e.g. beta particles or gamma rays.

[0146] Positions of the identified alpha particles can be determined from their spatial signatures, and energies of the alpha particles can be determined from amplitudes of the first signals. The first signals can also indicate arrival times for energy deposition events, including both alpha particles and beta particles.

[0147] At block 930, digital timestamps of second signals received from the gamma-ray detector can be compared with digital timestamps of some first signals, and coincidences can be identified as positron events. Thus, each such coincidence can include one first signal, and the spatial signature of this first signal can be used to determine a position for the corresponding positron event.

[0148] At block 940, an autoradiograph of the alpha-emitting radionuclide can be formed based on positions, and optionally energies, of the identified alpha particles. At block 950, an autoradiograph of the positron-emitting radionuclide can be formed based on positions of the identified positron events.

[0149] Then, at block 960, intensities of the two autoradiographs can be correlated over corresponding positions. Varying granularity can be used. That is, intensities can be correlated on a pixel-by-pixel basis, or in groups of multiple pixels. The pixel grouping can be uniform over the autoradiographs or can vary according to underlying anatomical features or autoradiograph intensity. To illustrate, finer granularity can be used in areas of therapeutic or diagnostic interest, or where high intensity indicates a concentration of one or both radionuclides. Coarser granularity can be used in areas lacking features of interest, or where low intensity indicates that little of a radionuclide is present. Coarser granularity can mitigate the effects of statistical fluctuations on measured correlation.

[0150] The correlation analysis at block 960 can generate a quantitative measure of the correlation, which can be one or more of: a correlation coefficient, a level of statistical significance, or a parameter measured on a scatter plot of the two autoradiographs’ pixel intensities. At block 970, such a measure of the correlation can be outputted, e.g. to a device similar to 250 of FIG. 2.

[0151] Numerous variations and extensions can be implemented within scope of the disclosed technologies. The method can be used with a wide range of radionuclides. Nonlimiting examples of the alpha-emitting radionuclide include211At,212Bi,223Ra,224Ra,225Ac, or227Th. Non-limiting examples of the positron-emitting radionuclide includenC,13N,15O,18F,62CU,64CU,68Ga,76Br,82Rb,86Y,89Zr, or124I. An infusion used to prepare the biological sample can include both the alpha-emitting and the positron-emitting radionuclides bound to a same ligand species, whereby both radionuclides can be expected to have similar or proportional uptake at receptor sites. The method can be applied to biopsy samples taken from human patients, or can be applied to organs or other tissues obtained from a sacrificed animal.

[0152] The analyzing at block 930 can include determining a measure of isotropy of an energy deposition pattern and identifying as alpha particles those spatial signatures having the measure of isotropy above a predetermined threshold. Further, the analyzing at block 930 can include separating disjoint regions of coincident energy deposition in the charged particle detector as distinct first signals. In this way, positions of identified alpha particles or beta particles can be accurately determined, without being skewed by other coincident energy deposition.

[0153] With regard to gamma-ray detection, the method can extend to digitizing analog data originating from a scintillator and time- tagging the digitized data to obtain the second signals. Data originating from a scintillator can be filtered in the analog or digital domain to retain, asthe second signals, gamma-ray detection events having an energy characteristic of positron annihilation, and to discard other gamma-ray detection events.

[0154] Still further extensions or variations can be applied to the method of FIG. 9, including those disclosed in context of FIG. 1, FIG. 2, or elsewhere herein.Example Alpha Decay

[0155] FIGS. 10A-10B illustrate example energy spectra of alpha particles arising from radioactive decay of alpha-emitting radionuclide227Th. FIG. 10A is a diagram 1001 showing predominant paths in the decay chain of227 R1, terminating in stable isotope207Pb. FIG. 10B is a chart 1002 showing two overlaid energy spectra for the alpha particles emitted throughout this decay chain. A first spectrum, shown as dotted line, comprises peaks 1041-1045 and low energy events 1046, and was obtained by placing a227Th sample directly atop an alpha spectrometer. The second spectrum, shown as solid line, comprises corresponding peaks 1051-1055 and low energy events 1056, and was obtained by placing a mylar film of 12 pm thickness, between the227Th sample and the alpha spectrometer. Disclosed examples advantageously use such a mylar film to protect equipment from contamination by the sample.

[0156] Peaks 1041, 1051 represent the 7.4 MeV alpha particles emitted by215Po. Peaks 1042, 1052 represent the 6.8 MeV alpha particles emitted by219Rn. Peaks 1043, 1053 represent the 6.6 MeV alpha particles emitted by211Bi. Main peaks 1044, 1054 represent the 5.9 MeV alpha particles emitted by227Th, and main peaks 1045, 1055 represent the 5.7 MeV alpha particles emitted by223Ra. These decays are also shown in FIG. 10A. Because the spatial distribution of daughter products can vary significantly from that of an infused parent species, alpha particle images can be filtered by energy, to obtain spatial distribution images of individual species. This can be advantageous for applications in which delivered radiotherapy dosage is dominated by the alpha decays of daughter species.

[0157] Low energy events, 1046 without mylar and 1056 with mylar are indicative of beta particles, including those directly emitted by211Pb and2O7T1 nuclei, as well as energetic secondary electrons knocked out of atoms outside the alpha spectrometer. The alpha spectrometer has a cutoff 1023 at about 280 keV.

[0158] Comparing the energy spectra with and without mylar, it can be seen that energy reported by an internal calibration of a charged particle detector can under-report alpha particle energy by about 20%. For this reason, examples of the disclosed technology apply an energy calibration to the reported energies to achieve accurate dosimetry.Example Positron Decay

[0159] FIGS. 11A- 1 ID illustrate example energy spectra of beta particles and gamma rays arising from radioactive decay of positron-emitting radionuclide89Zr. FIG. 11A is a table indicating the decay of a89Zr nucleus. Of interest herein is the positron decay defined on line 1111 , a positron and neutrino are emitted, leaving an yttrium nucleus89Y in an excited state, denoted89Y*. The neutrino has negligible interaction with either detector. The positrons e+ can be detected by a charged particle detector similar to 210 of FIG. 2. FIG. 1 IB is a chart 1102 showing energy spectrum 1121 of detected positrons on a logarithmic scale. The positron energy spectrum 1121 is in agreement with the known endpoint 905 keV for89Zr beta decay, marked by arrow 1123.

[0160] 89Zr exhibits two radioactive decay paths. As shown in line 1111, the positron decay occurs in about 22.8% of decays. Line 1112 shows the preferred decay path: electron capture (from a surrounding atom) which occurs in about 77.2% of decays. No beta particle is emitted, and the resulting nucleus is in the same excited state89Y* as for line 1111.

[0161] As indicated on line 1113,89Y* predominantly decays by emission of a 909 keV gamma ray, with a half-life of about 15.7 seconds. Additionally, the positrons emitted on line 1111 travel through nearby matter and annihilate in proximity of an electron, causing back- to-back emission of two 511 keV gamma rays.

[0162] FIG. 11C is a chart 1103 showing a spectrum of detected gamma rays, including a peak 1133 representing positron annihilation gamma rays and a peak 1135 representing the gamma decay of89Y* on line 1113. Other gamma ray counts outside peaks 1133, 1135 arise from various sources such as: Compton scattering of a 511 keV or 909 keV gamma ray, whereby only a portion of the gamma ray energy is deposited within a scintillator; X-ray emission following the electron capture on line 1112; or cosmic radiation.

[0163] Some of the gamma rays counted in FIG. 11C can be coincident with a beta particle. FIG. 1 ID is a chart showing the energy spectrum 1141 of these gamma rays, e.g. after applying a coincidence filter. The coincidence filter substantially eliminates most gamma rays in energy window 1149 (above 511 keV peak 1143), including 909 keV peak 1135, and also reduces other gamma rays relative to 511 keV peak 1143. The gamma rays to the left of peak 1143 can be dominated by Compton scattered annihilation gamma rays.

[0164] Peak 1143 is significantly reduced compared to peak 1133, indicating that a significant proportion of emitted positrons do not deposit significant energy in a charged particle detector. To illustrate, some positrons can be emitted away from the charged particle detector; other positrons can be annihilated before reaching the charged particle detector;further positrons can pass through the relatively thin charged particle detector without depositing enough energy to be detected as a beta particle.

[0165] In varying examples, gamma-ray events can be filtered by coincidence (as shown in FIG. 11D), by energy, or both. For example, an energy filter 1147 (shown as about 480-600 keV) can be applied. Alternatively, the energy filter can also encompass range 1149, because the coincidence filter is effective in eliminating high energy gamma rays in range 1149. In some examples, an energy filter of about 440 keV to 2000 keV (the upper limit of pulse height analysis electronics) can be used.Example Coincidence Processing

[0166] FIG. 12 is a chart 1200 illustrating coincidence processing of events from a charged particle detector and a gamma-ray detector. Two streams of event timestamps, one from each detector, are compared with reference to a predetermined coincidence window. Pairs of events whose timestamps differ by less than the width of the coincidence window can be identified as coincident.

[0167] In FIG. 12, two streams of events 1210, 1220 are shown, from a charged particle detector (similar to 210) and a gamma-ray detector (220) respectively. These event streams can be similar to signals {SI }, { S2 } of FIG. 2, respectively. Time increases along the horizontal axes. A marker for each event 1211-1218, 1221-1226 is positioned at the time coordinate of the event’ s timestamp. The timestamp differences are marked as At for a few pairs of events.

[0168] Coincidence processing can be performed by placing the timestamps of each event stream on a respective stack. The heads (earliest timestamps) of each stack can be compared, and a determination can be made whether the timestamps should be identified as coincident. Suitable action can be taken according to this determination, and the earlier of the two heads can be popped off its stack. The process can be repeated until one stream’s stack is empty.

[0169] To illustrate, events 1211, 1221 can be compared. The time separation At can exceed the coincidence window W 1205, meaning that events 1211, 1221 are not coincident.Because event 1221 precedes 1211, 1221 is popped off its stack, and events 1211, 1222 are compared next. Again, these events are not coincident, 1211 is popped, and 1222, 1212 are compared. The event stream processing proceeds, without finding coincidences, until events 1223, 1214 are compared. Their difference At is less than W, and this pair of events is found to be coincident, e.g. at block 225 of FIG. 2.

[0170] Popping 1223 off its stack, events 1214, 1224 can be compared next, and found not coincident. In a variation, determination of coincidence (1223, 1214) can lead to both events being popped off their stacks, so that a given event can be part of at most one coincidence.

[0171] In either case, processing continues until another coincidence is found between 1216, 1225. No further coincidences are present in this example, and eventually events 1226, 1218 are compared. When event 1226 is popped off its stack, the stack for stream 1220 is empty, and coincidence processing is complete.

[0172] Notably, gamma ray events 1221, 1222, 1224, and 1226 are not coincident with any charged particle and can be identified as anti-coincident with the stream of charged particle events. See area 152 of FIG. 1. Similarly, charged particle events 1211-1213, 1215, and 1217-1218 are not coincident with any gamma ray and can be identified as anti-coincident with the stream of gamma ray events. See area 156 of FIG. 1.Example Autoradiographs

[0173] FIGS. 13A-13B are black-and-white and color views of a set of example autoradiographs illustrating application of the disclosed technologies to a test coupon. In particular, this figure illustrates particle identification and selectivity. Four droplets infused with positron-emitting radionuclide89Zr were placed on a glass slide, together with another four droplets infused with alpha-emitting radionuclide227Th, forming the test coupon. Unlike some other examples herein, the two radionuclides occupy separate regions of the test coupon, so that imaging of each radionuclide can be readily distinguished.

[0174] Images 1311-1313 are different views of a first autoradiograph, in which all detected charged particles were imaged. In image 1311, regions 1301-1304 correspond to the89Zr droplets, while regions 1305-1308 correspond to the227Th regions. Visible droplets in images 1312-1313, 1321-1323, or 1331-1333 correspond to droplets 1301-1308 at the same positions.

[0175] Pixel intensities in image 1301 represent energy deposition linearly in a color scale going from black (lowest energy deposition) to white (highest energy deposition). Image 1312 represents the same underlying data as image 1311, however pixel intensities represent the logarithm of the energy deposition, also on a scale going from black (lowest energy deposition) to white (highest energy deposition). Image 1313 is a zoom inset of a portion of image 1312.

[0176] Images 1321-1323 are obtained from events classified as beta particles, while images 1331-1333 are obtained from events classified as alpha particles. Like images 1311-1313, images 1321, 1331 show pixel-by-pixel energy deposition on a linear scale, images 1322,1332 show corresponding data on a logarithmic scale, and images 1323, 1333 are zoom insets as shown.

[0177] Immediately noticeable is that image 1321 emphasizes the positron-emitting droplets 1301-1304, while image 1331 emphasizes the alpha-emitting droplets 1305-1308. The appearance of227Th droplets 1305-1308 in beta image 1321 reflects the fact that, while a227Th nucleus decays by alpha emission, the227Th radionuclide has a complex decay chain including beta emitters211Pb and2O7T1. Thus, image 1321 includes the beta emissions which are secondary decays from227Th droplets. The logarithmic scale of images 1312, 1332 makes certain low-level artifacts visible. For example, image 1332 shows images of sources 1301 - 1304 which can be due to a fewS9Zr positron decays being mis-identified as alpha particles. Other artifacts such as shadows of droplets 1305-1307 in images 1312, 1332 can be due to dispersion of gaseous daughter species 219Rn prior to its own decay (see FIG. 10A). Such artifacts are few in number and relatively insignificant, as demonstrated by their absence in images 1311, 1331 which have a linear scale.

[0178] Images 1341-1343 are obtained from events that are in coincidence with gamma-rays whose energy is in a range 400-600 keV. Notably, the requirement of gamma coincidence suppresses the secondary beta emissions from227Th droplets 1305-1308. A few remaining events at these droplet positions may be consistent with random coincidences.

[0179] FIG. 14 is a set of images 1401-1405 illustrating an example application of the disclosed technologies to an animal sample. For this example, a mouse was injected with a mixed saline solution incorporating223RaCl and Na18F as solutes, with223Ra and18F being alpha-emitting and positron-emitting radionuclides respectively. One hour after injection, the mouse was sacrificed and a 10 pm thick cryotome femur sample was prepared on a slide. Image 1401 is a reference brightfield image of the sample acquired with an optical microscope.

[0180] The sample was then mounted similarly to sample 202, proximate to a charged particle detector (similar to 210) and a gamma-ray detector (220), and decay events were acquired concurrently from both radioisotopes and processed as disclosed herein. Image 1402 is an alpha-particle autoradiograph, similar to the alpha image described in context of 130 of FIG. 1, and image 1403 is a positron autoradiograph, similar to the positron image described in context of 150. The observed energy deposition rates within the charged particle detector were observed to be 8.07 MeV / sec / pCi for alpha particles contributing to image 1402 and 66 keV / sec / pCi for positrons contributing to image 1403.

[0181] Images 1404, 1405 are overlays of images 1402, 1403, respectively, on pathology image 1401.

[0182] Generally, it can be seen that autoradiographs of both radionuclei follow the anatomy of the mouse subject and, further, that the two autoradiographs show apparent correlation. That is, anatomical features that are bright in one autoradiograph (e.g. 1411) are also bright in the other radiograph, and likewise for anatomical features that are dim in either autoradiograph (e.g. 1412). In this series of experiments, Ra223 and F18 both act as calcium analogues. The bone creation mechanism can absorb both Ra223 or Fl 8 as new bone structure is formed.Example Correlation

[0183] FIG. 15 is a set of images 1500 illustrating a correlation study performed on a mouse sample. Image 1510 is a brightfield image of the sample acquired with an optical microscope. As in FIG. 14, this mouse was injected with a mixed saline solution incorporating223RaCl and Na18F solutes. The sample thickness was 30 pm and imaging data for alpha particles, beta particles, and coincident gamma particles was collected concurrently, using the disclosed technologies, for one hour. Images 1520, 1530 are concurrently acquired alpha particle and positron images. Images 1540, 1550 are copies of images 1520, 1530 respectively, with grayscale reversed and numerous bounding polygons 1542 marked concurrently in both images 1540, 1550 using an image labeling tool such as Fiji (previously Imagel) developed by the National Institutes of Health, Bethesda, MD. One region 1543 is shown active in the labeling tool.

[0184] FIG. 16 is a chart 1600 showing correlation between positron emission strength and alpha emission strength across the regions 1542 of images 1540, 1550. Each point 1610 represents one of the regions 1542. Pixel amplitudes of positron each autoradiograph 1540, 1550 are summed over a given region 1542, and scaled to the dose strength at the time of injection to obtain the emission strengths shown. Emission strengths are shown as a fraction of the injected dose. Dotted line 1620 shows linear regression on the points 1610. As seen in legend 1630, correlation > 0.99 is observed over more than one order of magnitude emission strength.

[0185] In variations, one or both autoradiographs can be pre-processed so that point spreading is matched for the two autoradiographs. Spatial averaging or pixel resizing can also be applied so that comparisons can be performed at a desired granularity, which can vary across the spatial extent of the autoradiographs. Calibrations can be performed based onknowledge of, e.g., sample thickness, duration of data acquisition, half-life, energy calibration, and detection efficiency.

[0186] Chart 1600 shows a scatter plot distribution 1610 of alpha dosage (plotted along the Y axis) against positron emitter density (plotted along the X axis). Scatter plot 1610 provides a visual indication of correlation between positron emitter density and alpha particle dosage. A correlation coefficient an also be calculated from this data.

[0187] Linear regression can be applied to scatter plot 1610 to obtain a best fit line 1620, the slope of which provides a scale factor K. For patients receiving theragnostic treatment with the same radionuclides and infusion protocol as the instant sample, this scale factor enables a positron image to be converted to a dosage distribution map, which has direct clinical applicability. In examples, the patient’s positron image can first be converted to positron emitter density based on known characteristics of the positron imaging system used. Then, the scale factor can be applied to convert positron density to dosage density. Such localized (or, micro-scale) dosimetry maps enable personalized radiopharmaceutical therapy, fulfilling an unmet need.Example Spectrally Resolved Imaging

[0188] FIGS. 17A-17D illustrate an example of spectrally resolved imaging of a mouse spleen sample infused with89Zr, a beta source. FIG. 17A is a beta energy spectrum 1701 of an89Zr source having two principal beta decay paths: electron capture, with attendant emission of an Auger electron (76.6% of decays) and positron emission with an average energy of 395 keV and an endpoint of 902 keV. Thus, spectrum 1701 is a composite of Auger electrons and positrons. As indicated by bracket 1725, FIG. 17B is an image 1702 of all these detected beta particles.

[0189] The Auger electrons are emitted in a narrow spectrum 1735 having a peak at about 12 keV. FIG. 17C is an image 1701 of beta events within an energy window 5 - 20 keV, including predominantly Auger electrons. Because the Auger electrons have lower energies than most positrons, their deposition tracks are short and remain close to the decaying nucleus. Accordingly, FIG. 17C is sharper than FIG. 17B. Finally, FIG. 17D is an image 1704 complementary to FIG. 17C including, as shown by bracket 1745, all beta particles above 20 keV, which are predominantly positrons having longer tracks (in other words, more spatial spread) than the Auger electrons of FIG. 17C. Generally, the range of beta tracks scales proportionally with beta energy, and increased range adversely impacts spatial resolution of a source radionuclide. FIG. 17D is substantially similar to FIG. 17B.

[0190] Similar principles can be applied to generate spectrally resolved images in other scenarios. For example, alpha emissions from each of multiple nuclides in a decay chain (see e.g. FIG. 10A) can be imaged separately to assess the impact of nuclide migration. Notably, alpha decay of daughter nuclides can deliver a significantly greater radiotherapy dose than the alpha decay of a parent nuclide. In further examples, deconvolution can be applied on an energy spectrum to assess quantitative dose contributions from multiple source nuclides.Additional Examples

[0191] The following are additional examples of the disclosed technologies.

[0192] Example 1 is a method for validating positron imaging for measuring spatial distribution of an alpha-emitting radionuclide in a biological sample, including: mounting the biological sample, infused with a positron-emitting radionuclide and the alpha-emitting radionuclide, proximate to a spatially resolving charged particle detector and to a gamma-ray detector; analyzing spatial signatures of first signals received from the charged particle detector to identify alpha particles; comparing second digital timestamps of second signals received from the gamma-ray detector with first digital timestamps of respective ones of the first signals to identify coincidences as positron events; forming a first autoradiograph of the alpha-emitting radionuclide based on the identified alpha particles; forming a second autoradiograph of the positron-emitting radionuclide based on the identified positron events; correlating intensities of the first and second autoradiographs over corresponding positions; and outputting a measure of the correlation.

[0193] Example 2 includes the subject matter of Example 1 , and further specifies that the alpha-emitting radionuclide is 211At, 212Bi, 223Ra, 224Ra, 225 Ac, or 227Th.

[0194] Example 3 includes the subject matter of any of Examples 1-2, and further specifies that the positron-emitting radionuclide is 11C, 13N, 150, 18F, 62Cu, 64Cu, 68Ga, 76Br, 82Rb, 86Y, 89Zr, or 1241.

[0195] Example 4 includes the subject matter of any of Examples 1-3, and further specifies that the infused biological sample comprises the positron-emitting radionuclide and the alpha-emitting radionuclide bound to a same ligand species.

[0196] Example 5 includes the subject matter of any of Examples 1-4, and further specifies that the biological sample is a biopsy sample from a human patient.

[0197] Example 6 includes the subject matter of any of Examples 1-5, and further specifies that the biological sample comprises an organ tissue sample from a sacrificed animal.

[0198] Example 7 includes the subject matter of any of Examples 1-6, and further specifies that the analyzing comprises determining a measure of isotropy of an energy depositionpattern and identifying as alpha particles those spatial signatures having the measure of isotropy above a predetermined threshold.

[0199] Example 8 includes the subject matter of any of Examples 1-7, and further specifies that the analyzing comprises separating disjoint regions of coincident energy deposition in the charged particle detector as distinct first signals.

[0200] Example 9 includes the subject matter of any of Examples 1-8, and further includes digitizing analog data originating from a scintillator and time-tagging the digitized data to obtain the second signals.

[0201] Example 10 includes the subject matter of any of Examples 1-9, and further includes filtering data, originating from a scintillator, in an analog or digital domain to retain, for the second signals, gamma-ray detection events having an energy characteristic of positron annihilation, and to discard other gamma-ray detection events.

[0202] Example 11 is a method comprising: acquiring spatially resolved first signals of radioactive decay products from a first detector, the radioactive decay products comprising: beta particles; first alpha particles coincident with respective ones of the beta particles; and second alpha particles anti-coincident with the beta particles; separating the first signals of the first alpha particles from the respective coincident beta particles; acquiring second signals of positron-annihilation gamma rays from a second detector; forming a first image based on the first signals of the first alpha particles and the first signals of the second alpha particles; and forming a second image of positrons based on the first signals of the beta particles which are coincident with the second signals.

[0203] Example 12 includes the subject matter of Example 11, and further includes: determining measures of isotropy of energy deposition patterns in the first signals; and identifying some of the first signals as the first or second alpha particles based on the respective measures of isotropy being greater than or equal to a predetermined threshold.

[0204] Example 13 includes the subject matter of Example 12, and further specifies that the predetermined threshold is a first threshold, and further includes: identifying some of the first signals as the beta particles based on: the respective measures of isotropy being less than or equal to a second threshold; or determining that an energy measure of the respective energy deposition pattern lies within a predetermined range.

[0205] Example 14 includes the subject matter of any of Examples 11-13, and further specifies that the separating further comprises: separating disjoint regions of coincident energy deposition in the first signals; and independently processing each of the disjointregions to identify respective regions containing one of the first alpha particles and the respective coincident beta particle.

[0206] Example 15 includes the subject matter of any of Examples 11-14, and further includes: determining kinetic energies of the first and second alpha particles from the corresponding first signals.

[0207] Example 16 includes the subject matter of Example 15, and further specifies that the first signals comprise pixel event records from the first detector and the determining comprises: summing respective energies from each of a plurality of pixels.

[0208] Example 17 includes the subject matter of Example 16, and further specifies that the pixel event record for each of the pixels comprises a respective amplitude, and the method further comprises: applying a calibration to the respective amplitude to determine the respective energy.

[0209] Example 18 includes the subject matter of Example 17, and further specifies that the calibration is nonlinear.

[0210] Example 19 includes the subject matter of any of Examples 11-18, and further includes: digitizing scintillation signals; and applying digital timestamps to the digitized scintillation signals to produce the second signals.

[0211] Example 20 includes the subject matter of Example 19, and further includes: comparing the digital timestamps of the second signals with digital timestamps of the first signals to determine coincidences between tracks of the beta particles which are coincident with the second signals.

[0212] Example 21 includes the subject matter of any of Examples 11-20, and further includes: applying first and second clock signals, derived from a common clock source, to the first and second detectors respectively.

[0213] Example 22 includes the subject matter of any of Examples 11-21, and further includes: analyzing correlation between intensities of the first and second images at corresponding positions.

[0214] Example 23 is one or more computer-readable media storing instructions which, when executed by one or more hardware processors, cause the one or more hardware processors to perform the method of any of Examples 11-22.

[0215] Example 24. An apparatus, including: a spatially resolving charged particle detector; a gamma-ray detector; a clock source coupled to provide synchronized clock signals to the charged particle detector and the gamma-ray detector; and one or more hardware processors with memory coupled thereto; and computer-readable media storing instructions which, whenexecuted by the one or more hardware processors, cause the one or more hardware processors to perform operations comprising: acquiring first signals from the charged particle detector; among the first signals, separating signals of coincident particles; analyzing spatial energy deposition patterns of the first signals to identify (i) first alpha particles among the coincident particles, and (ii) second alpha particles anti-coincident with other first signals; forming a first image based at least partly on respective positions of the first and second alpha particles; from the gamma-ray detector, acquiring second signals comprising digital timestamps; determining, as positron events, coincidences between the digital timestamps of the second signals and digital timestamps of the first signals; forming a second image based on positions of the positron events; and outputting the first and second images.

[0216] Example 25 includes the subject matter of Example 24, and further specifies that the charged particle detector comprises: a planar semiconductor sensor; and a readout coupled to the planar semiconductor sensor and configured to provide, as the first signals, arrival time and collected charge for respective pixels; wherein the arrival time is according to the clock signal provided to the charged particle detector.

[0217] Example 26 includes the subject matter of Example 25, and further specifies that the pixels form an array whose length and width dimensions are in a range Example 0.2 cm to 20 cm.

[0218] Example 27 includes the subject matter of any of Examples 25-26, and further specifies that each of the pixels has a spatial extent in a range 10 pm to 1 mm.

[0219] Example 28 includes the subject matter of any of Examples 24-27, and further specifies that the gamma-ray detector comprises: a scintillator; a bias voltage supply; a photomultiplier coupled to receive photons from the scintillator and a bias voltage from the bias voltage supply, and configured to output electronic signals corresponding to the received photons; and a digitizer coupled to receive the electronic signals from the photomultiplier and the clock signal provided to the gamma-ray detector, and configured to generate timestamped digital packets corresponding to the received photons.

[0220] Example 29 includes the subject matter of Example 28, and further specifies that the scintillator comprises BGO, Lil, Nal, ZnS, CeBr3, a mixed cerium halide, a cerium-doped halide, or a plastic scintillator.

[0221] Example 30 includes the subject matter of any of Examples 28-29, and further specifies that the photomultiplier comprises: a semiconductor photomultiplier, a photomultiplier tube, or a microchannel plate.

[0222] Example 31 includes the subject matter of any of Examples 28-30, and further specifies that the digitizer is further configured to apply an energy filter to the received electronic signals to: generate the timestamped digital packets for events having energy within a predetermined range; and discard events having energy outside the predetermined range.

[0223] Example 32 includes the subject matter of any of Examples 24-31, and further specifies that the separating signals is based disjoint groups of pixels having amplitudes above a first threshold separated by at least a predetermined number of other pixels having amplitudes below the first threshold.

[0224] Example 33 includes the subject matter of any of Examples 24-32, and further specifies that the identifying first or second alpha particles is based on a shape or a size of the respective energy deposition patterns.

[0225] Example 34 includes the subject matter of any of Examples 24-33, and further specifies that the operations further comprise: determining a respective energy for each of the first and second alpha particles; wherein the forming the first image is also based at least partly on the respective energies of the first and second alpha particles.

[0226] Example 35 includes the subject matter of Example 34, and further specifies that the memory or the computer-readable media store a nonlinear energy calibration and the determining the energy of the first or second alpha particles comprises applying the nonlinear energy calibration to respective pixel amplitudes in the first signals.

[0227] Example 36 includes the subject matter of any of Examples 24-35, and further specifies that the operations further comprise: determining the respective position of each of the first or second alpha particles as a centroid of the respective energy deposition pattern.

[0228] Example 37 includes the subject matter of any of Examples 24-36, and further specifies that the operations further comprise: determining the respective position of each of the positron events as a centroid of an energy deposition pattern of the respective first signal.

[0229] Example 38 includes the subject matter of any of Examples 24-37, and further specifies that the execution of the instructions further causes the one or more hardware processors to: analyze correlation between the first and second images; and output a measure of the correlation.A Generalized Computer Environment

[0230] FIG. 18 illustrates a generalized example of a suitable computing system 1800 in which described examples, techniques, and technologies for autoradiography, including construction, deployment, operation, and maintenance of software, can be implementedaccording to disclosed technologies. The computing system 1800 is not intended to suggest any limitation as to scope of use or functionality of the present disclosure, as the innovations can be implemented in diverse general-purpose or special -purpose computing systems.

[0231] With reference to FIG. 18, computing environment 1810 includes one or more processing units 1822 and memory 1824. In FIG. 18, this basic configuration 1820 is included within a dashed line. Processing unit 1822 executes computer-executable instructions, such as for implementing any of the methods or objects described herein for performing autoradiography, including event detection or processing, image generation or processing, correlation of autoradiographs, or various other architectures, software components, handlers, managers, modules, or services described herein. Processing unit 1822 can be a general-purpose central processing unit (CPU), a processor in an application-specific integrated circuit (ASIC), or any other type of processor. In a multi-processing system, multiple processing units execute computer-executable instructions to increase processing power. Computing environment 1810 can also include a graphics processing unit or coprocessing unit 1830. Tangible memory 1824 can be volatile memory (e.g., registers, cache, or RAM), non-volatile memory (e.g., ROM, EEPROM, or flash memory), or some combination thereof, accessible by processing units 1822, 1830. The memory 1824 stores software 1880 implementing one or more innovations described herein, in the form of computer-executable instructions suitable for execution by the processing unit(s) 1822, 1830. The memory 1824 can also store event data, calibration data, image data, sample data; configuration data, data structures including data tables, working tables, change logs, output structures, data values, indices, or flags, as well as other operational data.

[0232] A computing system 1810 can have additional features, such as one or more of storage 1840, input devices 1850, output devices 1860, or communication ports 1870. An interconnection mechanism (not shown) such as a bus, controller, or network interconnects the hardware components of the computing environment 1810. Typically, operating system software (not shown) provides an operating environment for other software executing in the computing environment 1810, and coordinates activities of the hardware and software components of the computing environment 1810.

[0233] The tangible storage 1840 can be removable or non-removable, and can include magnetic disks, magnetic tapes or cassettes, CD-ROMs, DVDs, or any other medium which can be used to store information in a non-transitory way and which can be accessed withinthe computing environment 1810. The storage 1840 stores instructions of the software 1880 (including instructions and / or data) implementing one or more innovations described herein.

[0234] The input device(s) 1850 can be a mechanical, touch-sensing, or proximity-sensing input device such as a keyboard, mouse, pen, touchscreen, trackball, a voice input device, a scanning device, or another device that provides input to the computing environment 1810. The output device(s) 1860 can be a display, printer, speaker, optical disk writer, or another device that provides output from the computing environment 1810.

[0235] The communication port(s) 1870 enable communication over a communication medium to another computing device. The communication medium conveys information such as computer-executable instructions or other data in a modulated data signal. A modulated data signal is a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal. By way of example, and not limitation, communication media can use an electrical, optical, RF, acoustic, or other carrier.

[0236] In some examples, computer system 1800 can also include a computing cloud 1890 in which instructions implementing all or a portion of the disclosed technologies are executed. Any combination of memory 1824, storage 1840, and computing cloud 1890 can be used to store software instructions or data of the disclosed technologies.

[0237] The present innovations can be described in the general context of computerexecutable instructions, such as those included in program modules, being executed in a computing system on a target real or virtual processor. Generally, program modules or software components include routines, programs, libraries, software objects, classes, data structures, etc. that perform tasks or implement particular abstract data types. The functionality of the program modules can be combined or split between program modules as desired in various embodiments. Computer-executable instructions for program modules can be executed within a local or distributed computing system.

[0238] The terms “system,” “environment,” and “device” are used interchangeably herein. Unless the context clearly indicates otherwise, none of these terms implies any limitation on a type of computing system, computing environment, or computing device. In general, a computing system, computing environment, or computing device can be local or distributed, and can include any combination of special-purpose hardware and / or general-purpose hardware and / or virtualized hardware, together with software implementing the functionality described herein. Virtual processors, virtual hardware, and virtualized devices are ultimatelyembodied in a hardware processor or another form of physical computer hardware, and thus include both software associated with virtualization and underlying hardware.General Considerations

[0239] As used in this disclosure, the singular forms “a,” “an,” and “the” include the plural forms unless the surrounding language clearly dictates otherwise. Additionally, the terms “includes” and “incorporates” mean “comprises.” Further, the terms “coupled” or “attached” encompass mechanical, electrical, magnetic, optical, as well as other practical ways of coupling items together, and do not exclude the presence of intermediate elements between the coupled items. Furthermore, as used herein, the terms “or” and “and / or” mean any one item or combination of items in the phrase.

[0240] Although the operations of some of the disclosed methods are described in a particular, sequential order for convenient presentation, it should be understood that this manner of description encompasses rearrangement, unless a particular ordering is required by specific language set forth below. For example, operations described sequentially can in some cases be rearranged or performed concurrently. Moreover, for the sake of simplicity, the attached figures may not show the various ways in which the disclosed things and methods can be used in conjunction with other things and methods. Additionally, the description sometimes uses terms like “acquire,” “analyze,” “apply,” “assign,” “combine,” “create,” “determine,” “compare,” “correlate,” “digitize,” “discard,” “evaluate,” “execute,” “filter,” “form,” “generate,” “identify,” “obtain,” “output,” “provide,” “receive,” “respond,” “retain,” “return,” “retrieve,” “separate,” “sort,” “store,” “sum,” “transmit,” or “use,” to indicate computer operations in a computer system. These terms denote actual operations that are performed or controlled by a computer. The actual operations that correspond to these terms will vary depending on the particular implementation and are readily discernible by one of ordinary skill in the art.

[0241] Theories of operation, scientific principles, or other theoretical descriptions presented herein in reference to the apparatus or methods of this disclosure have been provided for the purposes of better understanding and are not intended to be limiting in scope. The apparatus and methods in the appended claims are not limited to those apparatus and methods that function in the manner described by such theories of operation.

[0242] In some examples, values, procedures, or apparatus may be referred to as “optimal,” “lowest,” “best,” “maximum,” “extremum,” or the like. It will be appreciated that suchdescriptions are intended to indicate that a selection among a few or among many alternatives can be made, and such selections need not be lower, better, less, or otherwise preferable to other alternatives not considered.

[0243] Any of the disclosed methods can be implemented as computer-executable instructions or a computer program product stored on one or more computer-readable storage media, such as tangible, non-transitory computer-readable storage media, and executed on a computing device (e.g., any available computing device, including tablets, smartphones, or other mobile devices that include computing hardware). Tangible computer-readable storage media are any available tangible media that can be accessed within a computing environment (e.g., one or more optical media discs such as DVD or CD, volatile memory components (such as DRAM or SRAM), or nonvolatile memory components (such as flash memory or hard drives)). By way of example, and with reference to FIG. 18, computer-readable storage media include memory 1824, and storage 1840. The terms computer-readable media or computer-readable storage media do not include signals and carrier waves. In addition, the terms computer-readable media or computer-readable storage media do not include communication ports (e.g., 1870) or communication media.

[0244] Any of the computer-executable instructions for implementing the disclosed techniques as well as any data created and used during implementation of the disclosed embodiments can be stored on one or more computer-readable storage media. The computerexecutable instructions can be part of, for example, a dedicated software application or a software application that is accessed or downloaded via a web browser or other software application (such as a remote computing application). Such software can be executed, for example, on a single local computer (e.g., any suitable commercially available computer) or in a network environment (e.g., via the Internet, a wide-area network, a local-area network, a client-server network, a cloud computing network, or other such network) using one or more network computers.

[0245] For clarity, only certain selected aspects of the software-based implementations are described. Other details that are well known in the art are omitted. For example, it should be understood that the disclosed technologies are not limited to any specific computer language or program. For instance, the disclosed technologies can be implemented by software written in ABAP, Adobe Flash, Angular, C, C++, C#, Curl, Dart, Fortran, Go, lava, lavaScript, lulia, Lisp, Matlab, Octave, Perl, Python, R, Ruby, SAS, SPSS, WebAssembly, any derivatives thereof, or any other suitable programming language, or, in some examples, markuplanguages such as HTML or XML, or in any combination of suitable languages, libraries, and packages. Likewise, the disclosed technologies are not limited to any particular computer or type of hardware. Certain details of suitable computer, hardware, and communication technologies are well known and need not be set forth in detail in this disclosure.

[0246] Furthermore, any of the software-based embodiments (comprising, for example, computer-executable instructions for causing a computer to perform any of the disclosed methods) can be uploaded, downloaded, or remotely accessed through a suitable communication means. Such suitable communication means include, for example, the Internet, the World Wide Web, an intranet, software applications, cable (including fiber optic cable), magnetic communications, electromagnetic communications (including RF, microwave, infrared, and optical communications), electronic communications, or other such communication means.

[0247] The disclosed methods, apparatus, and systems should not be construed as limiting in any way. Instead, the present disclosure is directed toward all novel and nonobvious features and aspects of the various disclosed embodiments, alone and in various combinations and sub-combinations with one another. The disclosed methods, apparatus, and systems are not limited to any specific aspect or feature or combination thereof, nor do the disclosed embodiments require that any one or more specific advantages be present or problems be solved. The technologies from any example can be combined with the technologies described in any one or more of the other examples.

[0248] In view of the many possible embodiments to which the principles of the disclosed technologies may be applied, it should be recognized that the illustrated embodiments are only preferred examples and should not be taken as limiting the scope of the claims. Rather, the scope of the claimed subject matter is defined by the following claims. I therefore claim all that comes within the scope of these claims and their equivalents.

Claims

1. I claim:

1. A method for validating positron imaging for measuring spatial distribution of an alpha-emitting radionuclide in a biological sample, comprising: mounting the biological sample, infused with a positron-emitting radionuclide and the alpha-emitting radionuclide, proximate to a spatially resolving charged particle detector and to a gamma-ray detector; analyzing spatial signatures of first signals received from the charged particle detector to identify alpha particles; comparing second digital timestamps of second signals received from the gamma-ray detector with first digital timestamps of respective ones of the first signals to identify coincidences as positron events; forming a first autoradiograph of the alpha-emitting radionuclide based on the identified alpha particles; forming a second autoradiograph of the positron-emitting radionuclide based on the identified positron events; correlating intensities of the first and second autoradiographs over corresponding positions; and outputting a measure of the correlation.

2. The method of claim 1, wherein the alpha-emitting radionuclide is211At,212Bi,223Ra,224Ra,225Ac, or227Th.

3. The method of claim 1, wherein the positron-emitting radionuclide isnC,nN,1SO,18F,62CU,64CU,68Ga,76Br,82Rb,86Y,89Zr, or124I.

4. The method of claim 1, wherein the infused biological sample comprises the positronemitting radionuclide and the alpha-emitting radionuclide bound to a same ligand species.

5. The method of claim 1, wherein the biological sample is a biopsy sample from a human patient.

6. The method of claim 1, wherein the biological sample comprises an organ tissue sample from a sacrificed animal.

7. The method of claim 1, wherein the analyzing comprises determining a measure of isotropy of an energy deposition pattern and identifying as alpha particles those spatial signatures having the measure of isotropy above a predetermined threshold.

8. The method of claim 1, wherein the analyzing comprises separating disjoint regions of coincident energy deposition in the charged particle detector as distinct first signals.

9. The method of claim 1, further comprising digitizing analog data originating from a scintillator and time-tagging the digitized data to obtain the second signals.

10. The method of any one of claims 1-9, further comprising filtering data, originating from a scintillator, in an analog or digital domain to retain, for the second signals, gamma-ray detection events having an energy characteristic of positron annihilation, and to discard other gamma-ray detection events.

11. A method comprising: acquiring spatially resolved first signals of radioactive decay products from a first detector, the radioactive decay products comprising: beta particles; first alpha particles coincident with respective ones of the beta particles; and second alpha particles anti-coincident with the beta particles; separating the first signals of the first alpha particles from the respective coincident beta particles; acquiring second signals of positron-annihilation gamma rays from a second detector; forming a first image based on the first signals of the first alpha particles and the first signals of the second alpha particles; and forming a second image of positrons based on the first signals of the beta particles which are coincident with the second signals.

12. The method of claim 11, further comprising: determining measures of isotropy of energy deposition patterns in the first signals; and identifying some of the first signals as the first or second alpha particles based on the respective measures of isotropy being greater than or equal to a predetermined threshold.

13. The method of claim 12, wherein the predetermined threshold is a first threshold, and further comprising: identifying some of the first signals as the beta particles based on: the respective measures of isotropy being less than or equal to a second threshold; or determining that an energy measure of the respective energy deposition pattern lies within a predetermined range.

14. The method of claim 11, wherein the separating further comprises: separating disjoint regions of coincident energy deposition in the first signals; and independently processing each of the disjoint regions to identify respective regions containing one of the first alpha particles and the respective coincident beta particle.

15. The method of claim 11, further comprising: determining kinetic energies of the first and second alpha particles from the corresponding first signals.

16. The method of claim 15, wherein the first signals comprise pixel event records from the first detector and the determining comprises: summing respective energies from each of a plurality of pixels.

17. The method of claim 16, wherein the pixel event record for each of the pixels comprises a respective amplitude, and the method further comprises: applying a calibration to the respective amplitude to determine the respective energy.

18. The method of claim 17, wherein the calibration is nonlinear.

19. The method of claim 11, further comprising: digitizing scintillation signals; and applying digital timestamps to the digitized scintillation signals to produce the second signals.

20. The method of claim 19, further comprising: comparing the digital timestamps of the second signals with digital timestamps of the first signals to determine coincidences between tracks of the beta particles which are coincident with the second signals.

21. The method of claim 11, further comprising: applying first and second clock signals, derived from a common clock source, to the first and second detectors respectively.

22. The method of claim 11, further comprising: analyzing correlation between intensities of the first and second images at corresponding positions.

23. One or more computer-readable media storing instructions which, when executed by one or more hardware processors, cause the one or more hardware processors to perform the method of any one of claims 11-22.

24. An apparatus comprising: a spatially resolving charged particle detector; a gamma-ray detector; a clock source coupled to provide synchronized clock signals to the charged particle detector and the gamma-ray detector; one or more hardware processors with memory coupled thereto; and computer-readable media storing instructions which, when executed by the one or more hardware processors, cause the one or more hardware processors to perform operations comprising: acquiring first signals from the charged particle detector; among the first signals, separating signals of coincident particles; analyzing spatial energy deposition patterns of the first signals to identify (i) first alpha particles among the coincident particles, and (ii) second alpha particles anticoincident with other first signals; forming a first image based at least partly on respective positions of the first and second alpha particles;from the gamma-ray detector, acquiring second signals comprising digital timestamps; determining, as positron events, coincidences between the digital timestamps of the second signals and digital timestamps of the first signals; forming a second image based on positions of the positron events; and outputting the first and second images.

25. The apparatus of claim 24, wherein the charged particle detector comprises: a planar semiconductor sensor; and a readout coupled to the planar semiconductor sensor and configured to provide, as the first signals, arrival time and collected charge for respective pixels; wherein the arrival time is according to the clock signal provided to the charged particle detector.

26. The apparatus of claim 25, wherein the pixels form an array whose length and width dimensions are in a range 0.2 cm to 20 cm.

27. The apparatus of claim 25, wherein each of the pixels has a spatial extent in a range 10 pm to 1 mm.

28. The apparatus of any one of claims 24-27, wherein the gamma-ray detector comprises: a scintillator; a bias voltage supply; a photomultiplier coupled to receive photons from the scintillator and a bias voltage from the bias voltage supply, and configured to output electronic signals corresponding to the received photons; and a digitizer coupled to receive the electronic signals from the photomultiplier and the clock signal provided to the gamma-ray detector, and configured to generate timestamped digital packets corresponding to the received photons.

29. The apparatus of claim 28, wherein the scintillator comprises BGO, Lil, Nal, ZnS, CeBr3, a mixed cerium halide, a cerium-doped halide, or a plastic scintillator.

30. The apparatus of claim 28, wherein the photomultiplier comprises:a semiconductor photomultiplier, a photomultiplier tube, or a microchannel plate.

31. The apparatus of claim 28, wherein the digitizer is further configured to apply an energy filter to the received electronic signals to: generate the timestamped digital packets for events having energy within a predetermined range; and discard events having energy outside the predetermined range.

32. The apparatus of claim 24, wherein the separating signals is based disjoint groups of pixels having amplitudes above a first threshold separated by at least a predetermined number of other pixels having amplitudes below the first threshold.

33. The apparatus of claim 24, wherein the identifying first or second alpha particles is based on a shape or a size of the respective energy deposition patterns.

34. The apparatus of claim 24, wherein the operations further comprise: determining a respective energy for each of the first and second alpha particles; wherein the forming the first image is also based at least partly on the respective energies of the first and second alpha particles.

35. The apparatus of claim 34, wherein the memory or the computer-readable media store a nonlinear energy calibration and the determining the energy of the first or second alpha particles comprises applying the nonlinear energy calibration to respective pixel amplitudes in the first signals.

36. The apparatus of claim 24, wherein the operations further comprise: determining the respective position of each of the first or second alpha particles as a centroid of the respective energy deposition pattern.

37. The apparatus of any one of claims 24-27 or 32-36, wherein the operations further comprise: determining the respective position of each of the positron events as a centroid of an energy deposition pattern of the respective first signal.

38. The apparatus of any one of claims 24-27 or 32-36, wherein the execution of the instructions further causes the one or more hardware processors to: analyze correlation between the first and second images; and output a measure of the correlation.

Images