Systems and methods for event-based tissue activity quantification

Abstract

Systems and methods for event-based tissue activity quantification are disclosed comprising obtaining a plurality of images of tissue activity of a tissue over a first time period and having at least two spatial dimensions and a temporal dimension. A response matrix is generated from the plurality of images, each element within the response matrix comprising a time series of tissue activity values determined along the temporal dimension of the plurality of images at a respective spatial location within the at least two spatial dimensions. A plurality of events within the response matrix is identified, wherein each event is associated with a plurality of spatiotemporally contiguous locations within the response matrix at which a predetermined event criterion is met. A first descriptor is generated based on the plurality of events, wherein the first descriptor quantifies tissue activity of the tissue over the first time period.

Description

[0001]PATENT APPLICATION Attorney docket: 33719-59550 PC SYSTEMS AND METHODS FOR EVENT-BASED TISSUE ACTIVITY QUANTIFICATION BACKGROUND Various techniques exist to image cellular activity within a biological tissue (e.g., an engineered human cardiac tissue). Calcium imaging is one such technique for visualizing and measuring changes in intracellular calcium levels within living cells or organisms. Voltage imaging is another such technique for visualizing and measuring changes in the electrical potential (voltage) across cell membranes in living cells or organisms. Both techniques are used to study changes in tissue activity such as the release and uptake of calcium or the change in voltage due to cell excitation (i.e., an action potential). To obtain such images, fluorescent dyes or genetically encoded calcium indicators are applied to the tissue. In the case of calcium imaging, the dyes / indicators exhibit changes in fluorescence intensity or wavelength upon binding to calcium ions such that, when intracellular indicators bind to calcium ions, a change in the fluorescence properties of the indicators occurs. For voltage imaging, the voltage sensitive dye / indicator will change spectral properties in response to voltage changes occurring within the tissue. In either case, a specially configured camera may be used to capture the emitted fluorescence as an image or a sequence of images over time. Integrating the information captured within such tissue activity images into downstream drug discovery and development tasks is challenging due to the variability in the spatial, temporal, and intensity properties of the tissue activities. Therefore, there is a need for robust and discriminative metrics to quantify and compare tissue activity properties to identify explainable differences between tissues under varying conditions. SUMMARY OF DISCLOSURE According to an aspect of the present disclosure there is provided a method for event-based tissue activity quantification. The method comprises obtaining, by one or more processors, a plurality of images of tissue activity of a tissue over a first time period, the plurality of images having at least two spatial dimensions and a temporal dimension, generating, by the one or more processors, a response matrix from the plurality of images, each element within the response matrix comprising a time series of tissue activity values determined along the temporal dimension of the plurality of images at a respective spatial location within the at least two spatial dimensions, identifying, by the one or more processors, a plurality of events within the response matrix, wherein each event of the plurality of events is associated with a plurality of spatiotemporally contiguous locations within the response matrix at which a predetermined event criterion is met, and generating, by the one or more processors, a first descriptor based on the plurality of events, wherein the first descriptor quantifies tissue activity of the tissue over the first time period. According to a further aspect of the present disclosure there is provided a non-transitory computer readable medium storing instructions which, when executed by one or more processors, cause the one or more processors to obtain a plurality of images of tissue activity of a tissue over a first time period, the plurality of images having at least two spatial dimensions and a temporal dimension, generate a response matrix from the plurality of images, each element within the response matrix comprising a time series of tissue activity values determined along the temporal dimension of the plurality of images at a respective spatial location within the at least two spatial dimensions, identify a plurality of events within the response matrix, wherein each event of the plurality of events is associated with a plurality of spatiotemporally contiguous locations within the response matrix at which a predetermined event criterion is met, and generate a first descriptor based on the plurality of events, wherein the first descriptor quantifies tissue activity of the tissue over the first time period. According to an additional aspect of the present disclosure there is provided a device comprising one or more processors and a memory storing instructions which, when executed by the one or more processors, cause the one or more processors to obtain a plurality of images of tissue activity of a tissue over a first time period, the plurality of images having at least two spatial dimensions and a temporal dimension, generate a response matrix from the plurality of images, each element within the response matrix comprising a time series of tissue activity values determined along the temporal dimension of the plurality of images at a respective spatial location within the at least two spatial dimensions, identify a plurality of events within the response matrix, wherein each event of the plurality of events is associated with a plurality of spatiotemporally contiguous locations within the response matrix at which a predetermined event criterion is met, and generate a first descriptor based on the plurality of events, wherein the first descriptor quantifies tissue activity of the tissue over the first time period. According to a further aspect of the present disclosure there is provided a system for event- based tissue activity quantification. The system comprises a bioreactor comprising a device configured for growing tissue and a sensor assembly configured to obtain calcium response images of a tissue within the device. The system further comprises a control unit communicatively coupled to the bioreactor, wherein the control unit comprises one or more processors configured to obtain a plurality of images of tissue activity of a tissue over a first time period, the plurality of images having at least two spatial dimensions and a temporal dimension, generate a response matrix from the plurality of images, each element within the response matrix comprising a time series of tissue activity values determined along the temporal dimension of the plurality of images at a respective spatial location within the at least two spatial dimensions, identify a plurality of events within the response matrix, wherein each event of the plurality of events is associated with a plurality of spatiotemporally contiguous locations within the response matrix at which a predetermined event criterion is met, and generate a first descriptor based on the plurality of events, wherein the first descriptor quantifies tissue activity of the tissue over the first time period. Further aspects and embodiments of the present disclosure are set out in the appended claims. In accordance with the above, and with the disclosure herein, the present disclosure includes applying certain features or aspects with or by use of, a particular machine, e.g., a bioreactor. In various aspects, the bioreactor may comprise a device configured for growing cell culture(s) and / or tissue (e.g., human body tissue such as muscle tissue, cardiac tissue, and / or skeletal muscle tissue). Additionally, or alternatively, bioreactor may comprise a sensor assembly configured to detect calcium response and / or voltage response of the cell culture(s) and / or tissue within the device as one or more fluorescence images. Further, the present disclosure includes effecting a transformation or reduction of a particular article to a different state or thing, e.g., the transformation or reduction of tissue activity (e.g., calcium response or voltage response) of human tissue, e.g., within a bioreactor as sensed by a sensor assembly as fluorescence images, to a different state or thing, e.g., the generation or creation of a quantifiable, comparable, and discriminative descriptor or feature vector of tissue activity. Such descriptors or feature vectors may be used within a number of downstream tasks for disease research, drug discovery and / or drug development. For example, a descriptor of tissue activity for a tissue under a first set of conditions (e.g., a perturbation such as application of a drug or compound) may be utilized as part of a virtual assay to predict an effect (e.g., efficacy, toxicity, mechanism of action, etc.) associated with the first set of conditions. The present disclosure therefore provides improvements to drug discovery / development and disease research by providing an efficient, comparable, and discriminative quantification of tissue activity which may be used to predict the biological activity of a set of conditions without performing physical experiments. The virtual assay may be deployed on an underlying computing device or system, thereby improving the accuracy and prediction in performing drug discovery / development and / or disease research tasks as described herein. In addition, because the descriptors used by the virtual assay are linked to the underlying dynamics of the tissue activity, they may be used to provide a biological or electrochemical explanation to why a set of conditions has been predicted to be associated with a certain effect. This in turn may help to improve understanding within the disease research, drug discovery, or drug development tasks described herein. Still further, the present disclosure includes improvements in computer functionality or in improvements to other technologies at least because the disclosure herein discloses a process for tissue activity quantification from fluorescence images of tissue activity (e.g., calcium response or voltage response) in a cell culture or tissue (e.g., human tissue or artificial / engineered human tissue). The process for tissue activity quantification, when deployed on the underlying system, allows the systems and methods of the present disclosure to execute with fewer iterations, and use fewer computing resources, than prior art related systems and methods. That is, the present disclosure describes improvements in the functioning of the computer itself of “any other technology or technical field” because the increased discriminative power and predictive accuracy provided by the tissue activity descriptors and / or features of the present disclosure allows the underlying computer system to utilize less processing and memory resources compared to prior art systems and methods. This is at least because the descriptors and / or features provide a compact and discriminative quantification of tissue activity without the need for further transformation, analysis, or computer simulations across a wide range of tests using multiple compute cycles and data. Therefore, use of the descriptors and / or features of the present disclosure results in fewer compute cycles, or otherwise iterations, that has less of an impact on the underlying computing device compared to previous prior art systems and methods. Still further, the present disclosure includes specific features other than what is well- understood, routine, conventional activity in the field, and / or otherwise adds unconventional steps that confine the disclosure to a particular useful application, e.g., systems and methods for event-based tissue activity quantification, generating tissue activity descriptors, and / or predicting an effect associated with a set of conditions (e.g., as related to disease research, drug discovery, and / or drug development). Advantages will become more apparent to those of ordinary skill in the art from the following description of the preferred embodiments which have been shown and described by way of illustration. As will be realized, the present embodiments may be capable of other and different embodiments, and their details are capable of modification in various respects. Accordingly, the drawings and description are to be regarded as illustrative in nature and not as restrictive. BRIEF DESCRIPTION OF DRAWINGS Embodiments of the present disclosure will now be described, by way of example only, and with reference to the accompanying drawings, in which: Figure 1A shows images of tissue activity in the form of calcium response of a tissue according to embodiments of the present disclosure; Figure 1B shows images of tissue activity in the form of voltage response of a tissue according to embodiments of the present disclosure; Figure 2 shows a system for event-based tissue activity quantification according to an aspect of the present disclosure; Figure 3 shows a response matrix according to an aspect of the present disclosure; Figure 4 illustrates a local noise reduction operation performed as part of the response matrix generation process according to an embodiment of the present disclosure; Figure 5 illustrates a binary response extraction approach according to an embodiment of the present disclosure; Figure 6 shows example tissue activity images along with corresponding slices within a binary response matrix and a convolved binary response matrix according to an embodiment of the present disclosure; Figure 7 shows a 3-dimensional visualization of a binary response matrix according to an embodiment of the present disclosure; Figure 8 shows a heatmap of tissue activity across a tissue according to an embodiment of the present disclosure; Figure 9 shows a 3-dimensional event matrix generated from the binary response matrix shown in Figure 7 according to an embodiment of the present disclosure; Figure 10 illustrates linking events to form a pattern according to an embodiment of the present disclosure; Figure 11 shows patterns extracted from the event matrix shown in Figure 8 according to an embodiment of the present disclosure; Figure 12 shows patterns and isolated events identified from the patterns shown in Figure 11 according to an embodiment of the present disclosure; Figure 13 shows a plot of the patterns shown in Figure 11 according to an embodiment of the present disclosure; Figure 14 shows a table of descriptors generated according to an embodiment of the present disclosure; Figure 15 shows a method for event-based tissue activity quantification according to an aspect of the present disclosure; Figure 16 shows a method for generating an event matrix according to an embodiment of the present disclosure; Figure 17 shows a method for determining a functional change based on tissue activity descriptors according to an aspect of the present disclosure; Figure 18 shows a bioreactor system according to an aspect of the present disclosure; and Figure 19 shows an example computing system for carrying out the methods of the present disclosure. TECHNICAL FIELD The present disclosure relates to analysis of images of tissue activity. Particularly, but not exclusively, the present disclosure relates to the generation of quantifiable descriptors of tissue activity from fluorescence images of a tissue; more particularly, but not exclusively, the tissue activity is a calcium response or voltage response of the tissue. More particularly again, but not exclusively, the present disclosure relates to the use of quantifiable calcium or voltage response descriptors to identify one or more effects associated with a change in conditions of a tissue. DETAILED DESCRIPTION Tissue activity images, such as calcium response images or voltage response images, allow the activity dynamics of a tissue to be visualized and for functional insights into the activity of the tissue to be determined and used for downstream tasks such as drug discovery and development and / or disease research. Throughout the following disclosure, tissue activity refers to changes occurring within a tissue such as changes in calcium concentration (calcium response) or changes in electrical potential (voltage response). Figure 1A shows images of tissue activity in the form of calcium response of a tissue according to embodiments of the present disclosure. Figure 1A shows a first image 102 of a tissue obtained at a first time point, a second image 104 of the tissue obtained at a second time point, a third image 106 of the tissue obtained at a third time point, and a fourth image 108 of the tissue at a fourth time point. Figure 1A further shows a region 110 of low intracellular calcium activity within the second image 104, a first calcium event 112, and a second calcium event 114. The tissue shown in Figure 1A is an artificial cardiac cell culture (2D tissue). The images shown in Figure 1A illustrate the changes in calcium activity of the tissue over time; particularly, the occurrence of “events” (alternatively referred to as calcium events, intracellular events, or intracellular calcium events) at different time points and spatial regions of the tissue. An event is caused by a change in cellular activity in the tissue, in this case a group of neighboring cells firing an action potential at substantially the same time. Within fluorescence images (such as those shown in Figure 1A), the fluorescence change of a single cell is mediated by a rapid increase of intracellular calcium followed by return to baseline during an action potential. However, the measured fluorescence intensity within a single cell is controlled by the amount of intracellular calcium which can fluctuate resulting in an image having a slightly higher (lighter) background without being an event. Additionally, isolated cells could have a true action potential without other nearby (spatially or temporally) cells firing an action potential. This fluorescent change in an isolated cell would not constitute an event as no action potential is fired by neighboring cells. In the first image 102, little to no calcium events are observed, as indicated by the uniform low intensity of the image. Conversely, the second image 104 shows a single calcium event occurring across the entire field-of-view of the image, and thus across approximately the entire portion of the tissue (cell culture) captured within the second image 104. This is indicated by the lighter intensity of pixels within the second image 104, although there are still regions of calcium inactivity (e.g., the region 110 of low intracellular calcium activity). The third image 106 and the fourth image 108 show smaller, localized, calcium events which occur at specific locations within the tissue (cell culture). For example, the first calcium event 112 within the third image 106 and the second calcium event 114 within the fourth image 108. These local calcium events are shown as lighter regions of the images. Although not shown within Figure 1A, the calcium response dynamically changes with time such that the calcium events shown within Figure 1A correspond to “snapshots” of calcium transients where the concentration of calcium ions rises rapidly and then returns to a baseline level (e.g., due to specific stimuli, such as neurotransmitters, hormones, or mechanical forces). Figure 1B shows images of tissue activity in the form of voltage response of a tissue according to embodiments of the present disclosure. Figure 1B shows a fifth image 116 of a tissue obtained at a first time point, a sixth image 118 of the tissue obtained at a second time point, a seventh image 120 of the tissue obtained at a third time point, and an eighth image 122 of the tissue obtained at a fourth time point. The tissue shown in Figure 1B is an engineered cardiac tissue (3D tissue). The images shown in Figure 1B illustrate the changes in voltage activity of the tissue over time. As in Figure 1A, the images show the occurrence of “events” (in this instance, alternatively referred to as voltage events) at different time points and spatial regions of the tissue. As described above, an event is caused by a change in cellular activity in the tissue (e.g., a group of neighboring cells firing an action potential at substantially the same time). Within the fluorescence images shown in Figure 1B, the fluorescence change of a single cell is mediated by a rapid change in electrical potential (voltage) across the cell membrane followed by a result to baseline. However, not all such changes in fluorescence constitute an event due to factors such as the fluctuation in electrical potential and isolated cells having a true action potential without neighboring cells firing an action potential. The complexity and diversity of events (calcium or voltage events) occurring within the activity of a tissue, along with the spatiotemporal variability of such events, make quantifying the activity dynamics of a tissue based on such events challenging. For example, and as mentioned above, not all fluorescent changes within a sequence of images of tissue activity constitute an event and may instead relate to an isolated cell response, a fluctuation in the amount of intracellular calcium, or a fluctuation in changes in electrical potential across a cell membrane. The present disclosure is directed to event-based quantification of complex tissue activity dynamics (e.g., changes in calcium / voltage response) contained within tissue activity images into useful, compact, comparable, and discriminative metrics (descriptors and / or features). A sequence of images of tissue activity of a tissue is spatiotemporally clustered into events which characterize the local activity dynamics of the tissue over time. Identifying such events may be performed efficiently in parallel and enables a reduction in complexity when analyzing tissue activity. Events may be grouped into higher order structures (patterns) thereby allowing complex behavior of the tissue to be represented by a set of events and / or patterns from which features and / or descriptors may be generated. These patterns provide an efficient and low- dimensional (or low-complexity) representation of the tissue activity at a local scale. The features and / or descriptors generated from such events and / or patterns may be used to encode or represent the activity dynamics of the tissue over time. The approaches of the present disclosure are fully automated thereby allowing them to be integrated into numerous downstream tasks such as within virtual assays used for disease research, drug discovery and drug development. Figure 2 shows a system 200 for event-based tissue activity quantification according to an aspect of the present disclosure. The system 200 comprises a response matrix generator 202, a spatiotemporal clustering module 204, and a descriptor generator 206. The spatiotemporal clustering module 204 comprises an event extractor 208 and optionally comprises a pattern extractor 210. The system 200 optionally comprises a pre-processor 212, a virtual assay 214, and a database 216. A plurality of images 218 of tissue activity of a tissue 220 over a first time period are provided to the response matrix generator 202 to generate a response matrix 222. The event extractor 208 identifies a plurality of events 224 within the response matrix 222 and the descriptor generator 206 generates a first descriptor 226 based on the plurality of events 224. The first descriptor 226 quantifies tissue activity of the tissue 220 over the first time period. The pattern extractor 210 extracts one or more patterns 228 from the plurality of events 224 and the descriptor generator 206 generates the first descriptor 226 based on the plurality of events 224 and / or the one or more patterns 228. The virtual assay 214 compares the first descriptor 226 against descriptors within the database 216 to determine functional changes and / or effects associated with functional changes. In some embodiments, the functional units of the system 200 are implemented as part of a signal processor of a bioreactor system (e.g., the signal processor 1812 of the bioreactor system 1800 shown in Figure 18). As such, the system 200 efficiently extracts discriminative and comparable descriptors of a tissue’s activity dynamics—e.g., the change in calcium or voltage response of the tissue over the first time period. By quantifying the tissue activity of the tissue in such a manner, a deeper understanding may be gained as to the underlying mechanisms of cellular processes and diseases which may be particularly advantageous when integrating tissue activity images and subsequent processing into downstream disease research, drug discovery, and drug development tasks. The plurality of images 218 comprises a sequence of tissue activity images of the tissue 220 over a first time period (or window). The tissue 220 may be a 2D cell culture, an artificial or engineered tissue (3D tissue), or an in vivo tissue. In one embodiment, the tissue 220 is an engineered tissue (or cell culture) held within a bioreactor such as the bioreactor 1802 shown in Figure 18 and described in detail below. The plurality of images 218 comprise fluorescence images of the tissue 220 which capture the calcium response or voltage response of the tissue 220 over the first time period. As such, the plurality of images 218 may be calcium response images which capture the change in calcium response—i.e., calcium dynamics—of the tissue 220 over the first time period. As described in more detail below, calcium response images are obtained by a calcium imaging system after a fluorescent dye or genetically encoded indicator has been applied to the tissue 220. Alternatively, the plurality of images 218 may be voltage response images which capture the change in voltage response—i.e., the electrical or voltage dynamics—of the tissue 220 over the first time period. As described in more detail below, voltage response images are obtained by a voltage imaging system after a voltage sensitive dye or genetically encoded indicator has been applied to the tissue 220. The plurality of images 218 may take the form of a video or sequence of frames which represent the tissue activity of the tissue 220 over the first time period. Each image (frame) within the plurality of images 218 thus comprises a snapshot of the tissue activity of the tissue 220 at a respective time point (as shown in Figures 1A and 1B). In one example implementation, the plurality of images 218 comprises a sequence of 2,000 fluorescence images of tissue activity each of size 1024^^ × 1024^^ × 3. That is, each image has a resolution of 1024^^ × 1024^^ and 3 color channels (e.g., red, green, blue). The frames are obtained at a frame rate of 100^^ (100 frames per second) such that the plurality of images 218 cover a time period of 20^. In such an example implementation, the plurality of images 218 may thus be represented as an array (matrix or tensor) of size^2000, 1024, 1024, 3^. The response matrix generator 202 is configured to generate the response matrix 222 from the plurality of images 218. The response matrix 222 is a matrix comprising at least two spatial dimensions which correspond to at least two spatial dimensions of the plurality of images 218. For example, for a sequence of images having spatial dimensions of 1024^^ × 1024^^, the response matrix generated by the response matrix generator 202 would comprise spatial dimensions 1024 × 1024. Each element within the response matrix 222 comprises a time series of tissue activity values determined from the corresponding spatial location of the plurality of images 218. For example, an element at spatial location ^^, ^^ within a response matrix comprises a time series of tissue activity values determined at spatial location^^, ^^within the plurality of images. For calcium response images, the tissue activity values would correspond to calcium response values whereas for voltage response images the tissue activity values would correspond to voltage response values. As such, the response matrix 222 may be considered a multi-dimensional matrix (or multi-dimensional array, cube matrix, or multi-dimensional tensor) comprising at least two spatial dimensions and a temporal dimension. Figure 3 shows a response matrix 302 generated by a response matrix generator, such as the response matrix generator 202 of the system 200 of Figure 2, according to an aspect of the present disclosure. The response matrix 302 has a first dimension 304, a second dimension 306, and a third dimension 308. The response matrix comprises a first submatrix 310 associated with time point ^^, a second submatrix 312 associated with time point ^^, and a third submatrix 314 associated with time point ^^. Figure 3 further shows a time series 316 of tissue activity values comprising a first element 318, a second element 320, and a third element 322. The response matrix 302 comprises an array, or grid, indexed by spatial location along the first dimension 304 and the second dimension 306. For example, Figure 3 shows a first spatial location ^^, ^^ and a second spatial location ^^, ^^. Each spatial location within the response matrix 302 comprises a time series of tissue activity values determined at a corresponding spatial location of a corresponding plurality of images of tissue activity (e.g., the plurality of images 218 shown in Figure 2). For example, the time series 316 of tissue activity values is determined from a spatial location ^^, ^^ within a plurality of images of tissue activity. The time series 316 of tissue activity values, or vector of tissue activity values, comprises a plurality of fluorescence values which capture the local change in tissue activity of the tissue over time at a spatial location within the plurality of images. The time series 316 of tissue activity values is determined along the temporal dimension of the plurality of images at a respective spatial location—i.e., the spatial location corresponding to ^^, ^^—within the at least two spatial dimensions of the plurality of images. Each time series within the response matrix 302 may thus be considered to capture, or represent, the local tissue activity dynamics—e.g., the local calcium response dynamics or local voltage response dynamics over time—at a respective location within a tissue (e.g., the tissue 220 shown in Figure 2). The individual elements, or values, of each time series are represented as elements, or values, within submatrices of the response matrix 302. As shown in Figure 3, the first element 318 in the first submatrix 310 corresponds to the value at time point of the time series 316, the second element 320 in the second submatrix 312 corresponds to the value at time point ^^of the time series 316, and the third element 322 in the third submatrix 314 corresponds to the value at time point ^^of the time series 316. Each of these elements thus correspond to a tissue activity value (i.e., a fluorescence value corresponding to a calcium response or voltage response) of a tissue at a given spatial location and time point. For example, the first element 318 corresponds to the tissue activity (calcium or voltage response) observed at spatial location ^^, ^^ and time point ^^. Referring once again to Figure 2, in one embodiment the response matrix 222 is generated directly from the plurality of images 218. That is, because the plurality of images 218 comprises a discrete representation of the local change in tissue activity (fluorescence) for each spatial location, the response matrix 222 may directly correspond to the plurality of images 218. In an alternative embodiment, the response matrix generator 202 uses the pre- processor 212 to generate the response matrix 222 by applying one or more pre-processing or filtering operations to the plurality images 218. For example, the pre-processor 212 may be used to apply a normalization operation to the plurality of images 218 such that the response matrix 222 corresponds to a normalized form of the plurality of images 218 (e.g., with each element scaled to the range^0,1^). Additionally, or alternatively, the pre-processor 212 may be used to apply one or more local noise reduction operations to the plurality of images 218. Figure 4 illustrates a local noise reduction operation performed by a pre-processor, such as the pre-processor 212, as part of the response matrix generation process according to an embodiment of the present disclosure. Figure 4 shows a first time series 402 and a second time series 404. The first time series 402 corresponds to the normalized change in tissue activity (the normalized change in fluorescence) of a tissue over time at a specific pixel location (spatial location) within a plurality of images of tissue activity (e.g., the plurality of images 218 shown in Figure 2). The second time series 404 comprises a noise filtered representation of the first time series 402. That is, the second time series 404 corresponds to the first time series 402 having undergone a local noise reduction operation. The local noise reduction operation smooths, or filters, the local tissue activity at the specific pixel location over time. That is, unlike many image processing based noise reduction operations which incorporate spatial information (e.g., blurring operations such as Gaussian filtering), the local noise reduction operation temporally filters the tissue activity at each spatial location. Thus, the local noise reduction operation maintains the spatial relationships between potential events whilst temporally smoothing the local tissue activity thereby reducing the presence of small fluctuations in tissue activity (noise). The local noise reduction operation may be any suitable filtering operation such as a low pass filter. In one embodiment, the local noise reduction operation is a moving average filter with a window size of 10. Beneficially, the moving average filter helps to retain the morphology of the underlying local tissue activity (fluorescence) time series whilst reducing noise-based fluctuations and outliers in the waveform. Referring once again to Figure 2, the response matrix 222 is provided to the spatiotemporal clustering module 204 to identify the plurality of events 224, and optionally the one or more patterns 228, from the response matrix 222. Within the context of the present disclosure, an event (alternatively referred to as a calcium event or voltage event) is associated with a transient change in activity within a specific region of the tissue 220. For example, a transient change in the concentration of calcium ions or a transient change in cell membrane potential. An event is thus associated with a group of cells undergoing a change in calcium concentration or voltage beyond a certain threshold. This change may be due to several factors such as functional events or cellular activity. The threshold is set so as to identify groups of cells which fire an action potential at substantially the same time. Additionally, or alternatively, the threshold is set to identify small changes that did not reach action potentials. In one embodiment, multiple thresholds may be used to identify different groups of activity / events. For instance, a first threshold is set to identify events corresponding to groups of cells firing an action potential at substantially the same time, and a second threshold is used to identify small changes that did not reach action potential (e.g., responses occurring between the first threshold and the second threshold). In this way, a user may set thresholds to define different events and / or activities to be identified. The plurality of events 224 comprises all events identified by the event extractor 208 as occurring within the tissue 220 over the first time period. An event has both a spatial component (because the event is associated with a specific region of the tissue 220—e.g., a specific group of cells) and a temporal component (because events are typically transient). As such, each event within the plurality of events 224 is associated with a plurality of spatiotemporally contiguous locations within the response matrix 222 at which a predetermined event criterion is met. Here, spatiotemporally contiguous locations are locations, or elements, within the response matrix 222 that are connected or adjacent both spatially and temporally. The event extractor 208 identifies the plurality of events 224 by identifying elements within the response matrix 222 which are spatiotemporally connected and at which a predetermined event criterion is met (i.e., two spatiotemporally contiguous locations are connected to the same event only if both locations satisfy a predetermined event criterion). In general, the predetermined event criterion acts as a threshold to determine whether the observed change in tissue activity over a contiguous time period at a given region of the tissue 220 constitutes an event. To generate the plurality of events 224, the event extractor 208 applies a two-step process: first, binary responses—or peaks—are extracted from the response matrix 222; and second, the binary responses are spatiotemporally grouped into the plurality of events 224. The plurality of events 224 may be represented as a matrix of events, or event matrix. Figure 5 illustrates a binary response extraction approach performed by an event extractor (e.g., the event extractor 208 shown in Figure 2) when undertaking the first step of the event identification process, according to an embodiment of the present disclosure. Figure 5 shows a plot 500 of a time series 502 of tissue activity values comprising a plurality of peaks including a first peak 504 occurring at time ^^^and a second peak 506 occurring at time ^^^. In general, the time series 502 comprises the tissue activity at a specific spatial location of a tissue over a time period. More particularly, the time series 502 comprises fluorescence intensity values corresponding to the calcium response of the tissue as recorded within a response matrix generated from a sequence of images of the tissue over the time period (as described in detail above in relation to Figures 3). Whilst the tissue activity shown in Figure 5 corresponds to calcium response, the skilled person will appreciate that the following description and discussion is equally applicable to voltage response. The first peak 504 and the second peak 506 correspond to peaks, or binary responses, within the time series 502 at which a predetermined event criterion is met. Other peaks, or binary responses, within the time series 502 are shown in the plot 500 as filled black circles. At a general level, the peaks identified within the time series 502, such as the first peak 504 and the second peak 506, correspond to elements, or locations, within the response matrix which are identified as being a part of an event such that these peaks may be grouped into events as part of the second step of the event identification process. For example, if the time series 502 is associated with a spatial location ^^, ^^ within a response matrix, then the first peak 504 indicates that at time ^^^, spatial location^^, ^^of the tissue is involved in an event (i.e., a transient change in concentration of calcium ions). Peaks are identified within the time series 502 by identifying locations within the time series 502 at which a predetermined event criterion is met. The predetermined event criterion may alternatively be referred to as a predetermined peak criterion, a predetermined local event criterion, or a predetermined binary response criterion. In one embodiment, the predetermined event criterion is a local maxima criterion. That is, a location, value, or element, within the time series 502 meets the predetermined event criterion if the time series value at that location, or element, is a local maximum. Local maxima are found within the time series 502 by defining a size for a window, or neighborhood, within which local maxima are searched for. The window is then centered at each location, or element, within the time series 502 and if the time series value at the central location of the window is greater than the other values within the window, then it is recorded as a local maximum. In one embodiment, a location or element is only recorded as a local maximum if it meets the above maximum value criterion along with a dead time and prominence criteria. The dead time criterion controls the sensitivity of the local maxima detection by disregarding local maxima occurring with a predetermined dead time period (i.e., the use of the dead time criteria avoids consecutive or closely spaced peaks from being identified as separate maxima). The prominence criteria is used to distinguish significant peaks from noise. The prominence of a local maximum is determined by calculating the difference between the time series value of the local maximum and the maximum time series value of neighboring points (e.g., elements within the window). If the prominence is greater than a predetermined threshold prominence value then the local maximum is retained, otherwise it is disregarded as noise. In one implementation, a window size of 5 is used along with a dead time of 150^^ and a threshold prominence value of 0.2. Additionally, or alternatively, the predetermined event criterion comprises a threshold criterion. That is, a location, or element, within the time series 502 meets the predetermined event criterion if the time series value at that location, or element, exceeds a predetermined threshold value. The predetermined threshold value may be a static value such as 0.5^ , 0.6^ , 0.7^ or the like. Alternatively, the predetermined threshold value may be dynamic such that the threshold is set as a predetermined proportion of the range of time series values within the time series 502 (e.g., for a time series having a range of time series values of ^0.1^ , 0.8^ ^, a predetermined proportion of 80% would result in a predetermined threshold value of , 0.66^ ). The peak, or binary response, identification process described above in relation to Figure 5 is performed for each spatial location within the response matrix (e.g., the response matrix 222 shown in Figure 2). The result of performing the peak identification process is a set of spatiotemporal locations which satisfy a predetermined event criterion—i.e., spatiotemporal locations which are predicted to relate to events. In one embodiment, the set of spatiotemporal locations are represented within a matrix having dimensions equal to those of the response matrix from which the spatiotemporal locations are identified. The matrix (alternatively referred to as a binary response matrix) is binary valued—e.g., a value of 0 at spatiotemporal location^^, ^, ^^indicates that the spatiotemporal location^^, ^, ^^is predicted not to be associated with an event, whilst a value of 1 at spatiotemporal location ^^, ^, ^^ indicates that the spatiotemporal location ^^, ^, ^^ is predicted to be associated with an event. In one embodiment, the binary response matrix is convolved with one or more predetermined kernels to expand each binary response spatially and temporally within the binary response matrix. To clean the binary response matrix (i.e., remove small non-zero elements), a first convolution is performed on the binary response matrix to generate a first intermediate matrix. The binary response matrix is convolved with a ^3,3,3^ kernel of all 1s to generate the first intermediate matrix. Any elements within the first intermediate matrix which have a value < 2 are then zeroed (i.e., only those elements within the first intermediate matrix which have a value >= 2 are retained). To counter possible time shifts in the exact peak time incurred from low-pass filtering, a second convolution is performed on the first intermediate matrix to generate a second intermediate matrix. The second convolution temporally expands the elements within the first intermediate matrix by convolving the first intermediate matrix with a^1,1, (^^kernel of all ones (where the kernel is defined according to the dimensions^^^, ^^, ^^). Here, (^is an expansion factor which controls the extent of temporal expansion performed by the second convolution. The expansion factor (^> 3 and in one implementation (^= 5. A third convolution is then applied to the second intermediate matrix to generate a spatially expanded response matrix. The second intermediate matrix is convolved with a^(^, (^, (^^kernel of all 1s to generate the spatially expanded response matrix. Here, (^is an expansion factor which controls the extent of spatial (and temporal) expansion performed by the third convolution. The expansion factor (^> 3 and in one implementation (^= 5. A threshold, ), is applied to the spatially expanded response matrix to generate a binary spatially expanded response matrix (which is used in the proceeding steps as the binary response matrix). If an element ^^, ^, ^^ within the spatially expanded response matrix has a value ≥ ), then the element ^^, ^, ^^ within the binary spatially expanded response matrix has a value = 1; otherwise, element ^^, ^, ^^ in the binary spatially expanded response matrix has a value = 0. The threshold ) > 5 and in one implementation ) = 10. All of the above described convolution operations are made the same size as the input centered with respect to the “full” output. Alternatively, the above operations regarding spatial expansion of the binary response image may be performed to the event matrix once generated. In such embodiments, the above convolution operations are performed in the same manner but with respect to an event matrix rather than a binary response matrix. Figure 6 shows example calcium response images along with corresponding slices, or sub- matrices, within a binary response matrix and a convolved binary response matrix according to an embodiment of the present disclosure. Figure 6 shows a first image 602 of calcium response of a tissue at a first time point, a second image 604 of calcium response of the tissue at a second time point, a third image 606 of calcium response of the tissue at a third time point, and a fourth image 608 of calcium response of the tissue at a fourth time point. Figure 6 further shows a first binary response image 610, a second binary response image 612, a third binary response image 614, and a fourth binary response image 614 associated with the first image 602, the second image 604, the third image 606, and the fourth image 608 respectively. The first binary response image 610, the second binary response image 612, the third binary response image 614, and the fourth binary response image 616 are obtained from a binary response matrix generated as described above in relation to Figure 5. Figure 6 also shows a first convolved binary response image 618, a second convolved binary response image 620, a third convolved binary response image 622, and a fourth convolved binary response image 624 generated from the first binary response image 610, the second binary response image 612, the third binary response image 614, and the fourth binary response image 616 respectively. As can be seen, the first image 602 and the second image 604 capture field-of-view (FOV) wide events which take up substantially all of the FOV of the image. In contrast, the third image 606 and the fourth image 608 capture smaller localized events occurring at different spatial locations within the images (and thus at different spatial locations within the tissue). The first binary response image 610 corresponds to the slice, or sub-matrix, of the binary response matrix at the first time point, the second binary response image 612 corresponds to the slice, or sub-matrix, of the binary response matrix at the second time point, the third binary response image 614 corresponds to the slice, or sub-matrix, of the binary response matrix at the third time point, and the fourth binary response image 616 corresponds to the slice, or sub- matrix, of the binary response matrix at the fourth time point. The binary response images are binary images such that each light / white pixel within each image corresponds to a spatial location at the respective time point which is predicted to be associated with an event (i.e., a spatial location at which a predetermined event criterion is met). Put another way, each light / white pixel in the binary response images shown in Figure 6 correspond to a peak, or binary response, identified at a respective spatial location at the first and second time points respectively. Figure 7 shows a 3-dimensional visualization of a binary response matrix 702 generated from the first step of the two-step event identification process according to an embodiment of the present disclosure. Figure 7 shows the binary response matrix 702 and a submatrix 704 (or slice) of the binary response matrix 702 associated with a time point, ^+. A first binary response 706 and a second binary response 708 are shown within the submatrix 704. The binary response matrix 702 is generated by an event extractor (e.g., the event extractor 208) from a response matrix as described above. The binary response matrix 702 is visualized in Figure 7 as a 3-dimensional plot comprising two spatial dimensions (width and height in pixels) and a temporal dimension (corresponding to frame number). Binary responses are represented within the binary response matrix 702 as dark gray regions. These binary responses (which may alternatively be referred to as objects, potential events, binary events, connected components, or regions) correspond to elements within the binary response matrix 702—or groups thereof—having a value indicative of a predetermined event criterion being met (e.g., “1” or “True”). Within a biological context, these binary responses correspond to calcium or voltage transients observed within the tissue over the first time period. The binary response matrix 702 shown in Figure 7 is generated from images of a tissue under a specific pacing protocol. The tissue was initially subjected to electrical stimulation at a predetermined pacing frequency (e.g., 1Hz) and then the electrical stimulation was stopped, and the spontaneous response of the tissue observed. This is shown in the binary response matrix 702 from frame 1 through to around frame 1,000 where a sequence of 16 field-of-view (FOV) wide binary responses are identified, followed by a pause, then two further FOV wide binary responses. From around frame 1,000 onwards, after electrical stimulation of the tissue has ceased, and a sequence of smaller, local, binary responses are identified. Because the binary response matrix 702 encodes the distinct spatial response dynamics of the tissue over the time period, activity across the tissue can be observed to visualize regions of high and low activity. Figure 8 shows a heatmap 802 of tissue activity across a tissue determined from the binary response matrix 702 shown in Figure 7 according to an embodiment of the present disclosure. The heatmap 802 shows the tissue activity of the tissue over the observed time period. In the present example, the tissue activity corresponds to calcium response, though the skilled person will appreciate that the discussion relates equally to other activities such as voltage response. Regions of high activity are represented by higher intensity (lighter) pixel values whilst regions of low activity are represented by lower intensity (darker) pixel values. The heatmap 802 captures several small, localized, areas that either have many events or almost no events, even during the stimulation period (i.e., where most of the tissue exhibited event- based activity). The heatmap 802 provides a useful tool for visual assessment to determine if the areas or regions of low activity (fewer events) are all localized in one area and thus may indicate that the area or region was damaged in some way. A heatmap (such as the heatmap 802) may also be used to show that the tissue has structure. For instance, where event heatmaps may be used to identify if events are occurring in specific structural regions of the tissue (e.g., near the edges or only in the center). Whilst a binary response matrix may be used to visualize and identify regions of high or low activity, the binary response matrix does not distinguish between different events. This is shown in Figure 7 where the first binary response 706 and the second binary response 708 within the submatrix 704 are associated with two events occurring at time point ^+but which are spatially distinct (and thus are considered separate events). However, the distinction between these events is not expressly encoded within the binary response matrix 702. To identify / express binary responses as distinct events, the second step of the event identification process (as performed by the event extractor 208 of the system 200 shown in Figure 2) is to spatiotemporally group the binary responses within the binary response matrix 702 into a plurality of events. Referring once again to Figure 2, at the second step of the event identification process, the event extractor 208 segments and labels all binary response by spacetime proximity. As stated above, the output of the first step of the event identification process is a binary response matrix having elements indicating the presence (or absence) of a potential event at a given spatiotemporal location. The event extractor 208 groups, or clusters, these elements into labeled connected components using a connected-component analysis (CCA), or region labeling, algorithm. A connected component comprises a set of elements within a binary response matrix which share the same (or similar) values and are connected to each other through a direct spatiotemporal adjacency as defined according to a connectivity relationship. The connectivity relationship may be one of 6, 18, or 26. For example, a connectivity relationship of 6 limits potential spatial connections to be formed above and below in only the three orthogonal directions (i.e., the two spatial dimensions and the temporal dimension) around a given element whereas a connectivity relationship of 26 allows spatial connections to be formed with any directly adjacent element within the binary response matrix. As previously stated, the event extractor 208 spatiotemporally groups the potential events within the binary response matrix using a CCA, or region labeling, algorithm (i.e., each event, or connected component, within the binary response matrix is assigned a unique identifier or label). Example algorithms include graph traversal algorithms, such as the “one component at a time” algorithm, or two pass algorithms such as the Hoshen–Kopelman algorithm. The result of performing spatiotemporal grouping on the binary response matrix is the plurality of events 224 which may be represented as an event matrix. Figure 9 shows a 3-dimensional visualization of an event matrix 902 generated by spatiotemporally grouping a binary response matrix as part of the second step of the two-step event identification process according to an embodiment of the present disclosure. Figure 9 shows an event matrix 902 comprising a first slice 904, a second slice 906, and a third slice 908. A first event 910 and a second event 912 are shown within the third slice 908. The event matrix 902 is generated from the binary response matrix 702 shown in Figure 7. The event matrix 902 comprises a plurality of events which are each associated with a plurality of spatiotemporally contiguous locations within the response matrix (e.g., the response matrix 222 of the system 200 shown in Figure 2) at which a predetermined event criterion is met. The first slice 904 (or submatrix) and the second slice 906 (or submatrix) each comprise a distinct FOV wide event with the two FOV wide events occurring in succession. Although the two events are substantially spatially congruent—i.e., the two events have a high degree of spatial similarity—they are temporally distinct. That is, there are a number of “empty” slices / submatrices / frames between the first slice 904 and the second slice 906 such that the two events are not temporally joined. These events are observed as a result of the tissue being stimulated at a set pacing frequency for the first portion of the time period covered by the event matrix 902. Each time the tissue is stimulated, a substantially tissue wide response is observed. However, the response is transient because the stimulation occurs periodically such that the events are temporally distinct. The third slice 908 shows a plurality of distinct events occurring at the same time point. In Figure 9, each event is distinguished by a different shading of gray. This is shown by the distinction between the first event 910 and the second event 912. Whilst these events are temporally the same—i.e., they both occur at the same time point—they are spatially distinct. Thus, unlike in a binary response matrix (e.g., the binary response matrix 702 shown in Figure 7), distinct events are identified as such within the event matrix 9 (cf., the first event 910 and the second event 912 and the first binary response 706 and the second binary response 708 in Figure 7). The plurality of events shown within the event matrix 902 of Figure 9 capture the local changes in cellular activity in the tissue. That is, each event relates to a plurality of cells within the tissue which are spatiotemporally grouped due to their spatial proximity within the tissue and their firing action potentials at substantially the same time. As discussed in more detail below, properties of one or more events within the plurality of events may be calculated so as to generate a discriminative and comparable set of features which quantify the event-based tissue activity of the tissue. Referring once again to Figure 2, in some embodiments, the pattern extractor 210 extracts one or more patterns 228 from the plurality of events 224 (which are represented as an event matrix as described above). Each of the one or more patterns 228 comprises one or more spatially related events within the plurality of events 224. As previously stated, an event comprises a plurality of spatiotemporally contiguous locations which are identified as being related (i.e., which satisfy a predetermined event criterion). A pattern is formed of two or more events which are spatially related but may not be temporally contiguous. As such, a pattern is a grouping of spatially related events used to help quantify / encode the activity (e.g., calcium response, voltage response, etc.) over time within a region of the tissue 220. That is, a pattern provides a localized grouping of events which allows the tissue activity at a spatial location of the tissue to be quantified. Put another way, a pattern allows for the local tissue activity to be efficiently represented as a lower-complexity or lower-dimensional representation of the events identified within the pattern (e.g., by generating a descriptor which provides a summary or statistical representation of events within the pattern). By determining tissue behavior locally, changes in the dynamics of tissue activity at a local scale may be efficiently recorded whilst providing a means for quantifying the global activity of the tissue across all patterns. For example, and as will be described in more detail below, patterns may be used to help identify periodic behavior in tissue activity whilst also allowing isolated events to be identified. Patterns thus help to improve the discriminative performance of virtual assays (e.g., the virtual assay 214 shown in Figure 2) by representing the tissue activity of a tissue within higher-order structures whilst reducing the complexity of the descriptors generated for a tissue. The one or more patterns 228 are identified from the plurality of events 224 by linking events based on shared pixel location. That is, if two events are determined to satisfy an event linking criterion (i.e., the events spatially overlap by at least a threshold amount) then the two events are linked to form a pattern or a part of a pattern. Figure 10 illustrates linking events to form a pattern according to an embodiment of the present disclosure. Figure 10 shows a region 1002 comprising a first event 1004, a second event 1006, and a third event 1008. The region 1002 corresponds to a portion of an event matrix projected along the temporal dimension of the event matrix onto two spatial dimensions of the event matrix. That is, Figure 10 shows the spatial, but not temporal, distribution of events within the region 1002 of the event matrix. As such, the first event 1004 and the second event 1006 shown in Figure 10 correspond to temporal projections of a first event and a second event within an event matrix. The first event 1004 may be represented by the set of spatial locations (or pixels), ,, the second event 1006 may be represented by the set of spatial locations, -, and the third event 1008 may be represented by the set of spatial locations, .. Figure 10 further shows various subsets of , and -. The intersection of , and -, , ∩ -, corresponds to the set of spatial locations which are shared by the first event 1004 and the second event 1006. The relative complement of - in ,, , ∖ -, corresponds to the set of spatial locations of the first event 1004 which do not overlap with the second event 1006. Conversely, the relative complement of , in -, - ∖ ,, corresponds to the set of spatial locations of the second event 1006 which do not overlap with the first event 1004. The intersection , ∩ . = ∅ because the first event 1004 and the third event 1008 do not overlap, and the intersection - ∩ . = ∅ because the second event 1006 and the third event 1008 do not overlap. The shared locations between , and - (i.e., the locations which are common to both the first event 1004 and the second event 1006) are represented by , ∩ -, whilst the unshared locations between , and - (i.e., the locations which are in the first event 1004 but not the second event 1006, and the locations which are in the second event 1006 but not the first event 1004) are represented by^, ∖ -^∪^- ∖ 3^which is equivalent to^, ∪ -^∖^, ∩ -^. In general, two events, 4^and 4^, are linked to form a pattern or a portion of a pattern if they satisfy the event linking criterion: Here, 8 ∈^0,1^is a predetermined threshold. The above equation calculates the overlap between the two events as a proportion of the total area covered by the two events. The predetermined threshold 8 may be set at any suitable value. A higher value (e.g., 8 > 0.5) may be used to enforce a greater degree of overlap between linked events and a lower value (e.g., 8 < 0.5) may be used to enforce a lesser degree of overlap between linked events. In one embodiment, 8 < 0.1 and more particularly 8 = 0.01 meaning that a small amount of overlap between events are required for the events to be linked within a pattern. As such, the event linking criterion of Equation (1) provides a mechanism for efficiently controlling the spatial locality requirement for events to be linked as part of a pattern. Different values of the threshold 8 may thus be used to capture the local behavior / activity of a tissue over a range of scales thereby allowing higher-order groupings of tissue activity to be obtained at different scales whilst also allowing control of the degree of complexity reduction achieved by grouping events into patterns. In the example shown in Figure 10, the first event 1004 and the second event 1006 satisfy Equation (1). That is, the number of shared, or overlapping, locations (|, ∩ -|) as a proportion of the total number of shared and unshared locations (|, ∩ -| + |^, ∪ -^ ∖ ^, ∩ -^|) is greater than a predetermined threshold. Because the first event 1004 and the second event 1006 satisfy the event linking criterion they are determined to form a part of the same pattern. Conversely, because neither the first event 1004 and the third event 1008, or the second event 1006 and the third event 1008, satisfy the event linking criterion (i.e., Equation (1)), the third event 1008 is determined not to form a part of the same pattern as the first event 1004 and the second event 1006. Referring once again to Figure 2, the pattern extractor 210 applies the event linking criterion described above in relation to Equation (1) to each possible pairing of events within the plurality of events 224 (i.e., within an event matrix representation of the plurality of events 224). In one embodiment, a space partitioning process is used (e.g., K-d trees, quadtrees, etc.) so that only spatially proximate pairings of events are evaluated thereby avoiding the need to determine whether two significantly spatially distinct events satisfy the event linking criteria. The pattern extractor 210 may repeatedly, or recursively, apply the event linking criterion to join / link patterns. The pattern extractor 210 terminates the process when all events which are not a part of a pattern do not satisfy the event linking criterion. Any events which are not joined into patterns as a result of the operations performed by the pattern extractor 210 are identified as isolated events. Figure 11 shows patterns extracted from the plurality of events represented by the event matrix 902 shown in Figure 9 according to an embodiment of the present disclosure. Figure 11 shows a plurality of patterns 1102 which include a first pattern 1104 and a second pattern 1106. The first pattern 1104 occurs over a time period of approximately 1,000 frames and comprises 18 field-of-view (FOV) wide events. The events within the first pattern 1104, whilst temporally distinct, are linked based on their spatial congruence so as to form a single pattern. The second pattern 1106 occurs over a time period of approximately 750 frames and comprises 3 local events. Although a long period of time elapses between each event, their spatial similarity indicates that the events belong to the same pattern such that they are grouped into the second pattern 1106. The patterns shown within Figure 11 may be alternately visualized by collapsing the temporal component of the patterns and visualizing the spatial distribution of patterns (and isolated events). Figure 12 shows patterns and isolated events identified from the patterns shown in Figure 11 according to an embodiment of the present disclosure. Figure 12 shows a first visualization 1202 of patterns and a second visualization 1204 of isolated events. The first visualization 1202 of patterns comprises a plurality of pattern visualizations including a first pattern visualization 1206 and a second pattern visualization 1208. The second visualization 1204 of isolated events comprises a plurality of isolated event visualizations including a first isolated event visualization 1210 and a second isolated event visualization 1212. The skilled person will appreciate that the pattern visualizations shown in Figure 12 correspond to the pattern boundaries so as to enable overlapping patterns to be viewed. The first visualization 1202 and the second visualization 1204 provide a means of visually assessing and / or validating the distribution, structure, and location of patterns within a tissue. Moreover, they help to assess the extent to which the tissue has a single, homogeneous, field- of-view (FOV) wide pattern. A healthy tissue would be expected to have a single pattern that encompasses the entire FOV (during stimulation), whilst a deviation from this—for instance, two distinct large patterns—indicates a physiological difference. Other physiological differences may be exhibited by a number of small, localized, patterns and / or isolated events (as shown in the second visualization 1204). As will be described in more detail below, the visual features represented in the first visualization 1202 and the second visualization 1204 may be quantified to generate a descriptor (or descriptors) of the tissue. Whilst Figure 12 shows the spatial distribution of patterns, the patterns shown within Figure 11 may be alternately visualized by collapsing the spatial component of the patterns and visualizing the temporal distribution of events within each pattern. Figure 13 shows a plot 1302 of the temporal distribution of events within the patterns shown in Figure 11 according to an embodiment of the present disclosure. The plot 1302 comprises a first pattern 1304, a second pattern 1306, a third pattern 1308, and a fourth pattern 1310. The first pattern 1304 and the second pattern 1306 correspond to the first pattern 1104 and the second pattern 1106 shown in Figure 11 respectively. Figure 13 further shows an expanded view 1306-1 of the second pattern 1306 comprising a first event 1312, a second event 1314, and a third event 1316. The plot 1302 shows the temporal distribution of events within each pattern shown in Figure 11. Grouping events into patterns allows the behavior (e.g., periodicity) of a tissue to be observed. As can be seen, the first pattern 1304 exhibits regular, periodic, behavior—i.e., regularly occurring events—which correspond to the periodic stimulation applied to the tissue between time points and ^^. As will be described in more detail below, the periodicity of events within patterns may be utilized to quantify the activity (e.g., calcium response, voltage response, etc.) of the tissue. Referring once again to Figure 2, the descriptor generator 206 uses the plurality of events 224 and / or the one or more patterns 228 to generate the first descriptor 226 which quantifies the tissue activity of the tissue 220. In general, a descriptor (e.g., the first descriptor 226) is a specific featurization of tissue activity of a tissue in the form of a vector of feature values, a matrix of feature values, or a distribution of feature values. For example, a vector of feature values quantifying the calcium response or the voltage response of the tissue over the first time period. A descriptor is thus a comparable and discriminative quantification of tissue activity which can be used in several downstream disease research, drug discovery, and drug development activities. For example, and as will be described in more detail below, descriptors may be used as part of the virtual assay 214 to predict the activity and / or properties of an unknown compound. A descriptor may be event- based and / or pattern-based; that is, a descriptor may comprise event-based features and / or pattern-based features. An event-based descriptor is generated from the plurality of events 224. A pattern-based descriptor is generated from the plurality of events 224 and / or the one or more patterns 228. As such, the first descriptor 226 may comprise one or more event-based features and / or one or more pattern based features. Event-based descriptors are generated by the descriptor generator 206 based on event features calculated from an event matrix representation of the plurality of events 224. An event feature is a value, or vector of values, which encodes or quantifies the spatiotemporal properties of one or more events within the plurality of events 224. As stated above, an event is associated with a transient change (e.g., in the concentration of calcium ions) within a specific region of a tissue and has both a spatial and temporal component. An event within an event matrix representation of the plurality of events 224 comprises a plurality of spatiotemporally contiguous locations within the event matrix which are identified as being related (i.e., the plurality of spatiotemporally contiguous locations satisfy a predetermined event criterion). As such, event features are generated by the descriptor generator 206 from such groupings of spatiotemporally contiguous locations within the event matrix. Event-based descriptors may include values related to cardinality features such as the total number of events and / or spatial event features such as the average event size, the minimum event size, and the maximum event size. Event-based descriptors may further include values related to temporal event features such as the average event length, the minimum event length, and the maximum event length. Event-based descriptors may further include values related to group level event features such as the number of large events, the number of small events, the ratio of large events to small events, the average size of large events, the average size of medium events, and the average size of small events. Event-based descriptors may further include values related to the homogeneity or heterogeneity of events. Homogeneity or heterogeneity features quantify (or are indicative of) a local similarity or diversity of the plurality of events 224. Homogeneity or heterogeneity features may be spatial or temporal. For calcium imaging embodiments, event-based descriptors may further include values related to relative intracellular concentration indicative of a relative difference in intracellular calcium release. Spatial event features comprise statistical information related to spatial properties of the plurality of events 224 such as the average event size, the minimum event size, and the maximum event size. Here, the size of an event is determined based on the number of elements (e.g., pixels or voxels) within the event. Because each event within the plurality of events 224 may be represented as a connected component within the event matrix, the size of each event may be determined by enumerating the number of elements within each respective connected component. Alternatively, the size of an event may be determined using shape based models. For example, spheres or ellipsoids may be fit around each event and the volume of the fit shape may then be used to determine the event size. Temporal event features comprise statistical information related to temporal properties of the plurality of events 224 such as the average event length, the minimum event length, and the maximum event length. The length of an event is determined by calculating the absolute difference between the first time point at which the event appears within the event matrix and the time point at which the event disappears from the event matrix. Group level event features comprise features such as the number of large events, the number of small events, the ratio of large events to small events, the average size of large events, the average size of medium events, and the average size of small events. Group level event features further include the average length of large events, the average length of medium events, and the average length of small events. The determination of whether an event is small, medium, or large, is based on predetermined thresholds associated with the resolution of the event matrix. For example, if the number of pixels, ^, within an event, :^, is less than a first threshold, )^, then :^is classified as a small event; if ^ is greater than or equal to but less than a second threshold )^, then :^is classified as a medium event; otherwise :^is classified as a large event. In one implementation, = 50px and )^= 200px. Alternatively, the determination is relative such that the sizes of all events are clustered into 3 groups (e.g., using a Gaussian mixture modelling or k means clustering approach) and the mean or median event size within a group is used to determine whether the group corresponds to the small, medium, or large group. Event-based homogeneity or heterogeneity quantify the extent to which events are widespread or localized. A spatial homogeneity feature is calculated from the spatial size of events within the plurality of events 224 and may include the variance of event sizes, the range of event sizes, the coefficient of variation of event sizes, and / or the interquartile range of event sizes. A temporal homogeneity feature is calculated from the temporal length of events within the plurality of events 224 and may include the variance of event lengths, the range of event lengths, the coefficient of variation of event lengths, and / or the interquartile range of event lengths. A spatial heterogeneity feature is calculated from the spatial size of events within the plurality of events 224 and may include the range of event sizes, the entropy of a distribution of spatial sizes, and / or the Gini coefficient of a distribution of spatial sizes. A temporal heterogeneity feature is calculated from the temporal length of events within the plurality of events 224 and may include the range of event lengths, the entropy of a distribution of temporal sizes / lengths, and / or the Gini coefficient of a distribution of temporal sizes / lengths. In embodiments where the events relate to calcium events, an intracellular concentration value feature quantifies relative differences in the amount of intracellular calcium release through changes in fluorescence within events. The intracellular calcium concentration of an event is determined from the intensity of pixel values related to the event within the plurality of images 218. As such, the value for the relative intracellular calcium concentration of the plurality of events 224 may be the average raw intensity of pixels associated with the plurality of events 224 within the plurality of images 218, the z-scored intensity of pixels associated with the plurality of events 224 within the plurality of images 218, and / or the normalized intensity of pixels associated with the plurality of events 224 within the plurality of images 218. As stated above, one or more of the above event features may be used to generate an event- based descriptor comprising a vector of values of the one or more event features. The skilled person will appreciate that an event-based descriptor may comprise any combination of event features described above. For example, the event-based descriptor may comprise a vector containing the total number of events, the average event size, and the average event length thereby quantifying the spatial and temporal tissue activity of a tissue. Additionally, or alternatively, an event-based descriptor comprises a distribution of spatial and / or temporal sizes of events. A distribution of spatial sizes comprises a distribution estimated from the sizes of the events within the plurality of events 224. For example, a histogram of event sizes may be generated and a nonparametric density estimation technique (e.g., kernel density estimation) used to estimate the distribution—i.e., probability distribution—of the event sizes from the histogram. Similarly, a distribution of temporal sizes comprises a distribution estimated from the lengths of the events within the plurality of events 224. A distribution descriptor may thus comprise the 1-dimensional distribution of spatial sizes, the 1-dimensional distribution of temporal sizes, or a 2-dimensional joint distribution of the spatial and temporal sizes. Distribution based descriptors may be compared (e.g., by the virtual assay 214) using a statistical measurement such as the Z-test, the sum of squared differences, or a divergence such as the Kullback-Leibler divergence (relative entropy). Pattern based descriptors are generated by the descriptor generator 206 based on the one or more patterns 228 and / or event features calculated from the plurality of events 224 grouped according to the one or more patterns 228. A pattern based descriptor generated for a first pattern may comprise statistical features calculated from the plurality of events within the first pattern (e.g., number of events, average event size, average event length, etc.). As such, any or all of the event features described above may be applied at a pattern level such that the features are computed from the plurality of events within a pattern rather than all events within the plurality of events 224. The homogeneity or heterogeneity descriptor described above in relation to event features may also be calculated for patterns such that homogeneity or heterogeneity descriptors are calculated from the events within each of the one or more patterns 228. Pattern based descriptors may further include a dysfunction feature indicative of an extent of the one or more patterns 228. The dysfunction feature is determined as a ratio of homogeneity descriptor(s) to heterogeneity descriptor(s). Additionally, or alternatively, a pattern descriptor may comprise a periodicity value or values determined from events within a pattern. The periodicity of a pattern is indicative of the firing rate of events within the pattern. This is illustrated in Figure 13. The periodicity of the second pattern 1306 is determined from the reciprocal of the inter-event-intervals (IEIs) for the second pattern 1306. The IEIs of a pattern correspond to the set of intervals between each adjacent pair of events within the pattern. Figure 13 shows a first interval, =^, between the first event 1312 and the second event 1314, and a second interval, =^, between the second event 1314 and the third event 1316. The IEIs for the second pattern 1306 is thus the set of intervals>=^, =^?. Here, an interval is the absolute difference between the last time point of one event and the first time point of a subsequent event. For example, the first interval comprises the absolute difference between the last time point of the first event 1312 and the first time point of the second event 1314. Each interval is measured in seconds. If the temporal length of a pattern is measured in frames, then the length may be converted to seconds by multiplying the frame length by the reciprocal of the frame rate. For example, a pattern having a length of 250 frames generated from a sequence of images obtained at a frame rate of 50^^ corresponds to a length in seconds of 250 × ^1⁄ 50 ^ = 5^. To determine the periodicity of a pattern, the average (e.g., mean) of the IEIs is computed and the reciprocal taken to calculate the periodicity (i.e., frequency). For example, a pattern having IEIs ^3^, 4^, 3^, 2^^ would have an average IEI of A = 3^ and a periodicity of 0.333^^. Other statistical features may be extracted from the IEIs of a pattern such as the average (e.g., mean) interval length, the standard deviation in IEIs, the minimum interval length, and the maximum interval length. Figure 14 shows a table of descriptors generated according to an embodiment of the present disclosure. The table shown in Figure 14 shows a set of descriptors generated for a first pattern (pattern ID = 1) and a second pattern (pattern ID = 28). The first pattern corresponds to the first pattern 1104 shown in Figure 11. The second pattern corresponds to the second pattern 1106 shown in Figure 11. Referring once again to Figure 2, in one embodiment, the first descriptor 226 is used by the virtual assay 214 to determine, or predict, an activity or property of the tissue 220. More particularly, the virtual assay 214 may determine, or predict, an activity or property of a condition associated with the tissue 220. For example, the tissue 220 may be associated with a condition corresponding to a drug treatment and the virtual assay 214 may predict an activity or property of the drug treatment (e.g., mechanism of action, toxicity, etc.) based on the first descriptor 226. The virtual assay 214 compares the first descriptor 226 to one or more reference descriptors stored within the database 216. Each of the one or more reference descriptors correspond to previously generated descriptors with associated label information. The label information for a previously generated descriptor provides an indication of an activity or property of a condition associated with a tissue from which the previously generated descriptor was generated. For example, the label information may comprise a mechanism of action associated with a drug treatment applied to a tissue prior to the descriptor being generated. The virtual assay 214 may thus compare the first descriptor 226 to the reference descriptors stored within the database 216 to determine one or more similar reference descriptors (alternatively referred to as matching reference descriptors or target reference descriptors). The label information associated with the one or more similar reference descriptors may then be associated with the condition(s) associated with the tissue 220 thereby predicting an activity or property of the condition. The virtual assay 214 may use any suitable approach to compare the first descriptor 226 to a reference descriptor within the database 216. For example, if the descriptors are 1- dimensional vectors of values then a measure such as the Euclidean distance, the Chebyshev distance, or the like may be used to determine the dissimilarity between the two descriptors. If the difference between two descriptors is below a predetermined threshold then the two descriptors are considered similar. The predetermined threshold may be manually set or learnt from the reference descriptors within the database 216. Alternatively, the similarity / dissimilarity between two descriptors may be made on a per-feature basis. For example, an absolute difference may be used to determine the dissimilarity for a number of events feature whilst a Euclidean distance may be used to determine the dissimilarity for an average event size / average event length pairing of features. The total similarity or dissimilarity between the two descriptors comprises a combination (e.g., sum, weighted sum, etc.) of the per-feature measures with a thresholding approach used to determine whether the descriptors are considered similar. As stated above, for distribution based descriptors, the virtual assay 214 may use a statistical measurement such as the Z-test, the sum of squared differences, or a divergence such as the Kullback-Leibler divergence (relative entropy). Alternatively, the virtual assay 214 determines, or predicts, an activity or property based on a relative change in descriptors generated from a tissue under two conditions. For example, a first descriptor may be generated from images of a tissue under reference, or baseline, conditions and a second descriptor generated from images of the tissue under one or more perturbed conditions. In general, reference conditions refer to conditions which provide a baseline comparison to the perturbation conditions. In embodiments, a reference condition is a condition associated with a control setup or environment. A reference condition may correspond to a tissue in its default, natural, or unaltered state (i.e., without dosage of a drug or agent). Alternatively, a reference condition may correspond to vehicle treated tissue. Perturbation conditions refer to conditions in which the tissue has been perturbed in some way. Examples of perturbation conditions include the administration of a drug or compound (i.e., a perturbant), a disease state, a different cell line, a physical perturbance applied to the tissue, or a change to the environment of the tissue. As such, perturbation conditions may alternatively be referred to as treatment conditions. In the case of perturbations involving a drug or compound, the conditions are further associated with an effect related to the drug or compound such as a mechanism of action or a toxicity. Given the range of different perturbation conditions, a tissue may be associated with more than one perturbation condition (e.g., a diseased tissue having been treated with a specific compound). The relative change in the descriptors between the reference and perturbed conditions thus quantifies the change in tissue activity of the tissue as a result of the one or more perturbed conditions. The relative change in descriptors generated by the virtual assay 214 may then be compared to similar relative changes within the database 216 to determine an activity or property associated with such a change. Advantageously, representing the change in conditions as a relative change between descriptors acts as a form of normalization thereby enabling improved comparison to other relative change descriptors by reducing the influence of noise and variability in imaging setups. The description will now turn to methods for utilizing the above described systems and models for event-based calcium response quantification. Figure 15 shows a method 1500 for event-based tissue activity quantification according to an aspect of the present disclosure. The method 1500 comprises the steps of obtaining 1502 a plurality of images, generating 1504 a response matrix from the plurality of images, identifying 1506 a plurality of events within the response matrix, and generating 1508 a descriptor. The method 1500 further comprises the optional steps of extracting 1510 one or more patterns from the plurality of events and outputting 21512 the descriptor. The step of extracting 1510 comprises the optional step of linking 1514 events to form a pattern. In one embodiment, the method 1500 is performed by a system such as the system 200 shown in Figure 2 or a module / unit such as the signal processor 1812 of the bioreactor system 1800 shown in Figure 18. In general, the method 1500 provides an efficient approach for quantifying complex and diverse tissue activity events from within a sequence of images of a tissue. The descriptors generated by the method 1500 are discriminative and comparable representations of the tissue activity (e.g., calcium response or voltage response) of the tissue which may be used for downstream tasks such as use within a virtual assay or providing understanding of the underlying mechanisms of cellular processes and diseases. At the step of obtaining 1502, a plurality of images of tissue activity of a tissue over a first time period are obtained (e.g., the plurality of images 218 shown in Figure 2 which are images of the tissue 220). The plurality of images have at least two spatial dimensions and a temporal dimension. The tissue associated with the plurality of images (e.g., the tissue 220 shown in Figure 2) is associated with a first set of conditions over the first time period. As described in more detail above in relation to the virtual assay 214 of Figure 2, the first set of conditions comprises a reference condition or a perturbation condition such as a drug treatment, a disease model, a cell line, or a physical perturbation, or a change in environmental conditions. The tissue is a 3D engineered or artificial tissue, a 2D cell culture, or an in vivo tissue. In one embodiment, the tissue is a cardiac tissue and may be an engineered cardiac tissue. The plurality of images may be obtained from a bioreactor or device containing the tissue (e.g., from a device such as the bioreactor 1802 shown in Figure 18). Over the first time period, an external stimulation may be applied to the tissue. For example, the tissue may be electrically stimulated at a fixed frequency of 1Hz over the first time period such that the plurality of images captures the tissue activity of the tissue as a result of being stimulated at 1Hz. Alternatively, the tissue may not be stimulated over the first time period. That is, the plurality of images capture the spontaneous response of the tissue in the absence of any external stimulation. Alternatively, a stimulation, or pacing, protocol may be used over the first time period. The stimulation protocol may define sub-periods of the first time period where the stimulation applied to the tissue changes. In one example, the stimulation protocol defines that an external stimulation at a predetermined pacing frequency (e.g., 1Hz, 2Hz, etc.) is applied to the tissue for a first portion of the first time period and is then stopped such that the plurality of images capture the tissue activity of the tissue under stimulation conditions over the first portion of the first time period and the spontaneous tissue activity of the tissue over the second portion of the first time period. As a further example, the stimulation protocol may define an increasing or decreasing ramp of stimulation frequencies (e.g., 1Hz for 10s followed by 2Hz for 20s followed by 3Hz for 10s). As such, in one embodiment the method 1500 further comprises, prior to the step of obtaining 1502, causing stimulation of the tissue according to a predetermined stimulation protocol over the first time period. The stimulation protocol may cause stimulation at a single predetermined pacing frequency or may cause no stimulation. Alternatively, the stimulation protocol may be any of the stimulation protocols described above. In an embodiment, the plurality of images are pre-processed using one or more pre-processing operations including downsampling and / or normalizing. As such, the method 1500 may further comprise the step of downsampling (not shown) the plurality of images by a predetermined factor. The predetermined factor is between 2 and 6 and is preferably 4. The method 1500 may further comprise the step of normalizing (not shown) each image within the plurality of images. In one example, a sequential normalization approach is applied whereby a Z-score normalization is first applied to the plurality of images to account for differences in intensity across the field-of-view and then the intensity values of the (Z-score normalized) plurality of images are scaled to between ^0, 1^ to allow for static thresholding in subsequent operations. At the step of generating 1504, a response matrix is generated from the plurality of images. Each element within the response matrix comprises a time series of tissue activity values determined along the temporal dimension of the plurality of images at a respective spatial location within the at least two spatial dimensions. As shown in Figure 3 above, the response matrix 302 comprises, for each spatial location, a time series of tissue activity values. For example, the response matrix 302 comprises the time series 316 at spatial location ^^, ^^. The time series 316 comprises a plurality of time ordered tissue activity values observed at a spatial location within the plurality of images corresponding to ^^, ^^. In one embodiment, the response matrix is filtered using a local filtering strategy. The local filtering strategy applies a low-pass filter to each element within the response matrix (i.e., to each time series within the response matrix) to provide a temporal filtering of the tissue activity at each spatial location within the response matrix. This is illustrated in Figure 4, and described in more detail above, where the second time series 404 is generated by applying a local reduction operation to the first time series 402. At the step of identifying 1506, a plurality of events are identified within the response matrix. Each event of the plurality of events is associated with a plurality of spatiotemporally contiguous locations within the response matrix at which a predetermined event criterion is met. Within a biological context, each event represents a group of neighboring cells within the tissue firing an action potential at substantially the same time. In general, the predetermined event criterion acts as a threshold to determine whether the observed change in tissue activity over a contiguous time period at a given region—i.e., spatial location—of the tissue constitutes an event (or merely noise). The predetermined event criterion comprises a local maxima criterion or a threshold criterion. A location, value, or element, within a time series meets a local maxima criterion if the time series value at that location, or element, is a local maximum. A location, or element, within a time series meets a threshold criterion if the time series value at that location, or element, exceeds a predetermined threshold value (e.g., 0.5^ , 0.6^ , 0.7^ , etc.). In one embodiment, an event matrix may be generated as part of the step of identifying 1506 the plurality of events (as described in more detail in relation to Figure 16 below). At the step of generating 1508, a first descriptor is generated based on the plurality of events. The first descriptor quantifies tissue activity of the tissue (e.g., calcium response or voltage response) over the first time period. As described in more detail above, the first descriptor may comprise a homogeneity or heterogeneity descriptor indicative of a local similarity or diversity of the plurality of events, or for embodiments related to calcium imaging, the first descriptor may comprise a relative intracellular calcium concentration descriptor indicative of a relative difference in intracellular calcium release. The skilled person will appreciate that at the step of generating 1508, two or more descriptors may be generated based on the plurality of events. For example, a homogeneity descriptor and a relative intracellular calcium concentration descriptor may be generated. At the optional step of outputting 1512, the first descriptor is output. In one embodiment, outputting the first descriptor comprises storing, or saving, the first descriptor to a persistent storage such as a non-volatile memory, a non-transitory medium, or the like. Additionally, or alternatively, outputting the first descriptor comprises transmitting the first descriptor via a network (e.g., a local area network, a wide area network, and the like), or displaying a representation of the first descriptor for review by a user. Additionally, or alternatively, outputting the first descriptor comprises providing the first descriptor to a step of another method or process (e.g., method 1700 described below). After the step of identifying 1506, the method 1500 may comprise the step of extracting 1510 one or more patterns from the plurality of events. At the step of extracting 1510, one or more patterns are extracted from the plurality of events (e.g., the one or more patterns are extracted from an event matrix comprising the plurality events). Each of the one or more patterns comprises one or more spatially related events within the plurality of events. As described in detail above, a pattern provides a low-complexity / low-dimensional representation of a group of locally related events. The step of extracting 1510 may comprise the step of linking 1514 events to form a pattern. More particularly, when a relative spatial overlap between a first event of the plurality of events and a second event of the plurality of events meets a predetermined overlap threshold, the first event and the second event are linked to form a first pattern of the one or more patterns. The relative spatial overlap comprises a ratio of a number of spatial locations shared by the first event and the second event to a total number of spatial locations involving either the first event or the second event. The predetermined overlap threshold is greater than 0% and less than or equal to 5%. More particularly, the predetermined overlap threshold is 1%. In embodiments where one or more patterns are extracted, the first descriptor generated at the step of generating 1508 may be generated based on the one or more patterns. In such embodiments, the first descriptor may comprise a periodicity feature indicative of firing rates of the one or more patterns, a dysfunction feature indicative of an extent of the one or more patterns, or a ratio of a number of events within a first group of patterns to a number of events within a second group of patterns. In one embodiment, the method 1500 further comprises generating one or more graphical representations of the plurality of events and / or the one or more patterns and outputting the one or more graphical representations for display to a user. The one or more graphical representations may include a graphical representation of the plurality of events such as the heatmap 802 shown in Figure 8 which provides a visualization of tissue activity within a tissue over an observed time period. The one or more graphical representations may include a graphical representation of the plurality of events such as the 3-dimensional visualization of the event matrix 902 shown in Figure 9. The one or more graphical representations may include a graphical representation of the one or more patterns such as the 3-dimensional visualization shown in Figure 11. The one or more graphical representations may include spatial or temporal pattern-based visualizations such as those shown in Figures 12 and 13. Advantageously, such graphical representations provide a user with a transformed and reduced complexity representation of the tissue activity of a tissue over time (e.g., the dynamic change in calcium response or voltage response). Such graphical representations may help lead to improvements in downstream disease research or drug discovery / development tasks by aiding in the identification of patterns of activity or regions of normal / abnormal activity within the tissue. The one or more graphical representations may be output for display to a user by causing the one or more graphical representations to be output to a display device (e.g., a monitor, screen, display panel, etc.) or by causing the one or more graphical representations to be persistently stored in a format suitable for graphical display (e.g., as a bitmap image, a raster image such as a portable network graphic (PNG) image, a compressed image such as a Joint Photographic Experts Group (JPEG) image, or the like). Figure 16 shows a method 1600 for generating an event matrix according to an embodiment of the present disclosure. The method 1600 comprises the steps of generating 1602 an event matrix from a plurality of events, applying 1604 one or more noise filtering operations to the event matrix, and convolving 1606 the event matrix with a predetermined kernel. The method 1600 may be performed as part of the method 1500 shown in Figure 15. More particularly, the method 1600 may be performed as part of the operations performed at the step in method 1500 of identifying 1506 a plurality of events. At the of generating 1602, an event matrix is generated from the plurality of events (e.g., the plurality of events 224 shown in Figure 2) such that a first value at a first location within the event matrix is indicative of the first location being associated with an event within the plurality of events. Within the context of the present disclosure, an event (alternatively referred to as a tissue activity event, a calcium event, or a voltage event) is associated with a transient change in (e.g., in the concentration of calcium ions) within a specific region of a tissue (e.g., the tissue 220 shown in Figure 2). An event is considered to have both a spatial component (because the event is associated with a specific region of the tissue—e.g., a specific cell or group of cells) and a temporal component (because events are typically transient). Each event within the plurality of events (e.g., the plurality of events 224 shown in Figure 2) is associated with a plurality of spatiotemporally contiguous locations within a response matrix (e.g., the response matrix 222) at which a predetermined event criterion is met. Here, spatiotemporally contiguous locations are locations, or elements, within the response matrix that are connected or adjacent both spatially and temporally. Therefore, the generated event matrix is of the same size (i.e., dimensionality and shape) as the response matrix from which it is generated. Any non-zero elements within the event matrix identify the event to which the spatiotemporal location of the non-zero element relates. At the step of applying 1604, one or more noise filtering operations are applied to the event matrix. The one or more noise filtering operations include one or more of a convolution operation and a thresholding operation. As described above in relation to Figure 5, in one embodiment the noise filtering operation comprises convolving the event matrix with a ^3,3,3^ kernel of all 1s and zeroing any elements in the resulting matrix which have a value < 2. At the step of convolving 1606, the event matrix is convolved with one or more predetermined kernels thereby spatially and temporally expanding each event within the event matrix. As described in more detail above in relation to Figure 5, in one embodiment the matrix generated at the previous step of applying 1604 is temporally expanded by convolving the matrix with a^1,1, (^^kernel of all ones (where the kernel is defined according to the dimensions ^^^, ^^, ^^) and the resulting matrix is convolved with a ^(^, (^, (^^ kernel of all 1s to generate the spatially expanded event matrix. Preferably, (^> 3 and > 2. In one implementation, (^= 5 and (^= 5. A threshold, ), is then applied to the spatially expanded event matrix such that any elements which have a value less than the threshold ) are zeroed. Preferably, ) > 5 and in one implementation ) = 10. The method 1600 may further comprise, after the step of convolving 1606, re-labeling the events within the event matrix. Such a step reassigns the event identifiers within the original event matrix to the convolved event matrix. This may be achieved by multiplying the (binary) event matrix generated at the step of convolving 1606 with the original event matrix (generated at the step of generating 1602). Alternatively, the event identifiers may be assigned by performing a CCA, or region labeling, algorithm as described above. Figure 17 shows a method 1700 for determining a functional change based on tissue activity descriptors according to an aspect of the present disclosure. The method 1700 comprises the steps of obtaining 1702 a first descriptor, obtaining 1704 a second descriptor, determining 1706 a functional change based on a comparison of the first descriptor and the second descriptor, and determining 1708 an effect based on the functional change. The method 1700 further comprises the optional step of outputting 1710 the functional change and / or effect. At the step of obtaining 1702, a first descriptor is obtained. The first descriptor quantifies tissue activity of a first tissue under a first set of conditions over a first time period. For example, the first descriptor may quantify the calcium response of the first tissue under reference conditions over the first time period. In one embodiment, the first descriptor is obtained according to the method 1500 shown in Figure 15. Additionally, or alternatively, the first descriptor is obtained by loading the first descriptor from a storage location such as a non-transitory computer readable medium. At the step of obtaining 1704, a second descriptor is obtained. The second descriptor is associated with tissue activity of the tissue under a second set of conditions over a second time period. The second set of conditions are different to the first set of conditions such that the first descriptor and the second descriptor may be used to identify / quantity changes in tissue activity of the tissue as a result of the change in conditions. Continuing the previous example, the second descriptor may quantify the calcium response of the first tissue under perturbation conditions (e.g., after having been dosed with a compound or drug treatment) over the second time period. In one embodiment, the second descriptor is obtained according to the method 1500 shown in Figure 15. Additionally, or alternatively, the second descriptor is obtained by loading the first descriptor from a storage location such as a non-transitory computer readable medium. At the step of determining 1706, a functional change in the tissue is determined based on a comparison of the first descriptor and the second descriptor. The functional change is indicative of a change from the first set of conditions to the second set of conditions. That is, the functional change efficiently encodes / represents the change in tissue activity that has occurred as a result of the change in condition(s) of the tissue—e.g., the change in calcium response or voltage response due to the tissue undergoing a drug treatment. The functional change may be a relative change in values between the first descriptor and the second descriptor. For example, if the first descriptor is an average event length descriptor taking the value 3.5s and the second descriptor is an average event length descriptor taking the value 1.8s, then the functional change may be determined as a decrease in average event length by 1.7s. In embodiments where the first descriptor and second descriptor are distributions (e.g., 1- dimensional distributions of size length), then the functional change may be a distribution shift indicative of a change from the first set of conditions to the second set of conditions. For example, if the first descriptor comprises a 1-dimensional Gaussian distribution of pattern periodicities parameterized by ^A^, B^^ and the second descriptor comprises a 1-dimensional Gaussian distribution of pattern periodicities parameterized by then the functional change may be represented by the shift^A^− A^, B^− B^^. Alternatively, the functional change for distributional descriptors may be determined using a statistical measurement such as the Z-test, the sum of squared differences, or a divergence such as the Kullback-Leibler divergence. Advantageously, using the distributional shift as an indicator of change in conditions provides robustness to variability (i.e., the effect of random fluctuations in individual features within a descriptor is minimized), robustness to outliers, and also good generalization. At the step of determining 1708, an effect associated with the change from the first set of conditions to the second set of conditions is determined based on the functional change. Here, the effect is associated with the change in conditions and may, more particularly, be associated with a specific perturbation associated with the second set of conditions. For example, a toxicity associated with a compound, or a mechanism of action associated with a drug. The effect may be determined using unsupervised multivariate analysis. More particularly, the effect may be determined by comparing the functional change to one or more predetermined, or reference, functional changes having known effects. For example, a database of reference functional changes (e.g., the database 216 shown in Figure 2) may be used to find a reference functional change which is similar to the observed functional change determined at the step of determining 1706. Because the reference functional change is stored along with one or more known effects, the observed functional change may be predicted to be associated with these one or more known effects. Therefore, by utilizing descriptors of calcium response functional changes occurring within a tissue due to changes in conditions may be efficiently quantified and used to identify one or more effects associated with the change in conditions. The descriptors thus provide an efficient and effective mechanism for quantifying tissue activity of a tissue which allows for effective and robust comparison. Moreover, because the descriptors are based on the underlying activity of the tissue, descriptors provide explainable insights into the biological or electrochemical reasons functional responses are considered similar. At the optional step of outputting 1710, the functional change and / or effect is output. In one embodiment, outputting the functional change and / or effect comprises storing, or saving, the functional change and / or effect to a persistent storage such as a non-volatile memory, a non-transitory medium, or the like. Additionally, or alternatively, the functional change and / or effect comprises transmitting the functional change and / or effect via a network (e.g., a local area network, a wide area network, and the like), or displaying a representation of the functional change and / or effect for review by a user. Figure 18 shows a bioreactor system 1800 according to embodiments of the present disclosure. The bioreactor system 1800 comprises a bioreactor 1802 and a control unit 1804. The bioreactor 1802 comprises a first device 1806a for growing engineered tissues and / or a second device 1806b for growing cell cultures, a sensor assembly 1808, and an interface 1810. The interface 1810 communicatively couples the bioreactor 1802 and the control unit 1804 such that data may be exchanged between the bioreactor 1802 and the control unit 1804. The control unit 1804 comprises a signal processor 1812 configured to perform various processing functionalities such as those described above in relation to Figure 2 and / or Figures 15-17. In various aspects, the bioreactor is used for growing tissue such as human body tissue including, by way of non-limiting example, muscle tissue, cardiac tissue, skeletal muscle tissue, or other human tissue. As shown in the expanded portion 1806-1 of the first device 1806a, the first device 1806a, or substrate, comprises one or more wells, such as the well 1814, one or more tissue wells, such as the tissue well 1816, a pair of electrodes including a first electrode 1818-1 and a second electrode 1818-2, and a pair of elements including a first element 1820-1 and a second element 1820-2. The well 1814 is positioned within the tissue well 1816 and has a bottom on the first device 1806a, a first end 1822-1, and a second end 1822-2. The well 1814 is configured for growing an engineered tissue 1824 from cells seeded therein. Culture medium may be added to the tissue well 1816 for growing and / or sustaining the engineered tissue 1824. The engineered tissue 1824, alternatively referred to as artificial tissue, comprises engineered tissue. In one embodiment, the engineered tissue is engineered muscle tissue such as engineered cardiac tissue. The engineered cardiac tissue is generated using Cellular Dynamics International (CDI) iCell Cardiomyocytes2and a side population of normal human ventricular cardiac fibroblasts embedded in a hydrogel composed of fibrin (Sigma- Aldrich), collagen (Sigma-Aldrich) and Matrigel (Corning). In an alternative embodiment, any other human cardiomyocytes are used in place of the iCell Cardiomyocytes2, such as Axol Bioscience Human iPSC-Derived Ventricular Cardiomyocytes or Sigma-Aldrich Human Cardiac Myocytes (HCM). Additionally, or alternatively, fibrin and / or collagen are omitted from the hydrogel. In an alternative embodiment, the engineered muscle tissue is engineered skeletal muscle tissue. The pair of electrodes are separated by a gap within which the well 1814 is positioned. The pair of electrodes are configured to apply an electrical stimulation to tissue within the one or more wells of the first device 1806a (e.g., the engineered tissue 1824 within the well 1814 shown in the expanded portion 1806-1). During maturation of the tissues within the first device 1806a, the pair of electrodes apply stimulation to the tissues according to a multi-week electrical stimulation protocol. After the tissues are matured, the pair of electrodes may be configured to stimulate the tissues (e.g., the engineered tissue 1824) at a set frequency, or pacing frequency. The first element 1820-1 and the second element 1820-2 are disposed across the well 1814 such that there is a gap between the bottom of the well 1814 and the pair of elements. The first element 1820-1 and the second element 1820-2 are configured to: (a) permit attachment of the engineered tissue 1824 formed therebetween, thereby suspending the engineered tissue 1824 above the bottom of the well 1814, and (b) deform in response to the contractile force exerted on the pair of elements by the engineered tissue 1824, thereby simulating a physiological environment that is native to the engineered tissue 1824 and / or permitting measurement of the contractile force (e.g., by the sensor assembly 1808). For example, the pair of electrodes may subject the engineered tissue 1824 to an electrical stimulation at a frequency of 0.1Hz. The engineered tissue 1824 will contract in response to this electrical stimulation causing deformation of at least one of the first element 1820-1 and the second element 1820-2. The sensor assembly 1808 is configured to obtain a plurality of images of a tissue (e.g., the engineered tissue 1824) or a cell culture at a plurality of time points. For example, the optical sensor may be configured to capture an image, or frame, of a tissue at a predetermined time point (e.g., 1 day post-seeding, 2 days post-seeding, etc.). In one embodiment, the control unit 1804 is configured to send a command 1826 to the bioreactor 1802 which causes the sensor assembly 1808 to obtain an image 1828 which is then returned to the control unit 1804. In the present disclosure, the sensor assembly 1808 is configured to obtain a plurality of fluorescent images of a tissue or cell culture. The sensor assembly 1808 comprises a calcium response imaging system and / or a voltage response imaging system. The calcium response imaging system comprises a light source, a microscope, and one or more cameras coupled to the microscope. A triggering system may be utilized to synchronize the light source with frame exposure. In one embodiment, the light source is a Lumencor Spectra III modified to include 340 and 380nm LED diodes, the microscope is a Nikon Ti-E or Ti-2 inverted microscope with 10X objective (MRH00105), and the one or more cameras include an Andor Zyla 4.2, a Hamamatsu ORCA-Fusion C14440-20UP. Anda Hamamatsu ORCA-Fusion BT C15440-20UP which may be used interchangeably. Each of the cameras are able to generate full resolution images at 100fps. The microscope further comprises a Semrock Fura-2-C-000 filter cube comprising a dichroic mirror (409nm) and an emission filter (510 / 84nm). The triggering system is a national instruments PCIe-6323. To obtain fluorescence images of calcium transients a suitable fluorescent dye or genetically encoded indicator is applied to the tissue or cell culture being imaged (e.g., Invitrogen™ Fura-4F). In one embodiment, a Calbryte™ 630 dye is used with excitation at 575nm and a Semrock TxRed filter cube being used with no excitation filter, a dichroic mirror (585nm), and an emission filter (63 / 69nm). When using the Calbryte™ 630 dye, no triggering is used for single wavelength measurement. The voltage response imaging system is the same as the calcium response imaging system but with a suitable voltage sensitive dye being applied to the tissue or cell culture being imaged and an excitation at 635nm and a filter cube with dichroic mirror (652nm) and emission filter (700nm long pass). Figure 19 shows an example computing system for carrying out the methods of the present disclosure. Specifically, Figure 19 shows a block diagram of an embodiment of a computing system according to example embodiments of the present disclosure. Computing system 1900 can be configured to perform any of the operations disclosed herein such as, for example, any of the operations discussed with reference to the functional units described in relation to Figures 18 or 2 or the steps discussed with reference to Figure 15-17. Computing system 1900 includes one or more computing device(s) 1902. The one or more computing device(s) 1902 of computing system 1900 comprise one or more processors 1904 and memory 1906. One or more processors 1904 can be any general purpose processor(s) configured to execute a set of instructions, such as computing instructions including implemented in any one or more programming languages such as Python, Go, C, C++, C#, Java, or the like. For example, one or more processors 1904 can be one or more general-purpose processors, one or more field programmable gate array (FPGA), and / or one or more application specific integrated circuits (ASIC). In one embodiment, one or more processors 1904 include one processor. Alternatively, one or more processors 1904 include a plurality of processors that are operatively connected. One or more processors 1904 are communicatively coupled to memory 1906 via address bus 1908, control bus 1910, and data bus 1912. Memory 1906 can be a random access memory (RAM), a read only memory (ROM), a persistent storage device such as a hard drive, an erasable programmable read only memory (EPROM), and / or the like. The one or more computing device(s) 1902 further comprise I / O interface 1914 communicatively coupled to address bus 1908, control bus 1910, and data bus 1912. Memory 1906 can store information that can be accessed by one or more processors 1904. For instance, memory 1906 (e.g., one or more non-transitory computer-readable storage mediums, memory devices) can include computer-readable instructions (not shown) (e.g., computing instructions) that can be executed by one or more processors 1904. The computer- readable instructions can be software written in any suitable programming language or can be implemented in hardware. Additionally, or alternatively, the computer-readable instructions can be executed in logically and / or virtually separate threads on one or more processors 1904. For example, memory 1906 can store instructions (not shown), such as computing instructions, that when executed by one or more processors 1904 cause one or more processors 1904 to perform operations such as any of the operations and functions for which computing system 1900 is configured, as described herein. In addition, or alternatively, memory 1906 can store data (not shown) that can be obtained, received, accessed, written, manipulated, created, and / or stored. In some implementations, the one or more computing device(s) 1902 can obtain from and / or store data in one or more memory device(s) that are remote from the computing system 1900. Computing system 1900 further comprises storage unit 1916, network interface 1918, input controller 1920, and output controller 1922. Storage unit 1916, network interface 1918, input controller 1920, and output controller 1922 are communicatively coupled to the central control unit (i.e., the memory 1906, the address bus 1908, the control bus 1910, and the data bus 1912) via I / O interface 1914. Storage unit 1916 is a computer readable medium, preferably a non-transitory computer readable medium, comprising one or more programs, the one or more programs comprising computing instructions which when executed by the one or more processors 1904 cause computing system 1900 to perform the method steps of the present disclosure. Alternatively, storage unit 1916 is a transitory computer readable medium. Storage unit 1916 can be a persistent storage device such as a hard drive, a cloud storage device, or any other appropriate storage device. Network interface 1918 can be a Wi-Fi module, a network interface card, a Bluetooth module, and / or any other suitable wired or wireless communication device. In an embodiment, network interface 1918 is configured to connect to a network such as a local area network (LAN), or a wide area network (WAN), the Internet, or an intranet. The above illustrative examples of various aspects and implementations provide an overview for understanding aspects and implementation of the disclosed method. The figures provided herein depict exemplary aspects of the present system and methods and are not intended to limit the scope of the disclosure. Unless otherwise stated, all technical terms used herein have the same meaning as commonly understand by a person skilled in the art. Singular forms “a”, “an” and “the” include plural references unless the context of the disclosure clearly dictates otherwise. The term “or” is intended to encompass “and / or” unless clearly stated otherwise. All publications, patents, and patent applications mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated by reference. ASPECTS OF THE PRESENT DISCLOSURE The following aspects of the disclosure are exemplary and not intended to limit the scope of the disclosure. Aspect 1. A method for event-based quantification of tissue activity, the method comprising: obtaining, by one or more processors, a plurality of images of tissue activity of a tissue over a first time period, the plurality of images having at least two spatial dimensions and a temporal dimension; generating, by the one or more processors, a response matrix from the plurality of images, each element within the response matrix comprising a time series of tissue activity values determined along the temporal dimension of the plurality of images at a respective spatial location within the at least two spatial dimensions; identifying, by the one or more processors, a plurality of events within the response matrix, wherein each event of the plurality of events is associated with a plurality of spatiotemporally contiguous locations within the response matrix at which a predetermined event criterion is met; and generating, by the one or more processors, a first descriptor based on the plurality of events, wherein the first descriptor quantifies tissue activity of the tissue over the first time period. Aspect 2. The method of Aspect 1 further comprising: outputting, by the one or more processors, the first descriptor. Aspect 3. The method of any one of Aspects 1-2 wherein the plurality of images are calcium response images of the tissue. Aspect 4. The method of Aspect 3 wherein the tissue activity is related to calcium activity such that the first descriptor quantifies calcium response of the tissue over the first time period. Aspect 5. The method of any one of Aspects 1-4 wherein the plurality of images are voltage response images of the tissue. Aspect 6. The method of Aspect 5 wherein the tissue activity is voltage activity such that the first descriptor quantifies voltage response of the tissue over the first time period. Aspect 7. The method of any one of Aspects 1-6 wherein the tissue is associated with a first set of conditions over the first time period. Aspect 8. The method of Aspect 7 wherein the first set of conditions comprises one or more of: a reference condition, a drug treatment; a disease model; or a cell line. Aspect 9. The method of Aspect 7 further comprising: obtaining, by the one or more processors, a second descriptor associated with tissue activity of the tissue under a second set of conditions over a second time period; and determining, by the one or more processors, a functional change in the tissue based on a comparison of the first descriptor and the second descriptor, wherein the functional change is indicative of a change from the first set of conditions to the second set of conditions. Aspect 10. The method of Aspect 9 further comprising: determining, by the one or more processors, an effect associated with the change from the first set of conditions to the second set of conditions based on the functional change. Aspect 11. The method of Aspect 10 further comprising: outputting, by the one or more processors, the functional change and / or the effect. Aspect 12. The method of Aspect 9 wherein the second set of conditions comprises one or more of: a drug treatment; a disease model; or a cell line. Aspect 13. The method of any one of Aspects 1-12 wherein the first descriptor comprises a value associated with a homogeneity or heterogeneity feature indicative of a local similarity or diversity of the plurality of events. Aspect 14. The method of any one of Aspects 1-13 wherein the first descriptor comprises a value associated with a relative intracellular calcium concentration feature indicative of a relative difference in intracellular calcium release. Aspect 15. The method of any one of Aspects 1-14 further comprising: extracting, by the one or more processors, one or more patterns from the plurality of events, each of the one or more patterns comprising one or more spatially related events within the plurality of events. Aspect 16. The method of Aspect 15 wherein the first descriptor is further based on the one or more patterns. Aspect 17. The method of Aspect 16 wherein the first descriptor comprises a periodicity value indicative of firing rates of the one or more patterns. Aspect 18. The method of Aspect 16 wherein the first descriptor comprises a value associated with a dysfunction feature indicative of an extent of the one or more patterns. Aspect 19. The method of Aspect 16 wherein the first descriptor comprises a ratio of a number of events within a first group of the one or more patterns to a number of events within a second group of the one or more patterns. Aspect 20. The method of Aspect 15 further comprising: generating, by the one or more processors, an event matrix from the plurality of events such that a first value at a first location within the event matrix is indicative of the first location being associated with an event within the plurality of events; wherein the one or more patterns are extracted from the event matrix. Aspect 21. The method of Aspect 20 further comprising: applying, by the one or more processors, one or more noise filtering operations to the event matrix. Aspect 22. The method of Aspect 21 wherein the one or more noise filtering operations include one or more of a convolution operation and a thresholding operation. Aspect 23. The method of Aspect 20 further comprising: convolving, by the one or more processors, the event matrix with one or more predetermined kernels thereby spatially and temporally expanding each event within the event matrix. Aspect 24. The method of Aspect 15 wherein the extracting one or more patterns from the plurality of events further comprises: when a relative spatial overlap between a first event of the plurality of events and a second event of the plurality of events meets a predetermined overlap threshold, linking, by the one or more processors, the first event and the second event to form a first pattern of the one or more patterns. Aspect 25. The method of Aspect 24 wherein the relative spatial overlap comprises a ratio of a number of spatial locations shared by the first event and the second event to a total number of spatial locations involving either the first event or the second event. Aspect 26. The method of Aspect 24 wherein the predetermined overlap threshold is greater than 0% and less than or equal to 5%. Aspect 27. The method of Aspect 24 wherein the predetermined overlap threshold is 1%. Aspect 28. The method of Aspect 15 further comprising: generating, by the one or more processors, one or more graphical representations of the plurality of events and / or the one or more patterns; and outputting, by the one or more processors, the one or more graphical representations for display to a user. Aspect 29. The method of any one of Aspects 1-28 further comprising, prior to generating the response matrix: downsampling, by the one or more processors, the plurality of images by a predetermined factor. Aspect 30. The method of Aspect 29 wherein the predetermined factor is between 2 and 6. Aspect 31. The method of Aspect 29 wherein the predetermined factor is 4. Aspect 32. The method of any one of Aspects 1-31 further comprising, prior to generating the response matrix: normalizing, by the one or more processors, each image in the plurality of images. Aspect 33. The method of any one of Aspects 1-32 further comprising, prior to identifying the plurality of events: filtering, by the one or more processors, the response matrix using a local filtering strategy. Aspect 34. The method of Aspect 33 wherein the local filtering strategy applies a low-pass filter to each element within the response matrix. Aspect 35. The method of any one of Aspects 1-34 wherein the predetermined event criterion comprises a local maxima criterion. Aspect 36. The method of any one of Aspects 1-35 wherein the predetermined event criterion comprises a threshold criterion. Aspect 37. The method of any one of Aspects 1-36 wherein the tissue is an engineered tissue. Aspect 38. The method of Aspect 37 wherein the plurality of images are obtained from a bioreactor containing the engineered tissue. Aspect 39. The method of Aspect 37 wherein the engineered tissue is an engineered cardiac tissue. Aspect 40. The method of Aspect 37 wherein the engineered tissue is an engineered skeletal tissue. Aspect 41. The method of any one of Aspects 1-40 wherein the tissue is an in vivo tissue. Aspect 42. A non-transitory computer readable medium storing computing instructions which, when executed by one or more processors, cause the one or more processors to: obtain a plurality of images of tissue activity of a tissue over a first time period, the plurality of images having at least two spatial dimensions and a temporal dimension; generate a response matrix from the plurality of images, each element within the response matrix comprising a time series of tissue activity values determined along the temporal dimension of the plurality of images at a respective spatial location within the at least two spatial dimensions; identify a plurality of events within the response matrix, wherein each event of the plurality of events is associated with a plurality of spatiotemporally contiguous locations within the response matrix at which a predetermined event criterion is met; and generate a first descriptor based on the plurality of events, wherein the first descriptor quantifies tissue activity of the tissue over the first time period. Aspect 43. A device comprising one or more processors and a memory storing computing instructions which, when executed by the one or more processors, cause the one or more processors to: obtain a plurality of images of tissue activity of a tissue over a first time period, the plurality of images having at least two spatial dimensions and a temporal dimension; generate a response matrix from the plurality of images, each element within the response matrix comprising a time series of tissue activity values determined along the temporal dimension of the plurality of images at a respective spatial location within the at least two spatial dimensions; identify a plurality of events within the response matrix, wherein each event of the plurality of events is associated with a plurality of spatiotemporally contiguous locations within the response matrix at which a predetermined event criterion is met; and generate a first descriptor based on the plurality of events, wherein the first descriptor quantifies tissue activity of the tissue over the first time period. Aspect 44. A system for event-based tissue activity quantification, the system comprising: a bioreactor comprising: a device configured for growing tissue; and a sensor assembly configured to obtain tissue activity images of a tissue within the device; and a control unit communicatively coupled to the bioreactor, wherein the control unit comprises one or more processors and a memory storing computing instructions, which when executed by the one or more processors, causes the one or more processors to: obtain a plurality of images of tissue activity of a tissue over a first time period, the plurality of images having at least two spatial dimensions and a temporal dimension, generate a response matrix from the plurality of images, each element within the response matrix comprising a time series of tissue activity values determined along the temporal dimension of the plurality of images at a respective spatial location within the at least two spatial dimensions, identify a plurality of events within the response matrix, wherein each event of the plurality of events is associated with a plurality of spatiotemporally contiguous locations within the response matrix at which a predetermined event criterion is met, and generate a first descriptor based on the plurality of events, wherein the first descriptor quantifies tissue activity of the tissue over the first time period.

Claims

CLAIMS 1. A method for event-based quantification of tissue activity, the method comprising: obtaining, by one or more processors, a plurality of images of tissue activity of a tissue over a first time period, the plurality of images having at least two spatial dimensions and a temporal dimension; generating, by the one or more processors, a response matrix from the plurality of images, each element within the response matrix comprising a time series of tissue activity values determined along the temporal dimension of the plurality of images at a respective spatial location within the at least two spatial dimensions; identifying, by the one or more processors, a plurality of events within the response matrix, wherein each event of the plurality of events is associated with a plurality of spatiotemporally contiguous locations within the response matrix at which a predetermined event criterion is met; and generating, by the one or more processors, a first descriptor based on the plurality of events, wherein the first descriptor quantifies tissue activity of the tissue over the first time period.

2. The method of claim 1 further comprising: outputting, by the one or more processors, the first descriptor.

3. The method of claim 1 wherein the plurality of images are calcium response images of the tissue.

4. The method of claim 3 wherein the tissue activity is related to calcium activity such that the first descriptor quantifies calcium response of the tissue over the first time period.

5. The method of claim 1 wherein the plurality of images are voltage response images of the tissue.

6. The method of claim 5 wherein the tissue activity is voltage activity such that the first descriptor quantifies voltage response of the tissue over the first time period.

7. The method of claim 1 wherein the tissue is associated with a first set of conditions over the first time period.

8. The method of claim 7 wherein the first set of conditions comprises one or more of: a reference condition, a drug treatment; a disease model; or a cell line.

9. The method of claim 7 further comprising: obtaining, by the one or more processors, a second descriptor associated with tissue activity of the tissue under a second set of conditions over a second time period; and determining, by the one or more processors, a functional change in the tissue based on a comparison of the first descriptor and the second descriptor, wherein the functional change is indicative of a change from the first set of conditions to the second set of conditions.

10. The method of claim 9 further comprising: determining, by the one or more processors, an effect associated with the change from the first set of conditions to the second set of conditions based on the functional change.

11. The method of claim 10 further comprising: outputting, by the one or more processors, the functional change and / or the effect.

12. The method of claim 9 wherein the second set of conditions comprises one or more of: a drug treatment; a disease model; or a cell line.

13. The method of claim 1 wherein the first descriptor comprises a value associated with a homogeneity or heterogeneity feature indicative of a local similarity or diversity of the plurality of events.

14. The method of claim 1 wherein the first descriptor comprises a value associated with a relative intracellular calcium concentration feature indicative of a relative difference in intracellular calcium release.

15. The method of claim 1 further comprising: extracting, by the one or more processors, one or more patterns from the plurality of events, each of the one or more patterns comprising one or more spatially related events within the plurality of events.

16. The method of claim 15 wherein the first descriptor is further based on the one or more patterns.

17. The method of claim 16 wherein the first descriptor comprises a periodicity value indicative of firing rates of the one or more patterns.

18. The method of claim 16 wherein the first descriptor comprises a value associated with a dysfunction feature indicative of an extent of the one or more patterns.

19. The method of claim 16 wherein the first descriptor comprises a ratio of a number of events within a first group of the one or more patterns to a number of events within a second group of the one or more patterns.

20. The method of claim 15 further comprising: generating, by the one or more processors, an event matrix from the plurality of events such that a first value at a first location within the event matrix is indicative of the first location being associated with an event within the plurality of events; wherein the one or more patterns are extracted from the event matrix.

21. The method of claim 20 further comprising: applying, by the one or more processors, one or more noise filtering operations to the event matrix.

22. The method of claim 21 wherein the one or more noise filtering operations include one or more of a convolution operation and a thresholding operation.

23. The method of claim 20 further comprising: convolving, by the one or more processors, the event matrix with one or more predetermined kernels thereby spatially and temporally expanding each event within the event matrix.

24. The method of claim 15 wherein the extracting one or more patterns from the plurality of events further comprises: when a relative spatial overlap between a first event of the plurality of events and a second event of the plurality of events meets a predetermined overlap threshold, linking, by the one or more processors, the first event and the second event to form a first pattern of the one or more patterns.

25. The method of claim 24 wherein the relative spatial overlap comprises a ratio of a number of spatial locations shared by the first event and the second event to a total number of spatial locations involving either the first event or the second event.

26. The method of claim 24 wherein the predetermined overlap threshold is greater than 0% and less than or equal to 5%.

27. The method of claim 24 wherein the predetermined overlap threshold is 1%.

28. The method of claim 15 further comprising: generating, by the one or more processors, one or more graphical representations of the plurality of events and / or the one or more patterns; and outputting, by the one or more processors, the one or more graphical representations for display to a user.

29. The method of claim 1 further comprising, prior to generating the response matrix:downsampling, by the one or more processors, the plurality of images by a predetermined factor.

30. The method of claim 29 wherein the predetermined factor is between 2 and 6.

31. The method of claim 29 wherein the predetermined factor is 4.

32. The method of claim 1 further comprising, prior to generating the response matrix: normalizing, by the one or more processors, each image in the plurality of images.

33. The method of claim 1 further comprising, prior to identifying the plurality of events: filtering, by the one or more processors, the response matrix using a local filtering strategy.

34. The method of claim 33 wherein the local filtering strategy applies a low-pass filter to each element within the response matrix.

35. The method of claim 1 wherein the predetermined event criterion comprises a local maxima criterion.

36. The method of claim 1 wherein the predetermined event criterion comprises a threshold criterion.

37. The method of claim 1 wherein the tissue is an engineered tissue.

38. The method of claim 37 wherein the plurality of images are obtained from a bioreactor containing the engineered tissue.

39. The method of claim 37 wherein the engineered tissue is an engineered cardiac tissue.

40. The method of claim 37 wherein the engineered tissue is an engineered skeletal tissue.

41. The method of claim 1 wherein the tissue is an in vivo tissue.

42. A non-transitory computer readable medium storing computing instructions which, when executed by one or more processors, cause the one or more processors to: obtain a plurality of images of tissue activity of a tissue over a first time period, the plurality of images having at least two spatial dimensions and a temporal dimension; generate a response matrix from the plurality of images, each element within the response matrix comprising a time series of tissue activity values determined along the temporal dimension of the plurality of images at a respective spatial location within the at least two spatial dimensions; identify a plurality of events within the response matrix, wherein each event of the plurality of events is associated with a plurality of spatiotemporally contiguous locations within the response matrix at which a predetermined event criterion is met; and generate a first descriptor based on the plurality of events, wherein the first descriptor quantifies tissue activity of the tissue over the first time period.

43. A device comprising one or more processors and a memory storing computing instructions which, when executed by the one or more processors, cause the one or more processors to: obtain a plurality of images of tissue activity of a tissue over a first time period, the plurality of images having at least two spatial dimensions and a temporal dimension; generate a response matrix from the plurality of images, each element within the response matrix comprising a time series of tissue activity values determined along the temporal dimension of the plurality of images at a respective spatial location within the at least two spatial dimensions; identify a plurality of events within the response matrix, wherein each event of the plurality of events is associated with a plurality of spatiotemporally contiguous locations within the response matrix at which a predetermined event criterion is met; andgenerate a first descriptor based on the plurality of events, wherein the first descriptor quantifies tissue activity of the tissue over the first time period.

44. A system for event-based tissue activity quantification, the system comprising: a bioreactor comprising: a device configured for growing tissue; and a sensor assembly configured to obtain tissue activity images of a tissue within the device; and a control unit communicatively coupled to the bioreactor, wherein the control unit comprises one or more processors and a memory storing computing instructions, which when executed by the one or more processors, causes the one or more processors to: obtain a plurality of images of tissue activity of a tissue over a first time period, the plurality of images having at least two spatial dimensions and a temporal dimension, generate a response matrix from the plurality of images, each element within the response matrix comprising a time series of tissue activity values determined along the temporal dimension of the plurality of images at a respective spatial location within the at least two spatial dimensions, identify a plurality of events within the response matrix, wherein each event of the plurality of events is associated with a plurality of spatiotemporally contiguous locations within the response matrix at which a predetermined event criterion is met, and generate a first descriptor based on the plurality of events, wherein the first descriptor quantifies tissue activity of the tissue over the first time period.

Images