Methods, programs, apparatus for obtaining health information from sensors in an ingestible capsule

Abstract

A method comprising following ingestion of an ingestible capsule by a subject, the ingestible capsule housing a motion sensor configured to generate a time series of motion sensor data representing motion of the ingestible capsule during passage through the GI tract of a subject, at data processing hardware communicably coupled to the motion sensor, performing a process comprising: generating a spectral analysis of the time series of motion sensor data generated by the motion sensor over a time period during passage of the ingestible capsule through the GI tract, using the spectral analysis to detect peristalsis at a location of the ingestible capsule within the GI tract at the time period.

Description

Methods, Programs, Apparatus for Obtaining Health Information from Sensors in an Ingestible CapsuleTechnical FieldThis invention relates to ingestible sensor capsules for medical and health applications in the gastrointestinal (GI) tract of mammals including humans, and specifically relates to recording sensor data within the ingestible capsule and determining a location within the GI tract based on the data.BackgroundIngestible capsules housing sensors may be used to provide information about the health of a subject.Gut health is increasingly identified as a contributor to overall health and wellness. Motility of an ingestible capsule (with or without associated gas constituent measurements) provides important information in the assessment of gut health. Determining location within the GI tract provides health information in itself, by providing information for GI tract motility reporting, and also provides context for sensor readings such as gas sensor readings. WO2023 / 064996 explains techniques for using sensor data from on-board an ingestible capsule to determine capsule location.Furthermore, as set out in W02023087074, a determination of the location of the ingestible capsule may be used to time the release of therapeutic matter from a releasable chamber in the capsule, so that the therapeutic matter can be delivered directly to a specific region of the GI tract.It is desirable to find accurate and reliable techniques for detecting and reporting data and information indicative of GI tract health, or data and information from the GI tract indicative of general patient health.It is desirable to take GI tract samples from a predefined target location for analysis ex-vivo in a lab or clinical environment.StatementsEmbodiments include A method comprising, for a patient undertaking an incretin mimetic dosage regime, the patient having ingested or been administered an ingestible capsule configured to determine and report the location of the ingestible capsule within the GI tract, or to cooperate with a computing device to determine and report the location of the ingestible capsule within the GI tract: determining a time series of locations of the capsule within the GI tract; determining a gastro-intestinal motility indicator based on the time series of locations of the ingestible capsule within the GI tract; and titratingor otherwise modifying the incretin mimetic dosage regime in dependence upon the gastro-intestinal motility indicator.Embodiments include a method comprising, for a patient undertaking an incretin mimetic dosage regime, the patient having ingested or been administered an ingestible capsule configured to diagnose gastroparesis or suspected gastroparesis in a subject: determining a gastro-intestinal motility indicator based on a positive or negative diagnosis of gastroparesis or suspected gastroparesis in the subject as reported by the ingestible capsule; and titrating or otherwise modifying the incretin mimetic dosage regime in dependence upon the gastro-intestinal motility indicator.Embodiments include a method comprising, for a patient undertaking a medicament dosage regime, the patient having ingested or been administered an ingestible capsule configured to determine and report the location of the ingestible capsule within the GI tract, or configured to cooperate with a computing device to determine and report the location of the ingestible capsule within the GI tract: determining a time series of locations of the capsule within the GI tract; determining a gastro-intestinal motility indicator based on the time series of locations of the ingestible capsule within the GI tract; and titrating or otherwise modifying the medicament dosage regime in dependence upon the gastro-intestinal motility indicator.Embodiments include a method comprising, for a patient undertaking a medicament dosage regime, the patient having ingested or been administered an ingestible capsule configured to diagnose gastroparesis or suspected gastroparesis in a subject; determining a gastro-intestinal motility indicator based on a positive or negative diagnosis of gastroparesis or suspected gastroparesis in the subject; and titrating or otherwise modifying the medicament dosage regime in dependence upon the gastro-intestinal motility indicator.List of FiguresAdetailed description of embodiments including apparatus, methods, programs, processes, and systems, is set out below, with particular reference to accompanying drawings, in which:Figure 1 illustrates a method;Figure 2 illustrates apparatus;Figures 3a and 3b illustrate schematically electronic components of ingestible capsules;Figure 4 illustrates schematically electronic components of an ingestible capsule;Figures 5a to 5c illustrate schematically electronic components of an ingestible capsule or a system including an ingestible capsule;Figure 6a illustrates a time series of accelerometer data;Figure 6b illustrates accelerometer data in the frequency domain;Figure 6c illustrates accelerometer data in the frequency domain;Figure 6d illustrates time series of sensor data from additional sensors and from an accelerometer; Figure 7 illustrates a reflectometer;Figure 8 illustrates time series sensor data from additional sensors and from an accelerometer;Figure 9 illustrates time series sensor data from additional sensors and from an accelerometer;Figure 10 illustrates a part of a method according to an embodiment; andFigure 11 illustrates a method.GeneralFigure 1 illustrates a method. Figure 2 illustrates an apparatus suitable for performing the methods. Figures 3a and 3b are schematic illustrations of ingestible capsules 10.The method of Figure 1 is a method for determining a location of an ingestible capsule 10 within the GI tract of a subject 40, illustrated in Figure 2. As illustrated in Figures 3a and 3b, the ingestible capsule 10 houses a motion sensor 19. The motion sensor 19 is configured to generate a time series of motion sensor data representing acceleration, rotation, or positional changes experienced by the ingestible capsule 10 (otherwise referred to as raw accelerometer data). The motion sensor 19 is fixed in position within the ingestible capsule 10 so that the ingestible capsule 10 and the motion sensor 19 experience the same acceleration, rotation, and positional changes.Location within the GI tract may be an indication of presence in either the stomach, the small intestine, or the large intestine. Optionally, the location may be provided to a greater level of specificity, such as proximal or distal small intestine.Steps S104A to S106 may be referred to collectively as spectral analysis processing steps. Each of S 104A to S 106 utilises the result of the spectral analysis from S 103 to extract information, detect events, or determine information. The spectral analysis processing steps S104Ato S106 may be performed onboard the capsule or remotely (see Figure 2). The results of the spectral analysis processing steps may be included in a report generated on-board the capsule 10 and transmitted away to a receiver computing apparatus 30, or may be included in a report generated by remote processing apparatus and transmitted to a recipient such as the subject or a clinician.Steps S 104B to S 107 may be performed in combination with one another, or separately. A single capsule 10 may be configured to perform any one of S104A to S107, all three of S104A to S107, or some combination thereof. Likewise, a system comprising an ingestible capsule 10 and remote processingapparatus maybe configured to perform any one of S104Ato S107, all three of S104Ato S107, or some combination thereof.Apparatus Arrangements OverviewAs shown in Figure 2, apparatus arrangements for performing methods for determining location of the capsule 10 within the GI tract based on a spectral analysis of a time series of data generated by the motion sensor 19, such as illustrated by Figure 1, may comprise:• only the ingestible capsule 10: in this case processing of the time series of motion sensor data to obtain a determination of location via spectral analysis is performed on-board the capsule by data processing hardware 15;• the ingestible capsule 10 and remote processing apparatus, wherein the remote processing apparatus may comprise only a receiver computing apparatus 30 in direct data communication with the ingestible capsule 10, or may comprise a receiver computing apparatus 30 in direct data communication with the ingestible capsule 10 and one or more further computing apparatus 20 to receive data from the receiver computing apparatus 30 over a network.Thus, the receiver computing apparatus 30 and the further computing apparatus 20 are optional. The receiver computing apparatus 30 may be a smart phone, a tablet, or some other personal computing device, or server computer, configured to pair, couple, or otherwise establish a direct data communication with a wireless data transceiver 18 of the ingestible capsule 10. The further computing apparatus 20 may be a smart phone, a tablet, or some other personal computing device, server computer, or cloud computing device / service / infrastructure in network communication with the receiver computing apparatus 30. The receiver computing apparatus 30 may be a personal device of the subject 30. The further computing apparatus 20 may be a device of a clinical service provider, an ingestible capsule provider, or some other entity.Data communications between the capsule 10 and the receiver computing apparatus 30 may be oneway, wherein raw motion sensor data is transmitted from the capsule 10 to the receiver computing apparatus 30 for processing (by the receiver computing apparatus 30 and / or the further computing apparatus 20), and no data flows in the reverse direction. Alternatively, a result of processing may be transmitted back to the capsule 10.In a particular implementation example based on Figure 1, transmission from remote processing apparatus to the ingestible capsule 10 may be in response to determining at S104B that a location of the capsule 10 within the GI tract is a target location for therapeutic matter being carried by the capsule 10 and thus the transmission is a trigger signal to trigger release of the therapeutic matter by the capsule 10.A process comprising the spectral analysis at S103, and the spectral analysis processing steps S104B to S 106 may be performed on-board the ingestible capsule 10, may be performed by a receiver computing apparatus 30 in direct data communication with the ingestible capsule 10, or may be performed by further computing apparatus 20 in data communication with the receiver computing apparatus 30 over a network (such as the internet). The process may be performed by two or three of those elements in combination.The apparatus arrangements of Figure 2 may also be referred to as systems. The subject 40 is illustrated for context but is not intended to form part of an apparatus or system.Capsule Arrangements OverviewFigures 3a and 3b schematically illustrate basic capsule arrangements. The ingestible capsule 10 comprises at least a motion sensor 19 and a power source 16. Other components such as control circuitry may be present in the capsule 10 but are not illustrated. The ingestible capsule 10 of Figure 3a includes data processing hardware 15 comprising processor hardware 151 such as a CPU for processing data, and a memory hardware 152 for storing data, in preparation for and during processing, and / or processing results.As illustrated in Figure 3b, the data processing hardware 15 is optional since the ingestible capsule 10 may comprise a wireless transceiver 18 to transmit the motion sensor data (and other sensor data generated by the capsule 10) away to remote apparatus (see Figure 2) for processing. Optionally, the ingestible capsule 10 may comprise data processing hardware 15 for on-board processing, and a wireless transceiver 18 to transmit a result of the on-board data processing to a remote apparatus for one or more from among further processing, storage, reporting, etc. Optionally, the spectral analysis step S 103 may be performed on-board the capsule 10, and the resultant frequency domain representation of the motion sensor data transmitted to the remote processing apparatus for the spectral analysis processing step or steps S104B-S106.The ingestible capsule 10 illustrated in Figure 3a and in Figure 3b may further comprise additional sensor hardware such as an EMG 31, a pulse-oximetry sensor 32, or a gas sensor, which would be housed within a gas sensing headspace within a gas permeable membrane and sealed from the remainder of the components by an impermeable membrane. As discussed below, the ingestible capsule 10 may be of an arrangement configured to release therapeutic matter into the GI tract at a timing based on a location determined at S104B by spectral analysis of a time series of data generated by the motion sensor 19.Motion sensor hardwareThe motion sensor 19 may be a gyroscope. The motion sensor 19 may be an accelerometer. The motion sensor 19 may comprise a gyroscope and an accelerometer. For example, the motion sensor 19 may be a tri-axis accelerometer. The motion sensor 19 may be a 12-bit tri-axis accelerometer. The motion sensor 19 may be a single-axis accelerometer or a two-axis accelerometer. The motion sensor 19 may comprise a piezo-film sensor, a surface MM capacitive sensor, a bulk capacitive sensor, and / or a piezo-electric electromechanical servo vibrational sensor. The motion sensor 19 may comprise a fibre optic accelerometer, a Hall effect accelerometer, a magnetoresistive accelerometer, and / or a strain gauge accelerometer.On-the-fly versus retrospective processing & on-board versus remote processingMethods for determining location of the capsule 10 within the GI tract based on a spectral analysis of a time series of data generated by the motion sensor 19, such as illustrated by Figure 1, may be performed on-the-fly (during GI tract passage), or may be performed retrospectively (following GI tract passage).In the on-the-fly processing case, the processing is retrospective insofar as the data being processed was generated by the motion sensor 19 in the past, but the processing is being performed more or less instantaneously after the end of a time period during which the time series represents to obtain a processing result representing the said time period, for example, a location of the capsule within the said time period.On-the-fly processing may be useful in examples such as determining location in order to release therapeutic matter into a particular location within the GI tract. Such processing may be useful in order to reduce a data transmission overhead, so that rather than transmitting raw motion sensor data away from the capsule 10, it is only necessary to transmit results of a location determination for a time period or location determinations for a series of time periods. Or, for example, to transmit the spectral information resulting from the spectral analysis, that is, the frequency domain representation of the time series of motion sensor data, or a compressed or otherwise processed version thereof. Wherein processed or compressed may indicate that signals or components in one or more relevant frequency ranges are extracted and other signals or components discards. Relevant may be a predefined set of peristalsis indicator frequency ranges.On-the-fly processing of the time series of motion sensor data to obtain a spectral analysis result at S 103 and a result of a spectral analysis processing step S 104B to S 106 may be performed on-board the capsule 10 by an arrangement such as illustrated in Figure 3a, or may be transmitted away from the capsule during passage through the GI tract for processing at a remote apparatus in an arrangement such as illustrated in Figure 3b.Transmission of the raw motion sensor data (i.e. the time series of motion sensor data) from the capsule 10 to the remote apparatus may be during passage through the GI tract. For example, the wireless data transceiver 18 may pair or otherwise connect with the remote apparatus for transmission according to a protocol such as Bluetooth or Bluetooth Low Energy transmission protocol. Likewise, results of processing on-board the capsule may be transmitted to a remote apparatus via the same mechanism. Results of processing on-board the capsule may be stored on the capsule for reporting by transmission to a remote apparatus after detection of excretion from the GI tract (by detecting a freefall event via the motion sensor 19 or by changes detected by temperature or relative humidity sensors, if included) triggering a burst of data transmission in an inquiry mode.Retrospective processing, which is taken to mean processing after passage through the GI tract, may be performed at a remote apparatus in an arrangement such as illustrated in Figures 2 and 3b.In the case of processing on-board the capsule 10, there may still be a wireless data transceiver 18 for transmitting away a result of the processing to a remote apparatus, and in the case of there being additional sensor hardware, it may be that data from the additional sensor hardware is transmitted away via the wireless data transceiver 18. Likewise, the additional sensor data may be processed on board the capsule 10 and a processing result transmitted away. It may be that no data is transmitted away from the ingestible capsule 10, for example if the processing result is a location determination at S104B used to determine a release timing of therapeutic matter, then the processing result does not necessarily need to be transmitted away from the capsule 10 (however it may be transmitted away for reporting purposes).Retrospective processing may be useful in examples such as health reporting and diagnostics, when a clinician wishes to obtain a report on GI tract motility of a patient (comprising one or more from among whole gut transit time, small bowel residence time, large intestine residence time, gastric residence time, etc) based on the method of Figure 1.A report may include additional sensor data generated at each location within the GI tract. Wherein such additional sensor data may be generated by on-board gas sensors indicating gases present in the gas mixture surrounding the capsule 10 during its passage through the GI tract, noting that combining such sensor data with determinations of location from S 104B in particular time windows enables a clinician to determine whether and to what extent gases indicative of disorders such as gastroparesis or SIBO are present. Of course, in such examples on-the-fly processing could also be implemented.In retrospective processing implementations, the raw motion sensor data (that is, the time series of motion sensor data) is processed by a remote apparatus after the passage through the GI tract.Transmission of the motion sensor data from the capsule 10 to the remote apparatus may be during passage through the GI tract. For example, the wireless data transceiver 18 may pair or otherwise connect with the remote apparatus for transmission according to a protocol such as Bluetooth or Bluetooth Low Energy transmission protocol. In another example, the capsule 10 may detect excretion from the GI tract (by detecting a freefall event via the motion sensor 19 or by changes detected by temperature or relative humidity sensors, if included) and then trigger a burst of data transmission in an inquiry mode. Transmission protocols are discussed specifically inPCT / AU2023 / 050801.Method of Figure 1 in more detailSI 01 Capsule IngestionAt step S101 the ingestible capsule 10 ingested by a subject 40. The subject may be a human subject. The ingestible capsule 10 may be obtained by the subject from a clinician (that is, a medical professional) for obtaining information about the condition of the GI tract, respiratory patterns, or other health information about a subject. Further, the ingestible capsule 10 may be provided as a means to deliver therapeutic matter directly to a particular location in the GI tract of the subject.The ingestible capsule 10 may be configured to power on upon removal from a package in which it is supplied, or upon receipt of a custom wireless signal from a dedicated application on a remote apparatus such as receiver computing apparatus 30.Optionally, the ingestible capsule 10 may include a secondary wireless transceiver, which may use an NFC communication protocol. The secondary transceiver is for specific activation control signalling only, such as for initiating an active mode of the capsule 10 at an unpackaging stage or otherwise prior to ingestion of the capsule 10. The secondary transceiver is not active during the live phase of the capsule 10, i.e. during passage through the GI tract of the subject mammal. The secondary transceiver does not contribute to transmission of the data transmission payload from the capsule 10 to the receiver apparatus 30. The secondary transceiver may not be required to perform any transmission whatsoever, that is, the secondary transceiver may only be required to receive an activation control signal from a smartphone or tablet running an application. However, the NFC protocol may require two-way exchange of signals such as a handshake or coupling process to enable said activation control signal to be transmitted from the smartphone or tablet and received by the capsule 10. Nonetheless, since the secondary transceiver is inactive while the capsule 10 passes through the GI tract of the subject mammal, unlike in the case of the primary transceiver, there is no requirement that the secondary transceiver be configured to transmit signals from inside the GI tract of the subject mammal 40 to a receiver apparatus 30 external to the subject mammal.Optionally, the secondary transceiver may be configured to receive an encoded activation control signal from a smartphone or tablet (e.g. the receiver computing apparatus 30) running an application configured for managing interactions between the smartphone or tablet (tablet in this context meaning tablet computer) and the capsule 10, which encoded activation control signal initiates a live phase of the capsule 10 during which capsule sensors take readings and the readings themselves or metrics and / or reports based on the readings are transmitted from the capsule 10 to the smartphone or table via the primary wireless data transceiver. Thus, the secondary wireless data transceiver is active in a listening phase which precedes a live phase of the capsule. The primary wireless data transceiver (and the other components such as the sensor hardware, processor hardware, etc) is inactive (i.e. consuming no power whatsoever) during the listening phase. Once the encoded activation control signal is received (and the capsule 10 powered on in response) the listening phase ends and the secondary wireless data transceiver becomes inactive. The primary wireless data transceiver is active during the live phase.In order to conserve battery power, capsule 10 may operate in a standby or listening mode during the time between release from manufacturing and initiation of the live phase during which readings are recorded by the on-board sensors and transmitted away from the capsule. The standby or listening mode is an extremely low power mode (for example, the sensors and the data processing hardware is inactive during the standby or listening mode). A live phase of the capsule is initiated prior to ingestion by the subject 40. A mechanism for ending the standby or listening mode and entering a live phase may include a reed switch coupled to a magnet on the packaging which is triggered by release of the capsule from the packaging and when triggered powers on the processor, sensors, and primary transceiver (i.e. initiates the live phase). An alternative mechanism is based on Near Field Communication, NFC. In the alternative mechanism, the capsule 10 is maintained in the standby or listening mode (which in the particular example of the NFC is a SENSE mode) prior to being issued to the subject. In the listening mode, when an electromagnetic field is detected with an appropriately encoded activation control message, the on-board control circuitry (such as a microcontroller) enters a live phase. A receiver computing apparatus 30 running an application configured for the purpose of managing interactions with the capsule 10 and the processing of data received therefrom, and having NFC capability, can generate the appropriately encoded activation control message. In particular, a back-end server may link a user account to a particular capsule instance, so that when that user is logged in to the application and selects to activate a capsule, the application performs a lookup to the back-end server to determine how to encode the activation control message. In other words, optionally the encoding is unique per capsule. Alternatively, the encoding may be uniform across a batch of capsules or all capsules.SI 02 Generate time series of motion sensor dataAt S102 the capsule 10 is in a live phase and so the motion sensor 19 is configured to take readings. For example, under control of control circuitry 15 onboard the capsule 10 controlling power supply to the accelerometer and readings therefrom.An exemplary motion sensor 19 is an accelerometer measuring roll about three mutually orthogonal axes. The readings from the accelerometer 19 may be vectors with a component per axis, with each component indicating an instantaneous angular acceleration about the corresponding axis, or an average acceleration about the corresponding axis over the time period since the preceding live reading. Alternatively, the readings may give a three dimensional orientation of the capsule. The motion sensor 19 may be a gyroscope or angular rate sensor. The motion sensor 19 may be a magnetometer coupled to a magnet external to the subject. The magnet may be fixed to the subject or to a building or to an object. Some pre-processing of the raw motion sensor data may be performed between S102 and S103 to prepare the motion sensor data for transform from the time domain to the frequency domain at S 103. Processing of the readings from the motion sensor 19 may be performed to generate a representation (such as a plot vs time) of aggregated (i.e. all three axes) motion sensor readings. In addition to being a basis for spectral analysis at S103, such a plot or representation may also be used to identify motility markers for events including an excretion event, gastric -duodenal transition, and ileocecal junction transition.The capsule orientation may be measured using a triaxial accelerometer and tracking the gravity vector or another fixed vector (such as provided by an external magnet) with respect to the capsule frame of reference.Readings from the motion sensor are a time series, so that there is a time value implicitly or explicitly associated with each reading. In the explicit case, readings may be time-stamped, and in the implicit case the readings are chronological and separated by a predefined time resolution step so that a timing can be inferred from a time of initiation of the live phase and a position of the reading within the chronological order.The motion sensor 19 may be a tri-axis accelerometer generating a single time series of accelerometer data or one time series of accelerometer data per axis.In case the motion sensor data comprises three time series each representing acceleration in a respective one of the three axes, the process may include a pre-spectral analysis step between S102 and S103 comprising: combining the three time series into a single resultant time series representing pitch angle (or tilt angle) between a capsule reference axis in fixed relation to the capsule, and an earth reference axis in fixed relation to the earth. The pre-spectral analysis step may comprise combining the three timeseries into a single resultant time series, for example by a vector addition, a summation of magnitudes, or an RSS combination. Spectral analysis at S103 would thenbe performed on the single resultant time series.In an alternative case, the motion sensor data is accelerometer data comprising three contemporary time series each representing acceleration in a respective one of the three axes, and combining is performed in the frequency domain (as part of S103).A preprocessing step may comprise discarding a time series representing an axis from which a predetermined number or proportion of readings are missing.Time Series Duration and Other ParametersThe time series of motion sensor data is a truncated time series insofar as it represents motion of the capsule 10 during a subset of the overall time during which it is resident in the GI tract of the subject. The length of the time window determines the frequency resolution of the result of the spectral analysis. The sampling frequency of the motion sensor readings (i.e. the time separation between adjacent readings in the time series of motion sensor data) determines the maximum frequency that is detectable in the frequency domain representation of the accelerometer data.Motion sensor takes readings at a sampling rate fa. Each motion sensor reading represents a motion sensor reading duration l / fa. This is the fundamental sampling rate of the data. The size of the FFT (or other spectral analysis result) is N and the frequency resolution is fa / N.The time series of motion sensor data may be divided into a series of time windows (either consecutive or sliding) each comprising M / fa, wherein M is a positive integer and is fixed or adaptive. M / fais a duration of time covered by a time window, wherein M is the number of readings in the time series data.The time window may be a period of time having a length, for example, between 4 minutes 20 minutes, between 5 minutes and nineteen minutes, between six minutes and eighteen minutes, between seven minutes and seventeen minutes, between eight minutes and sixteen minutes. The time window may have a length of ten minutes, that is, be between nine minutes and eleven minutes in length, or between eight minutes and twelve minutes in length, or between seven minutes and thirteen minutes in length, or between six minutes and fourteen minutes in length, or between five minutes and fifteen minutes in length. The time window may be between fifteen minutes and twenty five minutes in length, between twenty five minutes and thirty five minutes in length, between thirty five minutes and forty five minutes in length, or between forty five minutes and fifty five minutes in length, or between fifty minutes and sixty five minutes in length.The time window may be as short as one sample l / fa, or as long as all samples in the time series M, or any length in between. Time window length may be fixed or adaptive. Applying an adaptive or variable time window duration function may provide greater spectral resolution at one time window and greater data loss immunity at another.Figure 6a illustrates a time series of motion sensor data generated by an accelerometer and representing acceleration on the x axis relative to fixed gravitational vector g for a ten minute window 5 hours after ingestion of a sample ingestible capsule.Step SI 03: Generate spectral analysisStep S103 comprises conducting a spectral analysis of the time series of motion sensor data generated by the motion sensor over a time period during passage of the ingestible capsule through the GI tract.The motion sensor data is time-sampled. In other words, data from the time series of motion sensor data belonging to a time window of duration T is extracted and processed at SI 03. Figure 6a illustrates an exemplary data sample wherein T is ten minutes and the timing is five hours after ingestion of the capsule 10 by the subject 40.It is noted that spectral analysis may be performed on one or a series of such time windows. The series of time windows may be contiguous, or they may be partially overlapping.In the method of Figure 1, spectral analysis of a single time window may be sufficient to determine a capsule location at S104B, or spectral analysis of a series of time windows may be combined to determine a location of the capsule during the series of time windows, noting the relative loss in precision in the latter case. Time windows may be of a fixed length or may be adaptive.Spectral analysis at S103 comprises transforming the time series of motion sensor data from the time domain to the frequency domain. An example of a process for performing the transform is a Fast Fourier Transform, FFT.S104BA result of spectral analysis at S 103 is a frequency domain representation of the motion sensor data from the time window. The frequency domain representation may be a list, array, or another representation of signal magnitude in each of a series of component frequencies.Figure 6b illustrates the result of a spectral analysis performed on the time series of motion sensor (accelerometer) data from Figure 6a. In this example it may be that the y-axis and z-axis data was alsotransformed to the frequency domain, but that the x-axis data had the greatest signal magnitude in the set of peristalsis indicator frequency ranges (which may, for example, be determined by a summation per axis of signal magnitudes in component frequencies falling within one or more frequency ranges in a predefined set of relevant frequency ranges (such as peristalsis indicator frequency ranges in the context of the method of Figure 1), and a comparison of the summations to establish the greatest).In particular, as illustrated in Figure 6b, a peak in the frequency domain is observed at around 3 cycles per minute, which is within the frequency range indicating stomach peristalsis, as discussed below in relation to S104B. In the example of Figure 6b, the result of the spectral analysis is a measure of acceleration energy at each of a series of component frequencies. Figure 6b is an illustration of underlying data, and in implementation the underlying data may remain as a data list or array without being rendered in illustrative form such as in Figure 6b.In implementations leveraging machine learning for the spectral analysis processing at S104B, it may be that the underlying data is rendered into illustrative form and the resulting rendered version (such as a graph, for example) is provided as an input to the machine learning algorithm. Or, it may be that the numerical list or array is provided as input to the machine learning algorithm.Figure 6c illustrates the result of a spectral analysis performed on a further time series of data generated by the same capsule 10. Figure 6c represents motion sensor data from an accelerometer at a ten minute time window 12 hours after ingestion.The result of the spectral analysis is stored for use in the spectral analysis processing step S 104B . In the case of remote processing, the result of the spectral analysis is stored for reporting and other further processing. In the case of on-board processing, the result of the spectral analysis may be transmitted away from the capsule 10 to the receiver computing apparatus 30 for reporting and other further processing, or may be discarded after completion of S104B.The result of the spectral analysis stored for use in the spectral analysis processing step S104B may comprise a representation of signal magnitude at every component frequency in the frequency domain to which the accelerometer data is transformed at S103. Alternatively, data representing frequencies outside of a predefined set of relevant peristalsis indicator frequency ranges may be discarded. For example, power consumption on-board the capsule may be reduced by discarding some of the data at this stage.The predefined set of relevant frequency ranges may be dependent upon which of the spectral analysis processing steps is to be performed. In the case of location determination S104B, there may be a predefined set of peristalsis indicator frequency ranges.Regardless of whether S104B is performed on-board the capsule or at a remote processing apparatus, the result of the spectral analysis from S103 may be transmitted to the remote processing apparatus for storage and reporting (and optionally for S104B spectral analysis processing to be performed by the remote processing apparatus).That is, even if the spectral analysis processing, such as determination of location S104B, is to be performed on-board the capsule 10, the result of the spectral analysis may be stored by the capsule 10 on the memory hardware 152 for use in S104B and for transmitting away from the capsule by the wireless transceiver 18.Step SI 04a: Detect PeristalsisUsing the spectral analysis to detect peristalsis S104a is an example of a spectral analysis processing step. Using the spectral analysis to detect peristalsis S104a may include interrogating or otherwise processing the spectral analysis to identify signals indicative of peristalsis in one of the regions of the GI tract. Identifying a signal at a predefined minimum energy or magnitude in a frequency range indicating peristalsis (i.e. one of a predefined set of peristalsis indicator frequency ranges) may be a detection of peristalsis. Rather than a predefined minimum energy, it may be that a characteristic frequency is identified among the spectral analysis (i.e. a frequency component showing a strongest signal among the frequency components) and if the characteristic frequency is within one of a predefined set of peristalsis indicator frequency ranges, then peristalsis is detected.For example, there may be a predefined set of frequency ranges that indicate location of the capsule 10, these being the set of peristalsis indicator frequency ranges. The detecting at S104a may comprise, in the result of the spectral analysis, detecting presence or absence of a signal / component at one or more of a predefined set of peristalsis indicator frequency ranges. For example, total magnitude of signal at the one or more component frequencies within the frequency range meets a predefined minimum threshold, or an adaptive threshold. Thresholds are not required, since it may be that a characteristic frequency is extracted from the result of the spectral analysis at S103 and a determination made of whether the characteristic frequency belongs to one of the predefined set of peristalsis indicator frequency ranges.The predefined set of peristalsis indicator frequency ranges may comprise one or more from among:- a first peristalsis indicator frequency range, being a frequency range indicating stomach peristalsis;- a second peristalsis indicator frequency range, being a frequency range indicating small intestine peristalsis;- a third peristalsis indicator frequency range, being a frequency range indicating large intestine peristalsis.The first peristalsis indicator frequency range indicating stomach peristalsis may be around 3 cycles per minute. An exemplary range is from 2.5 to 3.5 cycles per minute.The second peristalsis indicator frequency range indicating small intestine peristalsis may be around 10 cycles per minute (as illustrated in Figure 6c). An exemplary range is from 9.5 to 10.5 cycles per minute, or from 9 to 11 cycles per minute, or from 8.5 to 11.5 cycles per minute.The third peristalsis indicator frequency range indicating large intestine peristalsis may be around 0.5 cycles per minute, around 1 cycle per minute, around 1.5 cycles per minute, or around 2 cycles per minute. An exemplary frequency range is from 0.5 to 2 cycles per minute, or from 1 to 2 cycles per minute, or from 0.5 to 2.5 cycles per minute, or from 1 to 2 cycles per minute.Further information may be derived from the spectral analysis, in particular where the process is repeated for a series of time periods. For example, fluctuations, variations, or anomalies within particular spectral components (frequency ranges of interest such as peristalsis indicator frequency ranges or respiration information frequency ranges or patterns, or physical activity type indicator frequency ranges) or of the characteristic frequency may be recorded and reported, either on-board the capsule 10, or in case the process is performed at a remote processing apparatus, then at the remote processing apparatus. Such further information may be valuable for health monitoring and diagnostic purposes. An example of such further information is power distribution across the entire frequency range (or across the frequency ranges of interest) in one or a series of time periods. An example of such further information is rhythm or distribution of a centre frequency across a series of time periods (optionally across the entire passage through the GI tract). An example of such further information is the identification of gaps in the peristaltic wave, that is, is there a time period or time periods where signals in one of the peristalsis indicator frequency ranges is lower than chronologically adjacent time periods, or is not detectable whereas it is detectable in time periods before and after. An example of such further information is a metric or metrics calculated from the spectral analysis from among: centre frequency, frequency spread, power distribution, frequency gaps. Any of these further information elements may be identified or measured on-board the capsule 10 or at a remote processing apparatus by processing the spectral analysis, and included in a report generated and output at S107.Generating the further information from the spectral analysis is an example of a spectral analysis processing step.Step S104B: Determine location of the capsule within the GI tractThe method of Figure 1 includes S104B determining the location of the capsule within the GI tract at the time period represented by the time series of motion sensor data.Step S 104B comprises using the spectral analysis to determine a location of the ingestible capsule within the GI tract at the time period, based on the peristalsis detection at S104a. Wherein the time period in question is the time period during which the motion sensor readings were taken that form the time series of motion sensor data on which spectral analysis is conducted at SI 03.Absence of a signal may also contribute to a location determination. Therefore, it is feasible to determine location at S 104b in the absence of a positive detections at S 104a.The ingestible capsule may be determined to be in the stomach by presence of a signal in the first peristalsis indicator frequency range. The ingestible capsule may be determined to be in the small intestine by presence of a signal in the second peristalsis indicator frequency range. The ingestible capsule may be determined to be in the large intestine by absence of a signal in the first peristalsis indicator frequency range and absence of a signal in the second peristalsis indicator frequency range. The ingestible capsule may be determined to be in the large intestine by presence of a signal in the third peristalsis indicator frequency range.In any of the spectral analysis steps S104A to S104B, in addition to identifying primary or average frequencies, further information may be extracted from the spectral analysis result, such as instability at or around a specific frequency range. Instabilities may be the result of a health condition and hence may contribute to a diagnosis. Such further information may be included in a report to be transmitted away from the capsule 10, or in the case of remote processing may be included in a report to be output to a clinician, subject, or another recipient.In Figure 6b (at 5 hours after ingestion), the spectral analysis shows a detectable spike at frequency components around 3cpm. This is an indication that magnitude(s) of the signal amplitudes (or a summation thereof) at the one or more frequency components within the first peristalsis indicator frequency range satisfies a threshold and that therefore at S104B a determination is made that, in the spectral analysis illustrated by Figure 6b, at the pertinent time window, the capsule 10 is located in the stomach. The 2ndharmonic at 6cpm is also detectable. Optionally in the determination at S104Bdetection of a signal at the 2ndharmonic may be used to confirm a detection at the first harmonic (the first harmonics being in the stated peristalsis indicator frequency ranges).In Figure 6c (at 12 hours after ingestion), the spectral analysis shows a detectable spike at frequency components around lOcpm. This is an indication that magnitude(s) of the signal amplitudes (or a summation thereof) at the one or more frequency components within the first peristalsis indicator frequency range satisfies a threshold and that therefore at S104B a determination is made that, in the spectral analysis illustrated by Figure 6c, at the pertinent time window (12 hours after ingestion), the capsule 10 is located in the small intestine.Figure 6d illustrates time series of additional sensor data. Figure 6d illustrates that the two sample determinations based on the spectral analysis results in Figures 6b and 6c are correct. Figure 6d further demonstrates the additional sensor data that may be generated by the capsule 10. Figure 6d shows sensor data from a plurality of additional sensors: capsules 10 may have no additional sensors, one additional sensor, or some subset of the additional sensors generating the data shown in Figure 6d (which corresponds to a capsule 10 hardware arrangement such as is illustrated in Figure 4). Other examples of additional sensors include pulse-oximetry sensors and EMG sensors, each of which may be housed within the ingestible capsule 10, or provided externally and affixed to a skin of the subject during GI tract passage of the ingestible capsule 10.The determination at S104b may be based upon a spectral analysis result from a single spectral analysis representing a single time window. Alternatively, the determination S104b may be based upon plural spectral analysis results each representing a different (and optionally distinct i.e. non-overlapping) time window. In the case of a single time window, a determination may be based upon meeting a minimum threshold amount of energy (or other measure of signal magnitude) at frequencies within the respective peristalsis indicator frequency range. In the case of plural time windows, it may be that a minimum threshold amount of energy (or other measure of signal magnitude) needs to be met for a predefined proportion (such as three-quarters, two-thirds, etc) of a predefined number of time windows, in order to make a positive determination of location of the capsule 10 at the associated location. Or, it may be that the characteristic frequency from the spectral analysis results of n consecutive time windows is within a particular peristalsis indicator frequency range.The minimum signal threshold to apply to a series of time periods may be, for example, a minimum summation of signal energy in the frequency range over the series of time periods, or a minimum proportion of individual time windows for which the signal in the frequency range exceeds a minimum for the individual time window. For example if an individual time window minimum is satisfied for 3 out of 4 time windows then that may result in the minimum signal threshold being satisfied, but if theindividual time window minimum is only satisfied for 2 out of the 4 time windows then that may not be sufficient.The determination may be a two-step process including first detecting S 104a a signal within the or each of the predefined set of frequency ranges, and second determining location SI 04b by comparing the detected signal with a predefined threshold minimum to determine whether the detected signal is strong enough for a positive determination of presence of the capsule 10 within the location associated with the respective frequency range.Alternatively it may be that a signal must exceed a threshold in order to be considered a detection, for example a threshold based on signal-to-noise ratio or signal-to-interference-plus-noise ratio. In such cases, the first and second steps are effectively combined.In an alternative case, a pre-trained machine learning algorithm may process the result of the spectral analysis to determine the location of the digestible capsule at the one or more time windows.Step SI 07 ReportingThe method of Figure 1 may include generating a report at S107. In Figure 1, dashed lines indicate optional steps.In processes including generating a report, the process may further comprise outputting the generated report. In implementations in which the process is conducted on-board the ingestible capsule 10 and so the report is generated on-board, the outputting may comprise transmitting the report via a wireless transceiver (for example, via Bluetooth) to a paired receiver computing apparatus 30. For example, the report may be stored at the receiver computing apparatus 30, may be combined with further reports from the same ingestible capsule 10 (i.e. from the same GI tract passage), and may be relayed to a messaging recipient or uploaded to a remote computing apparatus for storage and optionally for further analysis. A report is a message or other data artefact comprising pay load data (the specified information) and optionally also metadata for the purposes of enabling the transmission of the report itself from one device to another.Where processes generate plural reports, those reports may be combined and treated as a single report for the purposes of data transmission. For example, a single set of messaging metadata may be combined with two or more such reports as payload data, to comprise a single combined report message. Such a single combined report message may be generated at the capsule 10 and transmitted from the capsule 10 to the receiver computing apparatus 30 via Bluetooth. Such a single combined report message may be generated at the receiver computing apparatus 30 and uploaded or otherwise sent to a remotecomputing apparatus or message recipient, either by combining individual reports generated by and received from the capsule 10 or by generating the individual reports at the receiver computing apparatus 30.In any of the reports, in addition to a result or outcome of the spectral analysis processing step, the report may also include one or more from among:- the spectral analysis for the time period, and- additional sensor data readings from the same time period, and / or an outcome or processing the additional sensor data readings from the same time period such as a metric or identification of a motility marker, and / or information derived from additional sensor data readings from the same time period.A report, plural reports, or a single combined report, if generated on-board the capsule 10 by the memory hardware 152 and processor hardware 151, may be included as data transmission payload for transmission from the capsule 10 by the wireless transceiver 18 to a receiver.Location tracking and motility reportingThe method of Figure 1 may be performed on a repeated basis to monitor the relevant information (capsule location within GI tract, respiration, activity type) across a plurality of time windows. The time windows need not run continuously. The time windows may be contiguous. The time windows may be partially overlapping. To save processing and / or data transmission resources on board the capsule 10, the time windows may be separated by gaps of five, ten, fifteen, thirty, sixty minutes, for example. The time windows themselves may be of the order of one minute, two minutes, five minutes, ten minutes, depending on the implementation.The purpose of tracking the location may be to generate a motility report indicating a length of time the capsule 10 spent at each of plural locations within the GI tract such as stomach, small intestine, large intestine, or some combination thereof. Such information can be helpful in itself in gut health monitoring and diagnosis of conditions.Said report may be augmented with additional sensor data for the same time window.Combination with additional sensor dataThe capsule 10 may comprise one or more additional sensors, such as illustrated in Figure 4 and Figures 5a to 5c. Wherein additional in this context is taken to mean a sensor beyond the motion sensor 19. Noting that there may be plural motion sensors 19 (an accelerometer and a gyroscope or magnetometer) so the additional sensor may also be a motion sensor.The additional sensors may provide time series data that is also used in location determinations. Alternatively or additionally, the additional sensors may provide data for which the peristalsis detected at S 104a or the locations determined at S 104b provide context. For example, the capsule 10 may include one or more gas sensors, and there may be gases or gas concentrations that, if present in a particular location in the GI tract, are indicators of good gut health, poor gut health, or are diagnostic markers for conditions. Such conditions may include, for example, gastroparesis and SIBO.Figure 4 illustrates an ingestible capsule 10 with a plurality of additional sensors. A single one from among the illustrated additional sensors may be included in a capsule 10, or some subset of those illustrated in Figure 4.Figure 5a shows a particular selection, which includes the power source 16, motion sensor 19, wireless transceiver 18 comprising antenna 17 and directional coupler 171 to provide reflectometer readings, control circuitry 15 to control power supply to the other components and optionally also transfer of readings from the reflectometer and motion sensor 19 to the memory hardware 152, and processor hardware 151. Advantageously in the arrangement of Figure 5a the sensors are a motion sensor 19 and reflectometer, so no direct exposure to the environment is required. The arrangement of Figure 5a includes processor hardware 151. This is optional, noting that the arrangement of Figure 5a may be configured to transmit raw motion sensor data away from the capsule 10 to remote processing apparatus for generating the spectral analysis result at S103 and S104B. Alternatively the processor hardware 151 may be included and one or both of performing the spectral analysis S103 and the spectral analysis processing S104B performed on-board the capsule 10.The optional additional sensors in the capsule 10 may be one or some combination from among:- a gas sensor, which in particular may be a TCD gas sensor 131 and / or a VOC gas sensor;- a temperature sensor 14a;- a relative humidity sensor 14b;- a pulse-oximetry sensor 32;- an EMG sensor 31;- a reflectometer formed by an antenna 17 of the wireless data transceiver 18 and a directional coupler 171 (noting that one or more of these components may be present on the capsule for data communication purposes irrespective of reflectometer functionality.The capsule 10 may include one or both of a TCD gas sensor 131 and a VOC gas sensor 132. The gas sensors 13 are less than several mm in dimension each and are sensitive to particular gas constituents including oxygen, hydrogen, carbon dioxide and methane. In fact, the VOC sensor 132 may be configured to give sensor side readings and driver or heater side readings. The heater side readings maybe used to determine thermal conductivity of a surrounding gas and thereby the heater side readings of the VOC are TCD readings. The sensor side readings are used to determine concentrations of volatile organic compounds in the surrounding gases and are VOC readings. The TCD sensor 131 may be, for example, a heating element coupled to a thermopile output, with the thermopile temperature varying due to energy conducted into the gas at the location of the capsule 10. The TCD sensor 131 measures rate of heat diffusion away from the heating element.The heater side of the VOC sensor (operating as a TCD sensor) and the sensor side of the TCD sensor have different operating ranges, so TCD readings from the two sensors collectively span a wider range of operating temperatures than either of the sensors individually. Both sensors have heating elements. The TCD sensor has a low operating temperature but with a high precision. The heater side of the VOC increases the operating range but has a lower precision for TCD readings than the TCD sensor. The larger collective thermal range achieved by the two gas sensors 13 in concert enables better resolution of analytes in the second processing branch. The thermal conductivity of constituent gases in the gas mixture of the GI tract varies with temperature and so by obtaining TCD readings at different operating temperatures the different gases can be resolved from each other. This may be leveraged in measuring concentrations of constituent gases in the gas mixture surrounding the capsule 10.For ease of description the gas sensors are discussed here in the plural, though it is noted that the capsule 10 may not include any gas sensors or may include only a single gas sensor. The gas sensors are contained in a portion of the capsule 10 sealed from the power source 16 and other electronic components by a membrane 111. The outer surface of this portion of the capsule may be composed of a selectively permeable membrane 11 , or there may be an aperture allowing the gas mixture surrounding the capsule 10 to enter into the portion of the capsule 10 housing the gas sensors (and the temperature sensor 14a and relative humidity sensor 14b, where included). For example, the gas sensors include respective heaters which are driven to heat sensing portions of the respective gas sensors to temperatures at which sensor readings are obtained (i.e. a measurement temperature). The heaters may be driven in pulses so that there is temporal variation in the sensing portion temperature and so that measurement temperatures are obtained for periods sufficient to take readings but without consuming the power that would be required to sustain the measurement temperature continuously.The gas sensors may be calibrated, so that a gas sensor reading can be used to identify the composition and concentration of a particular gas. Calibration coefficients are gathered in manufacturing and applied to the recorded readings at the processing stage (i.e. by a server such as on the cloud or on-board the capsule 10). Otherwise, this calibration could be performed on the capsule 10, at the receiver computing apparatus 30, or on any device having access to the calibration coefficients and the recorded readings from the gas sensors. Such calibration relates to processing concerned with measuring the concentrationof constituent gases in the gas mixture at the capsule 10. Context for the outputs of that processing may be provided by the determinations made at S104b providing a location of the capsule 10 within the GI tract at which said gas mixture is found. Instead of calibration, the gas sensors may be used on a relative measure basis, where there is no formal pre-calibration and it is simply the variability in readings that is used. Alternatively the gas sensors could be calibrated on-the-fly using stomach data as a baseline.Additional sensor data may be processed to identify one or more motility markers to use as part of the determination at S104b. More detail on the detection and precise form of said motility markers is set out in PCT / AU2022 / 051270, to which reference may be made. A summary is presented below for ease of reference.Figure 5b illustrates a capsule arrangement in which the additional sensor hardware comprises an EMG 31. The EMG 31 may be on-board the capsule 10 or may be separate from the capsule 10 but provided as part of a system comprising the capsule 10 and the EMG 31. The capsule 10 may include an EMG (electromyographic) sensor comprising a pair of electrodes on an external surface of the ingestible capsule 10 at either end. The EMG sensor measures electrical activity in a medium or on a surface by measuring a potential difference across the pair of electrodes. Alternatively, the EMG sensor 31 may be separate from the capsule 10, and may comprise two or more electrodes affixed to the skin of the subject at locations corresponding to the GI tract. In the case of the EMG sensor 31 being on-capsule, the sensor data from the EMG sensor 31 may be processed on-board the capsule 10, or may be transmitted to the remote computing apparatus for processing. In the case of the skin-mounted EMG sensor 31, the sensor data, or one or more results from the processing of the sensor data by a computing apparatus in receipt of the said sensor data, may be transmitted to the wireless transceiver 18 of the capsule and then used by the on-board processor hardware 151.By comparing contemporaneous EMG sensor data and spectral analysis results, electrical signals triggering muscular contractions can be compared with mechanical movements resulting from muscular contractions, and results of the comparison used in health assessments and diagnostics.As with the additional sensor data discussed above, the EMG sensor data may be used to add confidence to location determinations made from the results of the spectral analysis of the motion sensor data, or may be used in reporting to add context to results, or in diagnostics that utilise the location determination.One or more from among:the motion sensor data;the spectral analysis result;location determinations;peristalsis detections;respiratory information;gait analysis information;activity type determination;the additional sensor data; anddetected motility event timings;may be included in a report. The reported data are contemporaneous, meaning that additional sensor data readings in a time series representing a time window are included in a report also comprising a spectral analysis result of the motion sensor data for the same time window, and / or an outcome or result of processing the said spectral analysis result. The report may be compiled on-board the capsule 10 (by the memory hardware 152 and processor hardware 151) and transmitted away during passage through the GI tract or in a burst or inquiry mode following detection of excretion (for example by detecting a freefall event), or the report may be compiled by the remote computing apparatus. Or the report may be compiled on-board the capsule 10 and by the remote computing apparatus in combination.Figure 5c illustrates a pulse-oximetry sensor 32 as additional sensor hardware. As with the EMG sensor 31, the pulse-oximetry sensor 32 may be provided on-board the capsule 10 or as a separate part of a system also including the capsule 10. For example, the pulse-oximetry sensor may be finger-worn. In the case of the on-board pulse-oximetry sensor, the capsule housing 11 may comprise a window or aperture via which LED light is transmitted and reflected light sensed. Alternatively, the pulse-ox sensor 32 may comprise a light transmitter and sensor affixed to the exterior surface of the capsule housing 11.Sensor data from the pulse-oximetry sensor may be included in a report generated by the capsule 10, or the capsule 10 in cooperation with the remote processing apparatus. The report may be used by a clinician for diagnostic purposes. For example, in respiration monitoring implementations, influence of disordered respiration or other respiration events detected in the result of the spectral analysis may be combined with the pulse-oximetry data. In particular, pulse-oximetry data from the small intestine may be particularly accurate due to the highly vascularized surfaces of the small intestine. Optionally, a first on-the-fly determination of location in the small intestine at S104b may be a trigger for the microcontroller 15 to turn on the pulse oximetry sensor 32 and start taking readings.Figure 6d illustrates additional sensor data from a live phase of an ingestible capsule, that is, a powered-on phase of a capsule 10 within the GI tract of a subject. An ingestion event (marked ‘I’) may be determined from the additional sensor data, for example the rise in relative humidity, or may be determined from a user interaction with a user interface on the receiver computing apparatus 30.Hydrogen concentration measurements are a metric derivable from TCD gas sensor readings, appropriately calibrated. H2 levels or TCD gas sensor readings themselves may be used as a basis for a gastric -duodenal transition indicator (marked ‘GDJ’). H2 levels may be sensed directly or may be derived, such as derived from TCD gas sensor readings.CO2 concentration measurements are derivable by appropriate calibration of the TCD gas sensor readings. The ICJ indicator (marked ‘ICJ’) being a steep rise (i.e. positive gradient above a predefined threshold) in CO2 concentration. PCT / AU2022 / 051270, at

[0300] to

[0303] , explains in more detail how by operating the TCD gas sensor at different sensing temperatures, different gases may be resolved. Extra information may be added by driving the VOC gas sensor heater side to take TCD measurements therefrom.The gastric -duodenal transition event may be detected in the TCD gas sensor readings as a spike, step change or an inflection point in the TCD gas sensor readings. A correction may be applied to the TCD gas sensor readings to account for changes in environmental temperature, based on recorded readings from the environmental temperature sensor 14a.An ICJ indicator may be detected by identifying an increase in (sensor side) VOC gas sensor output exceeding a predefined threshold. Alternatively ileocecal junction transition indicator may be detected by identifying an increase in a metric derived from VOC gas sensor output such as CO4 concentration (marked ‘fermentation’ in Figure 6d). Optionally, a further criterion may be applied such as presence of a contemporaneous, or temporally adjacent to within a predefined temporal distance either side, increase in measured H2 levels exceeding a predefined threshold. Noting that H2 levels are determined from the TCD gas sensor output and / or heater-side VOC sensor output.In the location determination processing, the determination at S104b may be based on the spectral analysis from S103 and peristalsis detections at S104a alone, or in combination with data from the one or more additional sensors. The data from the additional sensors may provide a motility marker being either an indication either of a location at which the capsule 10 is located (for example reflectometry measurements of the medium surrounding the capsule 10) or that a motility event has occurred (for example an inflection point, spike, or step change in concentration of a particular gas at a transition between two parts of the GI tract). Figure 6d illustrates timings of three motility markers T (ingestion), ‘ICJ’ (ileocecal junction transition indicator), and ‘GDJ’ (gastric duodenal junction transition indicator).Such motility markers from the additional sensor data may be used in a deterministic way at S 104B, for example it may be that a determination that the capsule 10 is in the small intestine at a time window based on the spectral analysis from S 103 can only be made if a gastric emptying event has been detectedin the additional sensor data preceding that time window. Likewise, it may be that a determination that the capsule 10 is in the large intestine at a time window based on the spectral analysis from S103 can only be made if an ileocecal junction transition event has been detected in the additional sensor data preceding that time window.Alternatively, the spectral analysis from S103 and the additional sensor data may be provided to a pretrained machine learning algorithm to classify capsule location at a particular time window.Some calibration may be required in seeking to find motility markers in the additional sensor data such as gastric -duodenal transition indicators, since ingested foodstuffs at different temperatures change the environmental temperature in the stomach, which influences rate of heat diffusion. In the case of gas sensor readings taken after ingestion and before the gastric -duodenal transition (i.e. whilst the capsule 10 is in the stomach), processing of readings may include applying a moderation to TCD readings, from either gas sensor, in order to correct for variations in environmental temperature, based on environmental temperature readings by the environmental temperature sensor 14a. TCD readings are effectively measuring rate of heat loss to surroundings, and so accuracy is improved by measuring the temperature of the surroundings rather than by relying on assumption (i.e. prior knowledge of internal temperature of the subject mammal). However, the processing may rely on assumption, for example, if the capsule 10 does not include an environmental temperature sensor 14a or if there is some issue with the environmental temperature sensor readings, or, for example, if the level of accuracy provided by assumption is acceptable in a particular implementation. Gastric temperature may vary based on, for example, ingestion of liquids or foodstuffs by the subject mammal, or physical activity undertaken by the subject mammal 40. Environmental temperature is a term used in this document to refer to the temperature of the environment in which the capsule 10 is located, as distinct from operational temperatures of the gas sensors. The sensitivity of the gas sensors to different constituent gases vary according to the operating temperature of the sensors and the processing of the readings includes calibrating (also referred to as moderating or correcting) readings from the gas sensors according to contemporaneous operating temperature and optionally also according to contemporaneous environmental temperature.In addition to the gas sensors 13 and the environmental sensor 14 (being a temperature sensor 14a and / or a humidity sensor 14b), the capsule electronics further include a microcontroller 15 or some other form of control circuitry, a power source 16, an antenna 17 or plural antennae, a wireless transceiver 18 or plural wireless transceivers, and optionally a reed switch (though in the case of there being two wireless transceivers the reed switch may be omitted) or some other mechanism to initiate data recording. The wireless transmitter 18 may operate in concert with the antenna 17 of the primary transceiver to transmit a data transmission payload including readings from the sensors (collectively referring to the gas sensors13 and the environmental sensor 14) to a receiver apparatus 30 and / or a remote computer 20 for processing. Alternatively, sensor data may be processed on-board and results transmitted away via the wireless transmitter 18.Figure 4 illustrates the primary transceiver antenna 17 and directional coupler 171 as elements of the wireless transmitter 18, since the antenna is the physical means by which the wireless transmitter 18 transmits data to the receiver apparatus 30. The wireless transmitter 18 is also configured to buffer data for transmission. The wireless transmitter 18 may also be configured to encode the data with a code unique to the capsule 10 among a population of like capsules 10.Figure 7 illustrates an exemplary reflectometer arrangement.Optionally, the capsule 10 includes a reflectometer comprising a transmission antenna 17 connected in series with a directional coupler 171 configured to measure a reflected signal from the transmission antenna 17, the output signal output by the sensing mechanism comprising accelerometer readings and / or reflectometer readings.Optionally, the capsule 10 includes the reflectometer, and the ingestible capsule further comprises a diode detector and the diode detector forms a part of the reflectometer, the diode detector being configured to receive the reflected signal from the antenna and to measure an amplitude of the reflected signal, the reflectometer readings in the output signal comprising amplitude measurements of the reflected signal.Optionally, the ingestible capsule 10 further comprises a quadrature demodulator and the quadrature demodulator forms a part of the reflectometer, the quadrature demodulator being configured to receive the reflected signal from the antenna via the directional coupler and to extract phase information of the reflected signal relative to a carrier signal, the reflectometer readings in the output signal comprising the extracted phase information of the reflected signal.Optionally, the ingestible capsule 10 further comprises an antenna impedance control mechanism comprising a variable capacitor configured to vary impedance of the transmission antenna, and a controller, wherein the reflectometer and the antenna impedance control mechanism form a closed loop or feedback loop, and wherein the controller is configured to receive the measurements of the amplitude of the reflected signal from a diode detector and to execute a control algorithm to use the amplitude measurements to generate an antenna impedance control signal setting a capacitance of the variable capacitor 172 to vary impedance of the antenna to reduce amplitude of the reflected signal, wherein the reflectometer readings in the output signal comprise readings of the antenna impedance control signal.Optionally, the closed loop or feedback loop further comprises a quadrature demodulator, and wherein phase information is extracted by the quadrature demodulator and output to the controller, and wherein the controller is configured to use the amplitude information and the phase information to generate the antenna impedance control signal.Optionally, the additional sensor data comprises reflectometer readings, and a motility marker may be detected therein. Specifically, an ileocecal junction transition indicator may be detected in readings from the reflectometer. Processing includes: processing the reflectometer readings to identify the presence of an ileocecal junction transition indicator in the reflectometer readings.Figure 8 illustrates readings from a directional coupler 171 of a reflectometer and in particular illustrates step changes coinciding with the detected gastric duodenal transition indicator and the ileocecal junction transition indicator. With appropriate calibration the presence of the step change may be detected to directly detect one or both transition events. Furthermore, the antenna reflectance values themselves pre- or post- the transition events may be used as an indicator of capsule location.Interconnections between electronic components may be via a central bus. This is one example of how power and data may be distributed between components. Other circuitry architecture may be implemented, for example, all connections may be via the microcontroller 15 which coordinates distribution of data and power between components. The sensors (the TCD sensor 131, the VOC sensor 132, the environmental sensor 14, the motion sensor 19, the EMG 31, the pulse-ox sensor 32, and / or the directional coupler 171) take readings under the instruction of the microcontroller 15, powered by the power source 16, and transfer the readings to the wireless transmitter 18 for transmission to the receiver apparatus via the antenna 17.The dimension of the capsule may be less than 11.2 mm in diameter and 27.8 mm in length. The housing of the capsule 10 may be made of indigestible polymer, which is biocompatible. The housing may be smooth and non-sticky to allow its passage in the shortest possible time and to minimise risk of any capsule retention.Detecting motility Markers in the Time Series of Motion Sensor DataDetected motility markers described here may be utilised in determining release timing of therapeutic matter, or determining sampling timing.Figure 9 shows time series of accelerometer readings as ‘roll’ in each of three mutually orthogonal dimensions and is marked with gastric emptying event, from which it can be seen that the change in accelerometer readings correlates temporally with a gastric emptying event. Therefore, accelerometertime series data may be utilised to provide an indicator of timing of a gastric-duodenal transition event which may be used to add confidence to a location determination at S104b.The capsule orientation is measured using a triaxial accelerometer and tracking the gravity vector with respect the capsule frame of reference. The capsule orientation is measured using a triaxial accelerometer and tracking the gravity vector with respect to the capsule frame of reference. When the capsule leaves the stomach it tends to experience rapid changes in its orientation as it transits through the duodenum and small intestine.Different techniques may be used for processing the time series of accelerometer data. Metrics may be calculated from the raw readings, from which metrics one or more motility markers are detectable. A first technique is “angle travelled”, which accumulates the orientation change in excess of a 90 degree hysteresis angle. This technique tends to be robust to small changes in orientation experienced in the stomach and avoids some of the complexities of other approaches.Angle travelled uses vector mathematics to calculate the angle between the gravity vector and a temporary vector. The temporary vector is pulled in the direction of the change in angle, only when this angle exceeds a given threshold (currently 90 Deg). It is then the accumulation of the change in the temporary vector that is visualized in the representation from which markers are identifiable. What is generally seen is that this measure does not change much in the stomach since the angle between the gravity and temporary vectors rarely exceed the threshold in any one direction, (small back and forth orientation changes in the stomach are effectively ignored by the inherent hysteresis of this algorithm) and that once in the tortuous lumen of the small intestine, this measure accumulates significantly due to the larger, more continuous orientation changes of the capsule. Thus, a step change in the cumulative angle travelled measure is a gastric -duodenal transition indicator.In an exemplary implementation of angle travelled: the accelerometer readings may provide a reading of an orientation of the digestible capsule relative to a frame of reference in fixed relation to a gravitational vector. Processing of the readings from the accelerometer may comprise recording an orientation of the ingestible capsule given by a first accelerometer reading as a reference orientation, and repetitively in respect of each successive accelerometer reading chronologically: determining whether the orientation of the ingestible capsule given by the respective accelerometer reading is more than a threshold angular displacement from the reference orientation, and if the threshold angular displacement is not met, progressing to the next accelerometer reading without changing the reference orientation, and if the threshold angular displacement is met, changing the reference orientation to align with the orientation of the ingestible capsule given by the respective accelerometer reading. Anindicator, such as the gastric -duodenal transition indicator, may be a step change in the rate of change of the reference orientation.A second technique for processing accelerometer data may be referred to as total roll. Total roll calculates the angle between the gravity vector and each of the capsule X, Y and Z axes and expresses this as a continuous measure that can accumulate beyond 360 Deg. For example, if the capsule x axis is at an angle of 350 Deg and rotates by a further 20 Deg, the resulting angle is expressed as 370 Deg rather than 10 Deg. This helps when representing the readings as a plot from which markers are identified since it avoids the sudden angle changes associated with crossing the zero line. In the example a real change of 20 Deg would be visualized instead of an artificial change of 340 Deg. In addition to this basic approach, low pass filtering may be applied to filter the raw data to remove sensor noise. Additionally, angles are only calculated when the raw accelerometer data provide sufficient data to calculate a meaningful angle. An example of where this is not the case is when the two accelerometer axis values used to calculate the orientation angle around the third axis both approach zero. In this case the calculation will be dominated by sensor noise and so a meaningful angle cannot be determined.The accelerometer readings provide a reading of an orientation of the ingestible capsule relative to a frame of reference in fixed relation to a gravitational vector. Exemplary processing of the readings from the accelerometer may comprise for each of three orthogonal axes in fixed spatial relation to the ingestible capsule derivable from the reading of the orientation, repetitively in respect of each successive accelerometer reading chronologically: calculating, as a scalar value, a change in the orthogonal axis relative to the gravitational vector from the preceding accelerometer reading; applying a low pass filter to the calculated changes; recording the cumulative filtered calculated changes. A marker serving as a gastric -duodenal transition indicator may be, for example, an increase (such as a spike or step change) in the rate of increase in the cumulative filtered calculated changes.Power SourceThe ingestible capsule 10 includes a power source. In the arrangements illustrated at Figures 2, 3a, 3b, 4, 5a to 5c, and 12 the power source may be a battery or may be a supercapacitor.Companion Device for Prescribed Incretin Mimetics including GLP-1 receptor agonists Embodiments include a method comprising, administering an ingestible capsule as defined or described anywhere in the present disclosure, to a patient undertaking an incretin mimetic (such as a GLP-1 receptor agonist) dosage regime; monitoring a location of the capsule within the GI tract; determining a gastro-intestinal motility indicator based on locations of the ingestible capsule within the GI tract determined by the ingestible capsule or determined by processing data obtained by a sensor or sensorson-board the ingestible capsule; and titrating or otherwise modifying the incretin mimetic (such as a GLP-1 receptor agonist) dosage regime independence upon the gastro-intestinal motility indicator.Embodiments include a method comprising, for a patient undertaking a medicament dosage regime, the patient having ingested or been administered an ingestible capsule configured to determine and report the location of the ingestible capsule within the GI tract, or to cooperate with a computing device to determine and report the location of the ingestible capsule within the GI tract: determining a time series of locations of the capsule within the GI tract; determining a gastro-intestinal motility indicator based on the time series of locations of the ingestible capsule within the GI tract; and titrating or otherwise modifying the medicament dosage regime in dependence upon the gastro-intestinal motility indicator. Optionally, the medicament may be an incretin mimetic such as a GLP-1 receptor agonist, or any other substance that may affect gastrointestinal transit times as a mechanism of action or side effect.Embodiments include a method comprising, administering an ingestible capsule configured to diagnose gastroparesis or suspected gastroparesis in a subject; determining a gastro-intestinal motility indicator based on a positive or negative diagnosis of gastroparesis or suspected gastroparesis in the subject; and titrating or otherwise modifying the incretin mimetic (such as a GLP-f receptor agonist) dosage regime in dependence upon the gastro-intestinal motility indicator.Where references are made in the present document to incretin mimetics, it is to be understood to include GLP-1 receptor agonists and other compounds and medicaments falling within the incretin mimetic group of medications. Examples include dipeptidyl peptidase-4 (DPP-4_ inhibitors, long-acting GLP-1 analogs, GLP-1 derivatives, peptide-based antidiabetic agents, peptidic GLP-1 receptor modulators, exogenous GLP-1 compounds, synthetic GLP-1 analogs, modified incretin peptides. Further, the same techniques may be applied to other medicaments that may affect gastrointestinal transit times as a mechanism of action or side effect.The gastro-intestinal motility indicator may be a small-intestinal motility indicator.Incretin-mimetics, such as GLP-ls and GLP-1 RAs may cause gastroparesis as a side effect, and / or may cause slowing of GI tract motility in the small bowel and / or large bowel. Embodiments provide non-invasive, cost-effective, and accurate mechanisms to measure GI tract motility at any or all of the stomach, small intestine, and large intestine, in order to provide information to a clinician to determine whether or how to titrate the dosage regime of the incretin mimetic.The patient may be preoperative, and the titrating or otherwise modifying the Incretin mimetic (such as a GLP-1 receptor agonist) dosage regime is to promote stomach emptying in advance of the surgery, in dependence upon the gastro-intestinal motility indicator.The patient may have an operation or procedure scheduled a period of weeks or months away, during which there is a risk of aspiration and so an empty stomach is desirable or is a medical requirement. Since incretin mimetics and in particular GLP-1 receptor agonists slow gastric emptying, it may desirable to administer ingestible capsules such as disclosed and defined in the present disclosure to the patient periodically (such as daily or weekly) over a period of time (such as greater than 2 weeks, greater than four weeks, greater than six weeks) in advance of the surgery or procedure in order to assess an extent to which gastric emptying is delayed / slowed, and to titrate the dosage regime of the GLP-1 receptor agonist or other incretin mimetic compound over the period of time to approximately zero at or in advance of the surgery so that timely gastric emptying is achieved.The patient may be diagnosed with type-2 diabetes.In a treatment regime, a patient might be prescribed incretin mimetics such as a GLP-1 receptor agonist for the treatment of, for example, type-2 diabetes or obesity. The prescription might prescribe an ingestible capsule such as defined or described in the present disclosure in addition to the incretin mimetics. The ingestible capsule may be prescribed to be ingested periodically (such as daily, weekly, monthly, fortnightly, once every two days, once every 3-10 days, once every 3-11 days, once every 3-12 days, once every 3-13 days, once every 3-14 days, once every 3-15 days, once every 3-16 days) and be configured (on its own or in cooperation with a computing device to which it transmits sensor readings) to measure and report a GI motility indicator comprising gastric emptying time (relative to capsule ingestion); and optionally also comprising one or more from among:small bowel transit / residence time;whole gut transit time;large intestine residence time;gastric residence timeingestion time;excretion time;gastro-duodenal junction transit time;ileocecal j unction transit time ;combined small and large intestine residence time;combined stomach and small intestine residence time;orocecal residence time.The ingestible capsule may itself calculate the GI motility indicator, or it may be calculated off-board based on data recorded by the capsule and transmitted away.In addition to administering the ingestible capsule, the patient may be provided with access to a software application for a personal electronic device such as a smartphone, table, personal computer, or smartwatch. The application is configured to receive inputs from the patient recording one or more from among vomiting, cramps, constipation, self-reporting mood, self-reporting stress, bowel movements, nausea, diarrhoea, medical symptoms, dietary events such as ingestion of food and drink. Inputs may also be received via a wearable such as a chest strap or smartwatch, either by having a user interface allowing the user to self-report the inputs, or by detecting the inputs. By providing the software application with appropriate permissions either to the wearable itself or to other applications storing the inputs on the device or at connected servers, the software application is able to access the inputs. The inputs may be combined with the GI motility indicator determined using the ingestible capsule in order to determine how to titrate or otherwise modify the incretin mimetics dosage regime, such as the GLP-1 receptor agonist dosage regime.The clinicians themselves, an algorithm configured for the purpose, or an appropriately trained machine learning algorithm, may determine how to titrate or otherwise modify the incretin mimetics dosage in dependence upon the inputs comprising the GI motility indicator and optionally also the inputs received via the application disclosed above.Titrating may be interpreted as continuously measuring and adjusting (the dosage regime). For example, a dosage regime for a GLP-1 receptor agonist may comprise a weekly dose within a defined range, such as 0.75mg to 4.5mg, or between 0.25mg and 5mg, or between 0.25mg and 4.5mg. The clinician or algorithm may respond to gastric emptying timings (i.e. determined time between the ingestible capsule being ingested and passing the gastric -duodenal junction) within defined ranges by determining to increase the weekly dosage, hold or maintain the weekly dosage, or decrease the weekly dosage. It may be that increase / decrease is only ever by a fixed increment, such as 0.5mg (without exiting the defined range), or it may be that there are ranges of gastric emptying timings that determine the increment or decrement should be greater. Similarly, determining presence or absence of gastroparesis may determine whether to maintain, increase, or decrease, the weekly dosage (of the GLP-1 receptor agonist or other incretin mimetic). Presence of gastroparesis may be a determinant factor in decreasing weekly dosage. Absence of gastroparesis may be a criterion in determining that weekly dosage is to be increased, if it is below a desired level for the patient, but not otherwise.Gastric emptying time (post-ingestion) greater than four hours, greater than six hours, greater than eight hours, greater than ten hours, or greater than twelve hours, may satisfy a threshold for decreasing weeklydosage. Alternatively, the threshold may be greater, such as at 14, 16, 18, 20, 22, or 24 hours. Gastric emptying time below six hours, or below four hours, may satisfy a threshold for increasing weekly dosage. Alternatively, the threshold for increasing weekly dosage may be greater, such as 8 hours or 10 hours. The selection of thresholds is patient-specific and depends upon the nature of the ailment being addressed by the dosage regime, and may depend on other factors.An amount of increment or decrement may be 0.25mg per week or 0.5mg per week, or Img per week. The amount of increment or decrement may depend upon the ailment being addressed by the dosage regime. The amount of increment or decrement may be dependent upon BMI of the patient, or bodyweight of the patient. The amount of increment or decrement may be dependent upon age of the patient. The amount of increment or decrement may be dependent upon the reporting of any other side negative side effects by the patient.Rather than increment or decrement, value ranges of gastric emptying time may be associated with absolute dosage levels. So, for example, gastric emptying time exceeding 12 hours determines the weekly dosage be set to 2mg, or to not greater than 2mg.Additionally or alternatively, the titration of the dosage regime may be dependent upon a trend in the measured gastric emptying times from the patient over a series of ingestible capsules, spaced apart in ingestion time. For example, the patient may be administered an ingestible capsule daily or weekly (or, for example, once every two or three days) and the clinician or algorithm then has a time series of either gastroparesis presence / absence determinations or gastric emptying time measurements in accordance with which to determine how to modify a dosage regime. The algorithm functions as a PID controller or typical control system that uses gradient and a target gastric emptying time or targets determination of absence of gastroparesis, to determine how to modify dosage regime. For example, a PID controller where GET is the measured output and dose is the control signal.In addition to controlling amounts of dose, the algorithm or clinician could titrate by changing time period between dosages, which could be implement via a software application issuing notifications to the patient to self-administer a dosage.That is, gastric emptying time and / or presence / absence of gastroparesis may be the only factor in determining how to modify the dosage regime within a defined range, or may be a factor among others. Such other factors may include one or more from among: is the patient losing or gaining weight; is the patient experiencing regular bowel movements; is the patient experiencing or reporting any adverse side effects from a present dosage regime; where is the patient’s current dosage with respect to a clinically-determined desired ongoing dosage level for the patient. Presence or absence of gastroparesis may bereported by the ingestible capsule along with a confidence level, which confidence level may be an input to the clinician or algorithm in titrating a dosage regime.The ingestible capsule may an ingestible capsule disclosed or defined in the present disclosure, or as disclosed in one from among:- PCT / AU2022 / 051270;- PCT / AU2022 / 051389;- PCT / AU2023 / 050801;- PCT / AU2023 / 050802;- PCT / AU2023 / 050803;- PCT / AU2023 / 050804;- PCT / AU2025 / 050151;the entire contents of which are herein incorporated by reference.In a particular example, the ingestible capsule may be: an ingestible capsule device comprising: an ingestible indigestible bio-compatible housing; and, within the housing: a power source; sensor hardware including gas sensing apparatus; processor hardware; memory hardware; and a wireless data transmitter; the memory hardware storing processing instructions which, when executed by the processor hardware, cause the processor hardware to perform a process comprising: obtaining data representing a time series of readings from the gas sensing apparatus, the time series of readings being taken during exposure of the gas sensing apparatus to a gas mixture at the ingestible capsule device during passage of the ingestible capsule device through a gastrointestinal tract of a subject, the subject having orally ingested the ingestible capsule device, each reading having a value, the values of the readings being sensitive to CO2 concentration in the gas mixture; processing the readings to detect one or more gastroparesis indicator spikes in the CO2 concentration with respect to time, a gastroparesis indicator spike being a spike in the CO2 concentration with respect to time at a timing after an ingestion timing of the ingestible capsule device and preceding a gastric -duodenal transition timing of the ingestible capsule device. In this example, the gastroparesis indicator spikes may be indicative of gastroparesis, with the gastrointestinal motility indicator (in dependence upon which the incretin mimetics dosage regime is modified) being detection of presence or absence of the gastroparesis indicator spikes, or a diagnosis of gastroparesis or suspected gastroparesis made in dependent upon the detection of presence or absence of the gastroparesis indicator spikes.The above presents two broad techniques for utilising ingestible capsules to inform an incretin mimetics dosage regime. In the first technique, motility is directly measured by the ingestible capsule, and a metric or indicator of GI tract motility is used to inform (i.e. to titrate or otherwise modify) the incretin mimetics dosage regime. In the second technique, gastroparesis itself, or suspected gastroparesis, isdetected by the ingestible capsule by measuring C02 concentration, and either the CO2 measurements or the diagnosis of gastroparesis is utilised as an indicator of GI tract motility to inform the incretin mimetics dosage regime. Methods, programs, and apparatus for detecting gastroparesis or suspected gastroparesis by an appropriately configured ingestible capsule are disclosed in PCT / AU2023 / 050803, the entire contents of which are herein incorporated by reference. The incretin mimetics may be a GLP-1 receptor agonist. In either case, the skilled reader will understand that the same technique may be applied to another medicament not falling within the scope of incretin mimetics but which may affect gastrointestinal transit times as a mechanism of action or side effect.Advantageously, since incretin mimetics such as GLP-1 receptor agonists are prescribed with an intention of slowing gastric emptying and small intestine motility in a patient, the accompanying administration of an ingestible capsule to determine a GI tract motility indicator provides a mechanism to measure the efficacy of the dosage regime. The use of the ingestible capsule to determine the GI tract motility indicator provides the clinician with information they can use to implement gradual GLP-1 receptor agonist dose escalation in order to improve tolerability and help patients adjust to the medication. This helps to avoid scenarios in which patients are given incretin mimetics dosages (such as GLP-1 receptor agonist dosages) causing nausea, vomiting, and diarrhea, which can lead to treatment discontinuation. The use of the ingestible capsule to determine the GI tract motility indicator allows clinicians to assess incretin mimetics dosage impact on motility to improve tolerability and establish an individualised dose.It is noted that where reference is made to the use of an ingestible capsule configured to calculate the GI tract motility indicator, it is considered to include ingestible capsules configured to take sensor readings from within the GI tract and to transmit them to a companion device (such as a mobile phone or connected computer) at which the GI tract motility indicator is determined.The patient may be diagnosed with type-2 diabetes. Gastroparesis is associated with uncontrolled diabetes. The use of the ingestible capsule to determine the GI tract motility indicator enables a clinician to monitor a diabetic patient to assess for presence or absence of gastroparesis.A goal of the dosage regime titration or other modification may be to avoid gastroparesis occurring in the patient. Therefore, administering ingestible capsules configured to detect gastroparesis alongside a GLP-1 receptor agonist treatment regime enables a clinician to control dosage in accordance with presence or absence of gastroparesis as reported by the ingestible capsule. For example, to increase dosage while there is an absence of gastroparesis, and to respond to a detection of presence of gastroparesis by decreasing a dosage.Training data for a machine learning algorithm may be patient case studies detailing incretin mimetics dosage regimes (against time), and may detail dosage regimes of a specific incretin mimetic such as a GLP-1 receptor agonist, one or more determined GI tract motility indicators (and timings thereof), and indications of outcomes in terms of tolerability (for example, indications that the patient did or did not tolerate the dosage regime at a particular time. By being trained with sufficient case studies covering both tolerable and intolerable dosage regimes, and with an aim of predicting whether a change in dosage regime will be tolerated or not, the machine learning algorithm may be leveraged in determining a new dosage regime.In another example, the machine learning algorithm may be a generative adversarial network in which a generative Al algorithm is trained to generate a dosage regime, and a classifier trained (as in the previous paragraph) to predict tolerance or intolerance. In this way, the generator network, once trained, is configured to determine new dosage regimes.Presurgical Medical UseEmbodiments include a method of evidencing gastric emptying in a patient, the method comprising administering an ingestible capsule as defined or described anywhere in the present disclosure to a patient; monitoring a location of the capsule within the GI tract; determining, based on the monitored location, that the ingestible capsule has passed the gastro -duodenal junction; and, based on the determining, outputting a report indicating that the ingestible capsule has passed the gastro-duodenal junction.Embodiments include a method of evidencing gastric emptying in a patient, the method comprising: administering to a patient an ingestible capsule housing a motion sensor configured to generate a time series of motion sensor data representing motion of the ingestible capsule during passage through the GI tract of a subject; at data processing hardware communicably coupled to the motion sensor, following ingestion of the ingestible capsule by the subject, processing the time series of motion sensor data to determine that the ingestible capsule has passed the gastro-duodenal junction, and outputting a report indicating that the ingestible capsule has passed the gastro-duodenal junction.Embodiments include a method of evidencing gastric emptying in a patient, the method comprising: administering to a patient an ingestible capsule housing a motion sensor configured to generate a time series of motion sensor data representing motion of the ingestible capsule during passage through the GI tract of a subject; at data processing hardware communicably coupled to the motion sensor, following ingestion of the ingestible capsule by the subject, receiving from the ingestible capsule or from data processing hardware processing sensor data generated by the ingestible capsule from withinthe GI tract of the patient, a report indicating that the ingestible capsule has passed the gastro-duodenal junction.The patient may be pre-surgical, and the ingestible capsule administered to be ingested along with a final intake of food and / or drink prior to surgery.The method may include, until the report indicating that the ingestible capsule has passed the gastroduodenal junction is output, outputting an alert indicating that gastric emptying is still pending. Advantageously, risk of aspiration during surgery (or another medical procedure) is suppressed by making surgery conditional upon gastric emptying of the final intake of food prior to surgery. Optionally, admission for surgery may be conditional upon report of gastric emptying (i.e. passing of the gastro-duodenal junction by the capsule) being received.A pre-surgical checklist may include a criterion that a report of gastric emptying of final food and / or drink intake prior to surgery has been reported by the ingestible capsule (or connected computing device).The ingestible capsule may be as disclosed or defined in the present enclosure. Specifically, the ingestible capsule has an on-board accelerometer or another motion sensor. The ingestible capsule administered to the patient may be, or may be part of, an apparatus as set out below:An apparatus comprising: an ingestible capsule housing a motion sensor configured to generate a time series of motion sensor data representing motion of the ingestible capsule during passage through the GI tract of a subject; data processing hardware communicably coupled to the motion sensor, the data processing hardware being configured, following ingestion of the ingestible capsule by the subject, at data processing hardware communicably coupled to the motion sensor, to perform a process comprising: processing the time series of motion sensor data to determine that the ingestible capsule has passed the gastro-duodenal junction, and outputting a report indicating that the ingestible capsule has passed the gastro-duodenal junction.The processing the time series of motion sensor data may comprise generating a spectral analysis of the time series of motion sensor data generated by the motion sensor over a time period during passage of the ingestible capsule through the GI tract, using the spectral analysis to detect peristalsis at a location of the ingestible capsule within the GI tract at the time period. The processing the time series of motion sensor data may comprise calculating a metric representing total roll or tumbling of the motion sensor data in the time domain, and determining passing the gastro-duodenal junction by a step change in themetric or by a prolonged period of time for which the metric is in a range indicating presence in the small intestine.The determination may be a direct determination of passing the gastro-duodenal junction by first determining residence of the digestible capsule in the stomach, and subsequently determining residence of the ingestible capsule in the small intestine or in another portion of the GI tract downstream of the stomach. Alternatively, the determination may be an indirect determination of passing the gastroduodenal junction by determining residence of the ingestible capsule in the small intestine or in another portion of the GI tract downstream of the stomach independent of any earlier determination of residence in the stomach. That is, indirect determination may be an inference that, because the sensor readings from the ingestible capsule indicate that the ingestible capsule is resident in the small intestine or another portion of the GI tract downstream of the gastro-duodenal junction, it must have passed the gastro-duodenal junction.Optionally, the determination (of passing the gastro-duodenal junction) may be based on a spectrograph or spectral analysis of the motion sensor data. In this example, sensing a signal in a frequency range (for example 8-12 cpm, 7-14 cpm, lOcpm, 5-20 cpm) associated with small-intestinal peristalsis at a signal strength satisfying a threshold (and optionally also exceeding a threshold number of time windows in which such a threshold is satisfied) may be a condition of the determination. The determination may also include a condition that a signal in a frequency range (3 cpm, 2-4 cpm, 1-5 cpm) associated with stomach peristalsis is zero or otherwise below a threshold.Optionally, the determination may be based on the time series of motion sensor data itself (in the absence of an intermediate frequency -domain representation). For example, the motion sensor data may be processed to obtain a metric representing an amount of tumbling or rotation of the ingestible capsule, with a step change increase being associated with passing the gastro-duodenal junction.Optionally, there may be a delay or margin between determining that the ingestible capsule has passed the gastro-duodenal junction, and outputting the report.The outputting the report may be unconditional once the positive determination of the capsule having passed the gastro-duodenal junction is made. In an alternative, the processing may be repeated on one or more occasions following an initial determination, with a predefined number of consecutive positive determinations that the ingestible capsule has passed the gastro-duodenal junction required before the report is output. A delay timer initiated after a first positive determination may be reset by a subsequent negative determination, with output of the report being conditional upon the delay timer reaching a predefined value / time.Obtaining a Time Series of ReadingsFigure 10 illustrates a method including steps or actions S401, S402, S403 and S404.At S401 data representing a time series of readings from gas sensor apparatus housed within a ingestible capsule device 10 orally ingested by a subject is obtained, for example at a processor 151. The time series of readings are taken during exposure of the gas sensor apparatus to a gas mixture at the ingestible capsule device 10 during passage of the ingestible capsule device 10 through a gastrointestinal tract of the subject 40. Each reading has a value representing a signal output by gas sensing apparatus that is sensitive to CO2 concentration. The readings may be taken at predefined intervals, such as every second, every 5 seconds, every 10 seconds, every 15 seconds, every 20 seconds, every 30 seconds, every minute. The readings form a time series. The readings may each include an explicit indication of time such as a time stamp, or time may be implicit by virtue of position within a chronological sequence. For example, post-initiation, the nth reading is at a time of n x m seconds, wherein m is the period between successive readings.The gas sensor apparatus may be a single gas sensor such as thermal conductive device (TCD) gas sensor that is sensitive to CO2 concentration. The TCD gas sensor may be operated at a plurality of temperatures (i.e. driven with varying input power) to add an additional dimension to the readings, from which additional information the CO2 concentration is derivable (for example by comparison of TCD readings at different sensor temperatures). That is, CO2 concentration within the gas mixture is derivable from the variability of TCD at different TCD gas sensor operating temperature setpoints. The processor executing the method may be on-board the ingestible capsule device 10, or off-board, wherein off-board includes being either at a receiver apparatus 30 in direct communication with the ingestible capsule device 10, or at a remote apparatus 20 in data communication with the receiver apparatus 30.Optionally, the ingestible capsule device 10 further comprises processor hardware 151, memory hardware 152, and a wireless transmitter 18, and the processor hardware 151 in cooperation with the memory hardware 152 is configured to perform either the whole method of Figure 10 during passage of the ingestible capsule device 10 through the gastrointestinal tract of the subject 40, or to perform steps S401 to S403 during said passage, and further to completing step S403 or S404, to transmit data indicating one or more detected spikes in the CO2 concentration or a gastroparesis diagnosis or suspected gastroparesis diagnosis to a receiver device 30 via the wireless data transmitter 18. The processor hardware 151 and memory hardware 152 may be combined in a single chip.DETECTING SPIKES IN CO2 CONCENTRATION AND DISTINGUISHING CAUSESProcessing steps S402 and S403 are mutually interdependent and may be performed one after the other, in any order, or concurrently. Furthermore, since the processing may be performed on-the-fly, it may be that processing steps S402 and S403 (and also S404) are performed whilst S401 is still ongoing.At steps S402 and S403, the obtained data representing the time series of readings from the gas sensing apparatus is processed to identify spikes in CO2 concentration and to distinguish those spikes as being a gastric-duodenal transition indicator spike and gastroparesis indicator spike or spikes. The distinction may be made based on chronology, since gastroparesis indicator spikes are caused by CO2 concentration increase while the capsule 10 is resident in the stomach, whereas the gastric-duodenal indicator spike is caused by the capsule 10 passing out of the stomach and into the small intestine. Therefore, it can be appreciated that the two steps are somewhat logically interdependent, and though the gastroparesis indicator spikes may be detected first (if spike detection is performed on-the-fly rather than retrospectively), the determination that they are gastroparesis indicator spikes is dependent upon detection of a gastric-duodenal transition indicator spike in later CO2 concentration values. Alternatively the gastric-duodenal transition timing may be determined from, for example, H2 concentration readings, accelerometer readings, and / or reflectometer readings, so that based on that timing the CO2 concentration spike at gastric -duodenal transition is not mis-identified as a gastroparesis indicator spike, because it does not precede gastric-duodenal transition. Embodiments may incorporate a five-, ten-, or fifteen minute buffer into the determined gastric -duodenal transition timing, so that any CO2 concentration spike within the buffer preceding the determined gastric-duodenal transition timing is not detected as a gastroparesis indicator spike.The processing the readings may include deriving or otherwise extracting or determining CO2 concentration values from the gas sensing apparatus readings. For example, the gas sensing apparatus readings may be from a TCD gas sensor which is operated to take readings at different operating temperature setpoints by the processor hardware 151 or some other on-board controller or microcontroller. By comparing the TCD gas sensor readings at the different operating temperature setpoints, the CO2 concentration is derivable. It is noted that other techniques for measuring CO2 concentration exist such as electrochemical sensors, non-dispersive infrared sensor and metal oxide semiconductor sensors.At S402 the recorded sensor readings are processed to detect one or more spikes in the CO2 concentration in the gas mixture preceding a determined gastric -duodenal transition timing.For example, a spike may be detected by identifying: a first period of increasing CO2 concentration at a rate of change with respect to time exceeding an increase gradient threshold, the increase gradient threshold being either predefined or calculated based on the subset of the time series of readings,followed by a second period of decreasing CO2 concentration at a rate of change with respect to time exceeding a decrease gradient threshold, the decrease gradient threshold being either predefined or calculated based on the subset of the time series of readings, wherein if the first period and the second period are identified, and if a duration between the identified first and second periods is below a predefined threshold, the readings representing the first period and the second period are detected as a spike. As a further example, the first derivative of the CO2 concentration with respect to time may be monitored to detect an inflection point, wherein the inflection point is determined to be indicative of a spike if it is at a height more than a threshold above a calculated (for example by a rolling average) baseline or if a gradient defined by readings preceding the inflection point is above a threshold and likewise a gradient defined by readings proceeding the inflection point is above a threshold. As a further example, a pattern matching algorithm may be configured to detect spikes in the readings, wherein the pattern matching algorithm is pre-trained with training data comprising readings containing labelled spikes and training data without spikes. The pattern matching algorithm may be a neural network such as a convolutional neural network.The detecting one or spikes in CO2 concentration during gastric residence of the capsule 10 (i.e. preceding a spike associated with gastric -duodenal transition) may be performed by the on-board processor 151, i.e. on-the-fly, or may be determined by an off-board processor at a receiver device 30 or a processing apparatus receiving data therefrom, either in real-time or retrospectively.DETERMINING LOWER BOUND ON TIMING IN WHICH CO2 CONCENTRATION SPIKES ARE DETECTABLEAn ingestion event may be determined by a user recording a timing of ingestion on a user interface such as provided by an application or other software running on a receiver device 30. An ingestion event may be determined by, for example, receiving temperature readings from a temperature sensor on board the capsule 10, and determining when the temperature readings start to be within a range predefined for a subject stomach. Ingestion event may also be determined by a relative humidity sensor on board the capsule, by determining when the relative humidity readings start to be within a range predefined for a subject stomach. The determined ingestion event timing is not necessarily the start event for the gastric time period. For example, to filter out effects caused by intake and outtake of breath whilst the capsule 10 is in the oesophagus, a predefined delay (such as five minutes or more, ten minutes or more, twenty minutes or more, thirty minutes or more) is applied between determined ingestion event timing and start of the gastric time period. The ingestion event timing may be determined by the on-board processor 151, i.e. on-the-fly, or may be determined by an off-board processor at a receiver device 30 or a processing apparatus receiving data therefrom, either in real-time or retrospectively.DETERMINING UPPER BOUND ON TIMING IN WHICH C02 CONCENTRATION SPIKES ARE DETECTABLEDifferent techniques may be used to bound temporally the readings from the gas sensor apparatus that are processed to detect the gastroparesis indicator spikes. In a first, ICJ-based, technique, and predicated upon there being a VOC gas sensor in the gas sensor apparatus, or some other mechanism for detecting passage of the capsule 10 across the ileocecal junction, an ileocecal junction indicator is detected and used as an upper temporal bound. A second, gastric -duodenal-transition-based technique, is predicated upon there being an accelerometer 19 in the capsule, a reflectometer formed by a directional coupler in series with a transmission antenna, or some other means beyond the spike in CO2 concentration of indicating gastric-duodenal transition. In the second technique, the gastric-duodenal transition indicator is detected in one or more of the accelerometer readings, the reflectometer readings, and any other means of indicating gastric-duodenal transition, and based of the timings of the indicator or indicators being coincident with (one another and) a spike in CO2 concentration, the spike in CO2 concentration is determined to be caused by gastric-duodenal transition of the capsule 10. Thus, the said spike is taken by the processor as an upper temporal bound on the CO2 concentration values in which gastroparesis indicator spikes are detected or detectable. In either case, the lower temporal bound may be, for example, a capsule 10 ingestion event.In other words, gastroparesis indicator spikes are spikes in CO2 concentration in the gas mixture at the capsule while the capsule 10 is resident in the stomach. But it is necessary to discount a latest spike chronologically since it is known that a spike in CO2 concentration occurs in both healthy and gastroparetic patients at the gastric -duodenal transition. Some embodiments may determine a gastric-duodenal transition timing from data other than the CO2 concentration data, so that by knowledge of that timing and its use as an upper bound the spike caused by gastric-duodenal transition is distinguishable from the gastroparesis indicator spikes. So the gastric-duodenal transition timing is determined, either based on direct determination by the gastric-duodenal-transition-based technique, or by indirect determination by the ICJ-based technique (and reasoning that the latest CO2 concentration spike preceding ICJ is caused by gastric -duodenal transition).The purpose of the upper bound is to filter out or otherwise prevent spikes in CO2 concentration associated with presence of the capsule in the cecum and beyond being erroneously detected as gastroparesis indicators.A gastric -duodenal transition indicator may be detected in one or more of the following sensor outputs:If the capsule 10 includes a reflectometer formed of an antenna (which may be the antenna of the wireless data transmitter 18) in series with a directional coupler, a gastric-duodenaltransition indicator may be detectable in readings of the reflectometer, for example, as a baseline shift.If the capsule 10 includes an accelerometer such as a tri-axial accelerometer, a gastric-duodenal transition indicator may be detectable in readings of the accelerometer, for example, in a metric derived from the readings and representing agitation or angular movement of the capsule 10, based on an observation that capsule 10 is more agitated or exhibits a higher rate of angular movement post-gastric emptying.A gastric -duodenal indicator may be detectable in readings of the gas sensing apparatus, for example in the output signal of a TCD gas sensor, either in the raw output signal or in calibrated readings representing concentration of one or more constituent gases such as CO2, such an indicator is distinguishable from gastroparesis indicators by occurring chronologically later than the gastroparesis indicators.Processing to detect or identify a gastric-duodenal transition indicator includes monitoring or otherwise assessing recorded readings from the TCD gas sensor, the accelerometer, and / or the reflectometer, to detect a characteristic feature in the readings that may indicate gastric-duodenal transition of the capsule 10. Characteristic features may be spikes, baseline shifts, inflection points, depending on the sensor and the nature of the readings.Since the timing of gastric emptying (gastric -duodenal transition) may be determined by sensors such as the accelerometer 19 and the reflectometer 18, it is not strictly necessary to detect the spike in CO2 concentration at S403, as long as the timing of the spike can be determined. That is, if the timing of gastric -duodenal transition of the capsule 10 is determined, then step S402 may only detect spikes in readings preceding that timing, so that the CO2 concentration spike associated with the gastric emptying event is not determined.Determining whether an indicator is caused by a gastric -duodenal transition event of the capsule 10 may comprise calculating a confidence score for the hypothesis that the detected indicator was caused by the said gastric -duodenal transition event. Optionally, a threshold may be applied to the confidence score wherein exceeding the threshold is a positive determination. A confidence score below the threshold may be a negative determination or may be a trigger for further processing such as processing the readings of sensors other than that providing the detected indicator to identify one or more further indicators (for example, if the initial sensor is in the TCD gas sensor output then processing the accelerometer and / or reflectometer readings). A revised confidence score is then calculated based on the combination of the initial indicator and the one or more further indicators, which is compared with the threshold and a positive determination made in the event that the threshold is met.The gastric -duodenal transition event timing may be determined by the on-board processor 151, i.e. on-the-fly, or may be determined by an off-board processor at a receiver device 30 or a processing apparatus receiving data therefrom, either in real-time or retrospectively.DETECT OR DIAGNOSE GASTROPARESIS BASED ON SPIKESAt S404 a processor housed either on-board the capsule 10 or at a receiver device or processing apparatus in data communication therewith is configured to determine whether or not the spikes detected at S403 indicate that the subject is suffering from gastroparesis. in other words, at S404 a diagnosis of gastroparesis, suspected gastroparesis, or a negative diagnosis, is made, based on the detected gastroparesis indicator spike or spikes. For example, S404 may comprise calculating a confidence score in the diagnosis or suspected diagnosis, the confidence score being calculated with reference to the detected one or more spikes relative to one or more reference cases. A lookup table may be stored on a memory (such as on-board memory hardware 152) accessible to the processor executing S404 which lookup table may store a confidence score value for a number of detected spikes as a key. In a further example, the lookup table may be multi-dimensional and may store a confidence score value for a key of up to n components, wherein each component is a spike height or some other value representing a magnitude of each detected gastroparesis indicator spike in CO2 concentration.Alternatively, a formula may be stored enabling a confidence score to be calculated for input values including number of detected spikes and optionally also height per spike, or some other metric indicating magnitude of each detected gastroparesis indicator spike in CO2 concentration. As a further alternative, it may be that, in place of confidence score, a simple single Boolean value indicating a positive or negative value is calculated and output (for example in a report transmitted away from the capsule 10 by a wireless data transmitter 18). As a further alternative, a single value may indicate one of three outcomes: gastroparesis diagnosis, suspected gastroparesis diagnosis, negative gastroparesis diagnosis. Wherein suspected gastroparesis diagnosis may be a signal or alert to a clinician to undertake further testing or investigation. A confidence score is a quantification of the likelihood that the hypothesis “the detected gastroparesis indicator spikes are caused by gastroparesis in the subject” is true.The lookup table is exemplary of a model. The model may also be one or more functions, a machine learning model, or some other processing model. In either case, the model is configured, based on calculated values of one or more input factors or input parameters, to generate a corresponding output being a quantitative or qualitative indication of likelihood of gastroparesis being present in the subject. Furthermore, the output may include an indication of severity of gastroparesis, or likely severity of gastroparesis.Examples of input factors include:a count of the number of gastroparesis indicator spikes;a count of the number of gastroparesis indicator spikes and the height of each gastroparesis indicator spike;a histogram or another representation of distribution of heights of each gastroparesis indicator spike; an aggregate height of the gastroparesis indicator spikes;a count of the number of gastroparesis indicator spikes and the area under each gastroparesis indicator spike;a histogram or another representation of distribution of areas under each gastroparesis indicator spike; an aggregate area under the gastroparesis indicator spikes;temporal duration between ingestion of the ingestible capsule device and the gastric- duodenal transition timing of the ingestible capsule device.The model is predefined in a model configmation process, which may be, for example, a training stage of a machine learning model, or manual configmation of a model by an expert, or some other configuration of the model based on sample or training data. The training data being, for real life subjects ingesting a capsule to generate data from which values of one or more of the input parameters are calculated, the values of the one or more input parameters, along with a target output being a positive or negative diagnosis and / or an indication of severity. By training or otherwise configuring the model based on a number of subjects, being, for example, 20 or more, 30 or more, 40 or more, 50 or more, 100 or more, the model is taught or otherwise configured to predict a positive or negative diagnosis and / or an indication of severity based on values of one or more input factors. The model may be a machine learning model such as an artificial nemal network. In a further example, the model may be a threshold value applied to a single calculated input factor (for example, if aggregate area under the spikes exceeds a defined threshold, then positive diagnosis), or a plurality of thresholds for respective individual input factors which must all be satisfied or a predefined proportion must be satisfied for a positive diagnosis. Such a threshold or thresholds may be determined by a human expert analysing sample data and defining a threshold based on the sample data. In the case of the machine learning model or the one or more defined thresholds, an additional criterion may be a minimum temporal duration between ingestion timing and gastric-duodenal transition timing. For example, embodiments may be configmed with a minimum-time-to-gastric-emptying threshold of 4 or 5 hours (or anywhere between), wherein a requirement for a positive diagnosis is that the minimum-time-to-gastric -emptying threshold is met. Optionally, said minimum-time-to-gastric-emptying threshold not being met, but one or more other thresholds being met, may lead to a diagnostic result of suspected gastroparesis output, rather than a positive diagnosis per se.Alternatively, the temporal duration between ingestion and determined gastric -duodenal transition of the capsule is determined, and if it is within a predefined window (such as between 4 and 5 hours), then the above one or more thresholds applies to an input factor relating the gastroparesis indicator spikes is applied, and if satisfied, then the model output is suspected gastroparesis, or a positive diagnosis, and if not satisfied, then the model output is negative diagnosis. That is, the model may be an algorithm comprising one or more conditions, the one or more conditions being based on a threshold applied to a value of an input factor relating to the gastroparesis indicator spikes, and / or a minimum-time-to-gastric-emptying threshold. Examples of input factors include: a count of the number of gastroparesis indicator spikes; a count of the number of gastroparesis indicator spikes and the height of each gastroparesis indicator spike; a histogram or another representation of distribution of heights of each gastroparesis indicator spike; an aggregate height of the gastroparesis indicator spikes; a count of the number of gastroparesis indicator spikes and the area under each gastroparesis indicator spike; a histogram or another representation of distribution of areas under each gastroparesis indicator spike; an aggregate area under the gastroparesis indicator spikes; temporal duration between ingestion of the ingestible capsule device and the gastric- duodenal transition timing of the ingestible capsule device.At S404, a determination is made as to whether the CO2 concentration data contains evidence of gastroparesis, and whether that evidence is sufficient to positively diagnose gastroparesis in the subject. In addition, a qualitative or quantitative indication of severity of gastroparesis in the patient may be determined.The ingestible capsule device ingested by the patient may be an ingestible capsule device comprising: an ingestible indigestible bio-compatible housing; and, within the housing: a power source; sensor hardware including gas sensing apparatus; processor hardware; memory hardware; and a wireless data transmitter; the memory hardware storing processing instructions which, when executed by the processor hardware, cause the processor hardware to perform a process comprising: obtaining data representing a time series of readings from the gas sensing apparatus, the time series of readings being taken during exposure of the gas sensing apparatus to a gas mixture at the ingestible capsule device during passage of the ingestible capsule device through a gastrointestinal tract of a subject, the subject having orally ingested the ingestible capsule device, each reading having a value, the values of the readings being sensitive to CO2 concentration in the gas mixture; processing the readings to detect one or more gastroparesis indicator spikes in the CO2 concentration with respect to time, a gastroparesis indicator spike being a spike in the CO2 concentration with respect to time at a timing after an ingestion timing of the ingestible capsule device and preceding a gastric -duodenal transition timing of the ingestible capsule device; based on the detected one or more gastroparesis indicator spikes in the CO2 concentration with respect to time, diagnosing gastroparesis or suspected gastroparesis.Optionally, the gas sensing apparatus includes a thermal conductivity detector, TCD, gas sensor, wherein processing the readings from the subset to detect one or more spikes in the CO2 concentration values with respect to time includes referencing calibration data transforming TCD gas sensor reading values to CO2 concentration values.Optionally, the process further comprises: controlling the TCD gas sensor to multiple operating temperature set points at which to make readings, wherein processing the readings from the subset to detect one or more gastroparesis indicator spikes in the CO2 concentration includes comparing the TCD gas sensor readings at different operating temperature set points with one another to calculate CO2 concentration values.Optionally, diagnosing gastroparesis or suspected gastroparesis based on the detected one or more gastroparesis indicator spikes in the CO2 concentration with respect to time comprises: predicting presence or absence of gastroparesis in the subject by calculating values of each of one or more factors and inputting the calculated values to a predefined model outputting a quantitative or qualitative indication of likelihood of gastroparesis being present in the subject based on the one or more input calculated values; the calculated values of each of one or more factors comprising one or more from among: a count of the number of gastroparesis indicator spikes; a count of the number of gastroparesis indicator spikes and the height of each gastroparesis indicator spike; a histogram or another representation of distribution of heights of each gastroparesis indicator spike; an aggregate height of the gastroparesis indicator spikes; a count of the number of gastroparesis indicator spikes and the area under each gastroparesis indicator spike; a histogram or another representation of distribution of areas under each gastroparesis indicator spike; an aggregate area under the gastroparesis indicator spikes; temporal duration between ingestion of the ingestible capsule device and the gastric- duodenal transition timing of the ingestible capsule device.Optionally, the output of the predefined model includes an indication of severity of gastroparesis in the subject.Optionally, the ingestible capsule device also houses an environmental temperature sensor to detect an environmental temperature at the ingestible capsule device, wherein the process further comprises preprocessing the readings from the gas sensing apparatus to compensate for changes in the environmental temperature.Optionally, processing the readings to detect one or more gastroparesis indicator spikes, comprises: detecting a gastric -duodenal transition indicator in the time series of readings from the gas sensing apparatus, determining that the gastric -duodenal transition indicator in the time series of readings fromthe gas sensing apparatus is caused by a gastric -duodenal transition by the ingestible capsule device, and determining a timing of the gastric -duodenal transition indicator as the gastric-duodenal transition timing.Optionally, processing the readings to detect one or more gastroparesis indicator spikes comprises: determining an upper bound on the gastric-duodenal transition timing by positively detecting residence of the ingestible capsule device in the intestines of the gastrointestinal tract; detecting a chronologically latest spike in the CO2 concentration with respect to time preceding the upper bound as a gastric-duodenal transition indicator spike and determining that the gastric-duodenal transition indicator spike is caused by a gastric -duodenal transition by the ingestible capsule device, and determining a timing of the gastric -duodenal transition indicator as the gastric -duodenal transition timing.Optionally, positively detecting residence of the ingestible capsule device in the intestines of the gastrointestinal tract includes detecting an ileocecal junction transition indicator, determining that the ileocecal junction indicator is caused by an ileocecal junction transition of the ingestible capsule device, and determining a timing of the ileocecal junction transition indicator as the upper bound on the gastric-duodenal indicator timing.Optionally, the gas sensing apparatus includes a VOC gas sensor and the ileocecal junction transition indicator is a feature in a time series of readings from the VOC gas sensor, the reading being a turning point, a step change, or a period of gradient increase exceeding a gradient increase threshold.Optionally, the process further comprises: obtaining data representing a time series of readings from a reflectometer housed within the ingestible capsule device and formed of an antenna in series with a directional coupler; determining the gastric-duodenal transition timing by: detecting a gastric-duodenal transition indicator in the time series of readings from the reflectometer, determining that the gastric-duodenal transition indicator in the time series of readings from the reflectometer is caused by a gastric-duodenal transition by the ingestible capsule device, and determining the gastric -duodenal transition timing based on the timing of the detected gastric -duodenal transition indicator.Optionally, the ingestible capsule device further comprises an accelerometer, and the process further comprises: obtaining data representing a time series of readings from the accelerometer; determining the gastric -duodenal transition timing by: detecting a gastric -duodenal transition indicator in the time series of readings from the accelerometer, determining that the gastric-duodenal transition indicator is caused by a gastric -duodenal transition by the ingestible capsule device, and determining the gastric-duodenal transition timing based on the timing of the detected gastric -duodenal transition indicator.Optionally, a spike in the CO2 concentration with respect to time, being a gastroparesis indicator spike or a gastric -duodenal transition indicator spike, is detected by identifying: a first period of increasing CO2 concentration at a rate of change with respect to time exceeding an increase gradient threshold, the increase gradient threshold being either predefined or calculated based on the subset of the time series of readings, followed by a second period of decreasing CO2 concentration at a rate of change with respect to time exceeding a decrease gradient threshold, the decrease gradient threshold being either predefined or calculated based on the subset of the time series of readings, wherein if the first period and the second period are identified, and if a duration between the identified first and second periods is below a predefined threshold, the readings representing the first period and the second period are detected as a spike.Optionally, a spike height threshold is applied to the first period of increasing CO2 concentration wherein a magnitude of increase in CO2 concentration represented by the first period is compared with the spike height threshold, and if the magnitude of increase does not meet the spike height threshold then the readings representing the first period and the second period are not detected as a spike.Optionally, a spike in the CO2 concentration with respect to time, being a gastroparesis indicator spike or a gastric-duodenal transition indicator spike, is detected by identifying a local maximum feature being a singularity, discontinuity, or inflection point at more than a predefined threshold above a baseline value defined based on values preceding the feature.Optionally, processing the readings to detect one or more gastroparesis indicator spikes, includes: determining that the ingestible capsule device has been ingested by the subject and the ingestion timing.Optionally, the ingestible capsule device houses an environmental temperature sensor to detect an environmental temperature at the ingestible capsule device, and determining that the ingestible capsule device has been ingested by the subject and the ingestion timing is by comparison of the environmental temperature represented by a signal from the environmental temperature sensor with a predefined temperature range for stomach of the subject or for the gastrointestinal tract of the subject.Optionally, the ingestible capsule device houses a relative humidity sensor to detect relative humidity at the ingestible capsule device, and determining that the ingestible capsule device has been ingested by the subject and the ingestion timing is by comparison of the relative humidity represented by a signal from the environmental temperature sensor with a predefined relative humidity range for stomach of the subject or for the gastrointestinal tract of the subject; wherein determining that the ingestible capsule device has been ingested and the ingestion timing is based on one or both of the relative humidity and the environmental temperature being within the respective predefined range.Optionally, diagnosing gastroparesis or suspected gastroparesis includes calculating a score representing likelihood of gastroparesis being present in the subject, the likelihood score being calculated with reference to the detected one or more spikes relative to one or more reference cases.Optionally, calculating the likelihood score is performed by a machine learning algorithm pre-trained with labelled training data, training data being representations of CO2 concentration measured by ingestible capsule devices during residence in stomachs of respective training subjects, each training subject being clinically diagnosed by a medical practitioner as being gastroparesis positive or gastroparesis negative, and the training data being labelled with the clinical diagnosis of the respective subject.CASE STUDY: GLP-1RA patient with apparent persistent lower GI symptomsLower GI symptoms such as constipation are commonly managed with a medication working directly on the colon.Colon-targeted drugs to manage constipation symptoms give the patient massive diarrhoea, even at the lowest dose. After five days without GI medication, the clinician gave the patent an ingestible capsule to generate a motility report measuring gastric emptying time (post-ingestion), small intestine residence time, and large intestine residence time. When the patient had not had a bowel movement after five days, Motegrity (a 5-HT4 receptor agonist and whole gut prokinetic) was prescribed, which also caused diarrhoea. After passing the ingestible capsule, the report data showed delayed gastric emptying and slow small-bowel transit with normal colonic motility, confirming the problem was in the upper GI tract. Based on the report data, the treatment plan was adjusted as follows by the clinician:-lowered the GLP-1 RA (semaglutide) dose from 2.4mg (max dose) to Img-dietary changes: reduced daily protein intake from 100g to 70g-recommended the OTC supplement iberogast-planned follow up in one month.Insight reported by clinician:The regional motility data provided by Atmo was pivotal in distinguishing the upper-GI delay from colonic dysfunction to allow more targeted management. The clinician also reported that it is common for similar situations to occur in which a patient is treated with colon-treating therapeutics even where the colon is not actually suffering motility issues. In particular this occurs in women, which may be because women at baseline tend to have higher incidence of gastroparesis than men, women have massive hormonal changes while going through menopause, addition of a GLP- 1 RA can heighten these effects.FURTHER USE CASESGatroenterologists and hepatology practitioners often prescribe GLP-1 RAs for fatty liver disease and abnormal function. Regular motility reporting will enable the practitioners to assess to what extent the GLP-1 RAs are slowing GI motility and thus to determine whether and how to titrate the prescription. The same is true for patients prescribed GLP-1 RAs for cardiovascular disease, chronic kidney disease, and obstructive sleep apnoea. Clinical practitioners would benefit from the ability to measure how the GLP-1 RAs are influencing gut motility and to use the measurement to determine whether and how to titrate the dosage. In particular, the administering to the patient of an digestible capsule to report on GI tract motility may be performed shortly after escalation of GLP-1 RA dosage in a treatment regime, wherein if the results show absence of gastroparesis, the practitioner determines to continue with the escalated dosage, and if the results show presence of gastroparesis, the practitioner reverts to the dosage regime pre-escalation. As an alternative to regularised GI tract motility reporting, ingestible capsules for motility reporting may be administered to patients on GLP-1 RA regimes in response to sudden severe side effects such as nausea, cramping, diarrhoea, vomiting, to investigate any co-occurring gut motility issues.INGESTIBLE CAPSULES FOR MOTILITY REPORTINGWhere techniques are described which use an ingestible capsule to measure patient GI tract motility, the ingestible capsule may be, for example: an ingestible capsule housing a motion sensor configured to generate a time series of motion sensor data representing motion of the ingestible capsule during passage through the GI tract of a subject; the ingestible capsule further housing, or being configured to transmit the time series of motion sensor data to: data processing hardware communicably coupled to the motion sensor, the data processing hardware being configured, following ingestion of the ingestible capsule by the subject, at data processing hardware communicably coupled to the motion sensor, to perform a process comprising: generating a spectral analysis of the time series of motion sensor data generated by the motion sensor over a time period during passage of the ingestible capsule through the GI tract, using the spectral analysis to detect peristalsis at a location of the ingestible capsule within the GI tract at the time period; determining the location of the capsule within the GI tract based on the detected peristalsis. The process is performed repeatedly in order to monitor GI tract location over time and to generate a GI tract motility report. The process may further comprise: generating and outputting to a receiver computing apparatus or message recipient a report including one or more from among: information extracted from the spectral analysis representing fluctuations, variations, or anomalies within spectral components; a metric or metrics calculated from the spectral analysis from among: centre frequency, frequency spread, power distribution, frequency gaps. Detecting peristalsis at a location of the ingestible capsule within the GI tract at the time period may comprise, in the result of the spectral analysis, detecting presence or absence of a signal at one or more of a predefined set of peristalsis indicator frequency ranges. The process may include determining the location of the capsule within the GI tract based on the detected peristalsis; and determining the location of the ingestiblecapsule within the GI tract comprises, in the result of the spectral analysis, detecting presence or absence of a signal / component at one or more of a predefined set of peristalsis indicator frequency ranges. A predefined set of peristalsis indicator frequency ranges may comprise one or more from among: a first peristalsis indicator frequency range, being a frequency range indicating stomach peristalsis; a second peristalsis indicator frequency range, being a frequency range indicating small intestine peristalsis; a third peristalsis indicator frequency range, being a frequency range indicating large intestine peristalsis. Optionally, the ingestible capsule is determined to be in the stomach by presence of a signal in the first peristalsis indicator frequency range; the ingestible capsule is determined to be in the small intestine by presence of a signal in the second peristalsis indicator frequency range; the ingestible capsule is determined to be in the large intestine by absence of a signal in the first peristalsis indicator frequency range and absence of a signal in the second peristalsis indicator frequency range; and / or the ingestible capsule is determined to be in the large intestine by presence of a signal in the third peristalsis indicator frequency range.The ingestible capsule may be part of an apparatus comprising: an ingestible capsule housing sensor hardware configured to generate readings representing motion of the ingestible capsule during passage through the GI tract of a subject, the sensor hardware including a motion sensor; and data processing hardware communicably coupled to the sensor hardware, the data processing hardware being configured, following ingestion of the ingestible capsule by the subject, to perform a method comprising obtaining readings from sensor hardware on-board a subject ingestible capsule, the readings being obtained during a passage of the ingestible capsule through the GI tract of a subject, the sensor hardware comprising at least a motion sensor; inputting a representation of the readings to a machine learning algorithm comprising a model trained to generate a classification of a cause of the obtained readings; processing the readings by the machine learning algorithm to generate a prediction of a cause of the obtained readings at one or more instances during the passage based on one or a time series of classifications generated by the model; and outputting the prediction. Wherein, the classification of the cause of the obtained readings is a classification of GI tract location of the subject ingestible capsule at the timing of the obtained readings, and the prediction of the cause of the obtained readings is a prediction of GI tract location of the capsule at one or more instances during the passage based on one or a time series of the classifications.Figure 11Figure 11 illustrates a method, in which, contrary to methods detailed above, the motion sensor data may be processed by an ML algorithm in the absence of the intermediate step of generating a spectrograph.Figure 11 illustrates a method for predicting location of an ingestible capsule within a GI tract of a subject by processing readings from an on-board motion sensor. Capsule ingestion operation S1101 is equivalent to S101 discussed above. Likewise obtaining SI 102 is equivalent to S102 above.The method illustrated by Figure 11 includes steps S 1102 to S 1105 which define a process for obtaining and processing readings obtained from an on-board motion sensor. The process may be performed while the capsule is still within the Gl-tract, or may be performed retrospectively. For example, the capsule may record readings during passage through the GI tract, which readings are transmitted to an external computing apparatus (at which the process is performed) post-excretion. Alternatively, the capsule may may be configured to perform the process at on-board computing hardware in real time (wherein real time is taken to mean in real time save for latency associated with recording readings, grouping readings into frames or samples, and executing the processing. Alternatively, the capsule may transmit (via a primary wireless transceiver) the readings to an external computing apparatus during passage through the GI tract, so that the external computing apparatus may perform the process in real time (save for latency as outlined above and also including data transmission latency).Even in the case of on-board processing, a report comprising one or more predictions or summarising predictions generated by the process may be transmitted to an external computing apparatus. The report may be transmitted post-excretion and include an indication that excretion has been detected.Step SI 102 is obtaining readings from an on-board motion sensor. The readings may be obtained directly from the sensor, or via an intermediate component such as a microcontroller and optionally also via a data transmission / reception protocol. The readings may be raw measurements or may be pre-processed (prior to the obtaining SI 102 or as part of the obtaining SI 102) for transmission and / or downstream processing, or otherwise sampled, summarised, or represented. The readings may be alphanumeric values or value vectors, or may be image data representing a time series of values during a defined time window. The motion sensor may be a magnetometer. The motion sensor may be an inertial motion sensor. The motion sensor may be an accelerometer.At S 1103 the obtained readings are input to a trained machine learning algorithm. The machine learning algorithm includes at least a model such as an artificial neural network trained to accept an input vector comprising the readings from the motion sensor (and optionally also readings from additional sensors) and to classify the readings in a manner indicative of GI tract location. The classification is taken to mean the output of the artificial neural network itself, and the prediction at SI 104 is either the classification generated at SI 103 or the result of further processing of the classification at SI 104. For example further processing may be adding the classification to a time series of such classifications obtained during the passage through the GI tract of a subject, wherein a prediction of capsule location at a point in time T is derivable based on the time series of classifications overlapping T, preceding T, and / or proceeding T.The classification generated at S 1103 and the prediction generated as S 1104 may be of location of the capsule within the GI tract.The machine learning algorithm may be trained to classify a portion of the GI tract in which the capsule is resident at a time window based on inertial motion sensor readings obtained during the time window. In this case, a prediction may be based on a combination of classifications of a number of adjacent time windows, wherein a predefined threshold number or proportion of classifications in agreement corresponds to a prediction. Alternatively, the machine learning algorithm may be trained to classify transitions between portions of the GI tract occurring during a time window based on motion sensor readings obtained during the time window. In this case, a prediction may be generated by applying logical rules to the classification to obtain a prediction. For example, if gastric -duodenal junction transition has been detected, and ileocecal junction transition has not been detected, then the prediction is that the capsule is in the small intestine.At SI 105 the prediction is output. The nature of the output is implementation-dependent. The output may be a trigger signal to a microcontroller to initiate the obtaining of a sample of matter from the GI tract, or the release of therapeutic matter into the GI tract from a compartment on the capsule having a triggerable release mechanism. The prediction may be reported to a receiver apparatus either during passage through the GI tract, or at excretion. The prediction may be reported on its own or may be added to a time series of predictions pertaining to different time periods or to different portions of the GI tract. A prediction may be reported when it is different from a preceding prediction in a time series. At SI 106 GLP-1 dosage titration is determined, using as inputs the report from SI 105 and / or the readings or predicted cause of SI 104. Titrating may be interpreted as continuously measuring and adjusting (the dosage regime). For example, a dosage regime for a GLP-1 receptor agonist may comprise a weekly dose within a defined range, such as 0.75mg to 4.5mg, or between 0.25mg and 5mg, or between 0.25mg and 4.5mg. The clinician or algorithm may respond to gastric emptying timings (i.e. determined time between the ingestible capsule being ingested and passing the gastric -duodenal junction) within defined ranges by determining to increase the weekly dosage, hold or maintain the weekly dosage, or decrease the weekly dosage. It may be that increase / decrease is only ever by a fixed increment, such as 0.5mg (without exiting the defined range), or it may be that there are ranges of gastric emptying timings that determine the increment or decrement should be greater. Similarly, determining presence or absence of gastroparesis may determine whether to maintain, increase, or decrease, the weekly dosage (of the GLP-1 receptor agonist or other incretin mimetic). Presence of gastroparesis may be a determinant factor in decreasing weekly dosage. Absence of gastroparesis may be a criterion in determining that weekly dosage is to be increased, if it is below a desired level for the patient, but not otherwise.Beyond gastroparesis, other motility indicators such as small bowel residence time and large bowel residence time may be monitored by a clinician based on the report or other data output by the capsule. Whilst some slowing of GI tract motility may be expected, there are conditions for which GLP-1 RAs are prescribed and for which slowing of GI tract motility would be considered an undesirable side effect. Therefore, if one or a series of consecutive GI tract motility report-producing ingestible capsules indicate small bowel residence time is above a threshold (which would be determined according to the individual patient situation and variables including age), and the same for gastric residence time or large bowel residence time, or for whole gut transit time, then a decision made by the clinician to reduce a weekly dosage of GLP-1 RAs by, for example, 0.25 mg or 0.5mg.Machine Learning Algorithm: Training DataThe machine learning algorithm, and specifically the model comprising the artificial neural network, is trained using labelled training data. In summary, the artificial neural network is configurable during training, after which its configuration is fixed for replication in the ingestible capsules 10 themselves which store executable versions of the model. In the off-board processing example, the trained machine learning algorithm may be distributed to (or otherwise stored for execution by) computing apparatus in data communication with an ingestible capsule or capsules 10 directly or via one or more intermediate devices.The present example discusses training in the context of triaxial accelerometer measurements. It may be that the trained machine learning algorithm is equivalent to the capsule motion sensor hardware configuration. Equivalent in this context is taken to mean configured to generate readings representing the same measurable physical property or properties. So that, for example, capsules 10 having a magnetometer as a motion sensor store an executable machine learning algorithm pre-trained on training data of labelled magnetometer readings, and so on.The model is configurable (trainable). The trainable model may be an artificial neural network. The artificial neural network may be a convolutional neural network. The model is configured to classify readings as being from within different portions of the GI tract, and / or as representing that a transition between portions has or has not occurred. The model may therefore be described as a classifier.In the present example, the data which the classifier is trained comprises triaxial accelerometer readings from capsules going through the gastrointestinal tract of respective subjects over time. The training data contain samples from subjects selected to sample the target users. The precise selection is dependent upon the implementation case. For example, if the implementation is to assist in detection or assessment of gastroparesis, then the training data is configured to include samples from subjects with positive and negative gastroparesis diagnoses, and within those subsets, with varying degrees of gastroparesis or GItract motility. In general, the model may be trained to be generic for use in various implementations including GI tract motility monitoring, therapeutic matter delivery, and diagnosis of conditions such as gastroparesis, constipation, SIBO. The training data is configured to represent such generic usage, by including samples demonstrating varying degrees of GI tract motility and both with and without the conditions. Furthermore, the training data may be configured to include samples who slept during the GI tract passage of the capsule, samples who undertook physical activities of varying degrees of duration and intensity, and possibly covering a range of samples across other characteristics influencing motion sensor data (such as time spent on a train or motor vehicle, subjects with sedentary and active lifestyles, etc). The model should learn to ignore and be robust to motion that is not characteristic of GI tract location of the capsule. The intention in the configuration (i.e. the selection of subjects and samples) of training data is for the model to learn what features are generalised across the population (of target users).In place of an accelerometer the classifier may be trained by labelled readings from other motionsensitive measurement devices (gyroscope, magnetometer, single or dual axis accelerometers) during respective GI tract passages.Labelled is taken to mean a human expert has manually classified the reading in accordance with the set of classifications for which the classifier is being trained, and the training data labelled with the manual classification.The training data may be obtained from ingestible capsules 10 having the same hardware configuration as the capsule obtaining SI 102 the readings on which the trained classifier will be executed.DESCRIPTION ENDS.

Claims

CLAIMS1. A method comprising, for a patient undertaking an incretin mimetic dosage regime, the patient having ingested or been administered an ingestible capsule configured to determine and report the location of the ingestible capsule within the GI tract, or to cooperate with a computing device to determine and report the location of the ingestible capsule within the GI tract:determining a time series of locations of the capsule within the GI tract;determining a gastro-intestinal motility indicator based on the time series of locations of the ingestible capsule within the GI tract; andtitrating or otherwise modifying the incretin mimetic dosage regime in dependence upon the gastro-intestinal motility indicator.

2. The method according to claim 1, wherein the incretin mimetic dosage regime is a dosage regime of one or more incretin mimetic compounds or medicaments from among: GLP-1 receptor agonists, dipeptidyl peptidase-4 (DPP-4)_ inhibitors, long-acting GLP-1 analogs, GLP-1 derivatives, peptide-based antidiabetic agents, peptidic GLP-1 receptor modulators, exogenous GLP-1 compounds, synthetic GLP-1 analogs, modified incretin peptides.

3. The method according to claim 1 or 2, wherein the gastro-intestinal motility indicator comprises gastric emptying time.

4. The method according to claim 1, wherein the gastro-intestinal motility indicator is a small-intestinal motility indicator, from which gastric emptying is inferable.

5. The method according to any of claims 1 to 4, wherein the gastro-intestinal motility indicator comprises, or further comprises, one or more from among:small bowel transit / residence time;whole gut transit time;large intestine residence time;gastric residence timeingestion time;excretion time;gastro-duodenal junction transit time;ileocecal j unction transit time ;combined small and large intestine residence time;combined stomach and small intestine residence time;orocecal residence time.

6. The method according to any of claims 1 to 5, wherein the patient is preoperative.

7. The method according to claim 6, wherein the patient is scheduled to undergo general anaesthetic.

8. The method according to any of claims 1 to 7, wherein the patient has a positive diagnosis of a condition for which dysmotility is a symptom.

9. The method of any of claims 1 to 8, wherein the patient has a positive diagnosis of one or from among:type-2 diabetes;obesity;hypertension;MASH, metabolic dysfunction-associated steatohepatitis;fatty liver disease;cardiovascular disease;chronic kidney disease;obstructive sleep apnoea;metabolic syndrome.

10. The method according to any of claims 1 to 9, wherein the ingestion or administering for ingestion is repeated on a plurality of occasions and the titrating or other modification of the incretin mimetic dosage regime is in dependence upon changes in the gastro-intestinal motility indicator between the plurality of occasions.

11. The method according to any of claims 1 to 10, wherein the ingestible capsule is configured to determine and report, to a computing device external to the patient, the gastro-intestinal motility indicator.

12. The method according to any of claims 1 to 11, wherein the ingestible capsule is configured to obtain sensor readings from within the GI tract of the patient, to transmit the sensor readings to a computing device external to the patient, and the computing device is configured to determine the gastro-intestinal motility indicator.

13. The method according to any of claims 1 to 12, wherein the patient is provided with access to a software application configured to receive inputs indicating occurrence and timing of one or more from among:- symptoms;- drug intake events;- mood;- constipation;- cramps;- stomach cramps;- pains;- headache;- sleep;- stress;- bowel movement events; and- food / drink ingestion events.

14. The method according to claim 13, wherein the symptoms include one or more from among nausea, vomiting, diarrhoea, cramps, constipation.

15. The method according to claim 13 or 14, wherein the titration or other modification of the incretin mimetic dosage regime is also in dependence upon inputs received via the software application; and / orwherein the titration or other modification of the incretin mimetic, such as a GLP-1 agonist, dosage regime is also in dependence upon inputs received via a wearable device or smartwatch and representing one or more from among:- symptoms;- drug intake events;- mood;- constipation;- cramps;- stomach cramps;- pains;- headache;- sleep;- stress;- bowel movement events; and- food / drink ingestion events.

16. The method according to any of claims 1 to 15, wherein the titrating or other modification of the incretin mimetic dosage regime is determined by a pre-trained machine learning algorithmconfigured to accept as inputs one or a series of gastro-intestinal motility indicators from the patient and to output one or a plurality of incretin mimetic dosage regimes to achieve dosage tolerability.

17. The method according to claims 13 to 16, wherein inputs to the pre-trained machine learning algorithm include inputs received via the software application.

18. The method according to any of claims 1 to 17, wherein the patient is administered for ingestion, or otherwise ingests, apparatus comprising:an ingestible capsule housing a motion sensor configured to generate a time series of motion sensor data representing motion of the ingestible capsule during passage through the GI tract of a subject;the ingestible capsule further housing, or being configured to transmit the time series of motion sensor data to:data processing hardware communicably coupled to the motion sensor, the data processing hardware being configured, following ingestion of the ingestible capsule by the subject, at data processing hardware communicably coupled to the motion sensor, to perform a process comprising:generating a spectral analysis of the time series of motion sensor data generated by the motion sensor over a time period during passage of the ingestible capsule through the GI tract, using the spectral analysis to detect peristalsis at a location of the ingestible capsule within the GI tract at the time period;determining the location of the capsule within the GI tract based on the detected peristalsis.

19. The method according to any of claims 1 to 18, further comprising administering the ingestible capsule to the patient for ingestion.

20. A method comprising, for a patient undertaking an incretin mimetic dosage regime, the patient having ingested or been administered an ingestible capsule configured to diagnose gastroparesis or suspected gastroparesis in a subject:determining a gastro-intestinal motility indicator based on a positive or negative diagnosis of gastroparesis or suspected gastroparesis in the subject as reported by the ingestible capsule; and titrating or otherwise modifying the incretin mimetic dosage regime in dependence upon the gastro-intestinal motility indicator.

21. The method according to claim 20, wherein the incretin mimetic dosage regime is a dosage regime of one or more incretin mimetic compounds or medicaments from among: GLP-1 receptor agonists, dipeptidyl peptidase-4 (DPP-4)_ inhibitors, long-acting GLP-1 analogs, GLP-1 derivatives,peptide-based antidiabetic agents, peptidic GLP-1 receptor modulators, exogenous GLP-1 compounds, synthetic GLP-1 analogs, modified incretin peptides.

22. The method according to claim 20 or 21, wherein the gastro-intestinal motility indicator comprises gastric emptying time.

23. The method according to any of claims 20 to 22, wherein the gastro-intestinal motility indicator is a small-intestinal motility indicator, from which gastric emptying is inferable.

24. The method according to any of claims 20 to 23, wherein the gastro-intestinal motility indicator further comprises, one or more from among:small bowel transit / residence time;whole gut transit time;large intestine residence time;gastric residence timeingestion time;excretion time;gastro-duodenal junction transit time;ileocecal j unction transit time ;combined small and large intestine residence time;combined stomach and small intestine residence time;orocecal residence time.

25. The method according to any of claims 20 to 24, wherein the patient is preoperative.

26. The method according to claim 25, wherein the patient is scheduled to undergo general anaesthetic.

27. The method according to any of claims 20 to 26, wherein the patient has a positive diagnosis of a condition for which dysmotility is a symptom.

28. The method of any of claims 20 to 27, wherein the patient has a positive diagnosis of one or from among:type-2 diabetes;obesity;hypertension;MASH, metabolic dysfunction-associated steatohepatitis;fatty liver disease;metabolic syndrome.

29. The method according to any of claims 20 to 28, wherein the ingestion or administering for ingestion is repeated on a plurality of occasions and the titrating or other modification of the incretin mimetic dosage regime is in dependence upon whether the diagnosis of gastroparesis changes between positive and negative between occasions.

30. The method according to any of claims 20 to 29, wherein the ingestible capsule is configured to determine and report, to a computing device external to the patient, the gastro-intestinal motility indicator.

31. The method according to any of claims 20 to 30, wherein the ingestible capsule is configured to obtain sensor readings from within the GI tract of the patient, to transmit the sensor readings to a computing device external to the patient, and the computing device is configured to determine the gastro-intestinal motility indicator.

32. The method according to any of claims 20 to 31, wherein the patient is provided with access to a software application configured to receive inputs indicating occurrence and timing of one or more from among:- symptoms;- drug intake events;- mood;- constipation;- cramps;- stomach cramps;- pains;- headache;- sleep;- stress;- bowel movement events; and- food / drink ingestion events.

33. The method according to claim 32, wherein the symptoms include one or more from among nausea, vomiting, diarrhoea, cramps, constipation.

34. The method according to claim 32 or 33, wherein the titration or other modification of the incretin mimetic dosage regime is also in dependence upon inputs received via the software application; and / orwherein the titration or other modification of the incretin mimetic dosage regime is also in dependence upon inputs received via a wearable device or smartwatch and representing one or more from among: - symptoms;- drug intake events;- mood;- constipation;- cramps;- stomach cramps;- pains;- headache;- sleep;- stress;- bowel movement events; and- food / drink ingestion events.

35. The method according to any of claims 20 to 34, wherein the titrating or other modification of the incretin mimetic dosage regime is determined by a pre-trained machine learning algorithm configured to accept as inputs one or a series of gastro-intestinal motility indicators from the patient and to output one or a plurality of incretin mimetic dosage regimes to achieve dosage tolerability.

36. The method according to claims 32 to 35, wherein inputs to the pre-trained machine learning algorithm include inputs received via the software application.

37. The method according to any of claims 30 to 36, further comprising administering the ingestible capsule to the patient for ingestion.

38. The method according to any of claims 20 to 37, wherein the ingestible capsule device comprises:an ingestible indigestible bio-compatible housing;and, within the housing:a power source;sensor hardware including gas sensing apparatus;processor hardware;memory hardware; anda wireless data transmitter;the memory hardware storing processing instructions which, when executed by the processor hardware, cause the processor hardware to perform a process comprising:obtaining data representing a time series of readings from the gas sensing apparatus, the time series of readings being taken during exposure of the gas sensing apparatus to a gas mixture at the ingestible capsule device during passage of the digestible capsule device through a gastrointestinal tract of a subject, the subject having orally ingested the ingestible capsule device, each reading having a value, the values of the readings being sensitive to CO2 concentration in the gas mixture;processing the readings to detect one or more gastroparesis indicator spikes in the CO2 concentration with respect to time, a gastroparesis indicator spike being a spike in the CO2 concentration with respect to time at a timing after an ingestion timing of the ingestible capsule device and preceding a gastric-duodenal transition timing of the ingestible capsule device; based on the detected one or more gastroparesis indicator spikes in the CO2 concentration with respect to time, diagnosing gastroparesis or suspected gastroparesis.

39. A method comprising, for a patient undertaking a medicament dosage regime, the patient having ingested or been administered an ingestible capsule configured to determine and report the location of the ingestible capsule within the GI tract, or configured to cooperate with a computing device to determine and report the location of the ingestible capsule within the GI tract:determining a time series of locations of the capsule within the GI tract;determining a gastro-intestinal motility indicator based on the time series of locations of the ingestible capsule within the GI tract; andtitrating or otherwise modifying the medicament dosage regime in dependence upon the gastrointestinal motility indicator.

40. The method according to claim 39, wherein the medicament is an incretin mimetic, wherein the medicament is a GLP-1 receptor agonist, or wherein the medicament is a substance affecting gastrointestinal transit times as a mechanism of action or as a side effect or reported side effect or potential side effect.

41. A method comprising, for a patient undertaking a medicament dosage regime, the patient having ingested or been administered an ingestible capsule configured to diagnose gastroparesis or suspected gastroparesis in a subject;determining a gastro-intestinal motility indicator based on a positive or negative diagnosis of gastroparesis or suspected gastroparesis in the subject; andtitrating or otherwise modifying the medicament dosage regime in dependence upon the gastrointestinal motility indicator.

42. The method according to claim 41, wherein the medicament is an incretin mimetic, wherein the medicament is a GLP-1 receptor agonist, or wherein the medicament is a substance affecting gastrointestinal transit times as a mechanism of action or as a side effect or reported side effect or potential side effect.

43. The method according to any of claims 1 to 42, wherein the patient is administered for ingestion, or otherwise ingests, an ingestible capsule, the ingestible capsule being a part of an apparatus comprising:the ingestible capsule, housing sensor hardware configured to generate readings representing motion of the ingestible capsule during passage through the GI tract of a subject, the sensor hardware including a motion sensor; anddata processing hardware communicably coupled to the sensor hardware, the data processing hardware being configured, following ingestion of the ingestible capsule by the subject, to perform a method comprising:obtaining readings from sensor hardware on-board a subject ingestible capsule, the readings being obtained during a passage of the ingestible capsule through the GI tract of a subject, the sensor hardware comprising at least a motion sensor;inputting a representation of the readings to a machine learning algorithm comprising a model trained to generate a classification of a cause of the obtained readings;processing the readings by the machine learning algorithm to generate a prediction of a cause of the obtained readings at one or more instances during the passage based on one or a time series of classifications generated by the model; andoutputting the prediction.

44. The method of claim 43, wherein, the classification of the cause of the obtained readings is a classification of GI tract location of the subject ingestible capsule at the timing of the obtained readings, and the prediction of the cause of the obtained readings is a prediction of GI tract location of the capsule at one or more instances during the passage based on one or a time series of the classifications.

45. The method of claim 43 or claim 44, wherein:the data processing hardware is housed by the ingestible capsule, orthe data processing hardware is a component of computing apparatus distinct from the ingestible capsule and configured to be located external to the subject and to receive the readings from the sensor hardware on-board the ingestible capsule during or following passage of the ingestible capsule through the GI tract of the subject.

Images