NZ750631B2 - Monitoring and identifying laboratory animals through uwb and other digital sensory signatures - Google Patents

Abstract

method and system for the continuous monitoring of animal physiology in laboratory animal cages through the use of miniaturized sensors located throughout the cage and in various fashions including the ability to identify each rodent's digital sensory signature through data fusion and artificial intelligence. ntelligence.

Description

Monitoring and Identifying Laboratory Animals through UWB and Other l Sensory Signatures Related Applications This application claims priority to US provisional application 62/378,522, ?led August 23, 2016, whose contents are expressly incorporated herein by nce.

Technical Field One or more aspects of the disclosure relate to the monitoring of animals in a live animal research setting using various sensors including ultra-wideband (UWB) technology in various fashions including the ability to identify cage occupants through their unique digital sensory signatures.

Background Animal monitoring of vital signs in various settings, whether it be in pre- ion, surgical, post-operation recovery, home, clinical, or a laboratory setting, all bene?t from the use of monitoring techniques which do not e any physical intervention to obtain valid physiological readings.

The use of hands free automated monitoring systems is preferable to the animals being handled to take such readings as such a method does not disturb the animal, cause any harm, or increase anxiety levels. Automated monitoring techniques will also result in more te gs as identi?ed in clinical research papers what has become known as the "white coat me". In terms of ve RFID implants that may provide core temperature, glucose, and other physiology readings, these types of passive RFID techniques usually require a human-operated/hand-based reader to be present within 3 cm of the actual implant or that the animal has to be put into a separate monitoring container.

Research papers show that animals tently react negatively or positively to the close proximity of humans (even as far as varying based on the sex of the nary technician). As such, the monitoring of animals may be in?uenced by the very act of attempting to obtain the readings. This increase in an animal’s anxiety levels or the hiding of pain levels will lead to the capturing of false readings. In a laboratory g there is continued regulatory pressure to provide an environment that assists in meeting what has become known as the three "R’s" principles of Replacement, Re?nement, and Reduction as published by W.M.S. Russell and R.L. Burch. If done properly, the capturing of continuous and more reliable clinical information will allow for the development of more accurate animal computer models and therefore lead to a replacement of laboratory animals where computer models will now suf?ce. Re?nement may be achieved if the automated monitoring environment enhances an animal’s well- being and zes or eliminates unnecessary pain or distress. Reduction may be achieved if the amount of information gathered may be maximized from a given number of animals so that in the long run, fewer animals are needed to acquire the same scienti?c information.

Although the use of RFID implants helps in sing some of the negative aspects of animal handling and their effect on clinical data accuracy, as described above, it is ive and inconvenient to have technicians approach the animals with RFID wands or place them in temporary RFID reading ures.

Rodentia, in a live animal research sitting, may also be placed in temporary enclosures to measure their physiology by measuring the animal’s ical currents as detected through their feet or skin, using a pressure cuff around the tail area, or using photoplethysmography (PPG) on a limb or at the base of the tail. In all cases, the s have to be physically moved from their base cage to a l temporary cage/container with metal electrodes on the bottom, or put into special animal restraint holders that use s technologies to take readings from their tail, feet, or thigh with a clamp. After taking such readings taken in this fashion, the animals have to be returned to their home cage and the monitoring equipment disinfected for the next batch of test subjects. The above mentioned work ?ow is both cost prohibitive from a labor and a capital standpoint. Due to these factors tories have a dif?cult time meeting the standards as outlined in the "R’s" as the ring solutions cannot be economically propagated across the entire tory setting.

Summary To address the above issues, the deployment of miniaturized electronic devices is disclosed such that they may be attached to or placed in existing rodentia cages that are in use today in laboratories by s methods and left unattended for continuous monitoring. Technologies employed would include, but not limited to, UWB, mm-wave, IR, ultrasonic, capacitive sensing, RFID, bio-impedance, micro electromagnetic ?elds (EMF), piezo, passive miniature wireless pressure sensors, positron emission tomography (PET), moisture sensors, and urine and bio-marker analysis sensors, etc. Such a system is designed to be deployed at fraction of the cost of the current solutions in the marketplace to allow for scalability to the entire laboratory population t the ated labor overhead per animal.

In one ment, a small self—contained paddle is attached under each water delivery system in the cage. Inside of the paddle is a matched pair of ceramic UWB antennas, a l PCB incorporating a microcontroller, , and additional sensors such as a 6-aXis accelerometer and one or more radios such as LoRA, WiFi and Bluetooth. In other embodiments, only the UWB antenna, accelerometer, and temperature sensors are located in the paddle and the control onics are located in a central electronics ure located at the top of the cage (or bottom of the cage on a side or even on the back of the cage) with a much larger support battery. Such a central control electronics enclosure would also be able to service multiple UWB paddles or like sensors located in the cage.

In yet in another embodiment, the cage itself would be situated in a rack that would hold many cages. In this case, the paddles themselves would be connected through an armored cabled (or coax or optical cable) to a small onic switch on top of the cage. The switch in turn would be networked to a s that runs along the back of rack across each row of cages. These connections would in turn be connected to a central rack hub on top of the rack itself. This central rack hub would e power to each of the paddles through the cage top switches, as well as interact with the paddles on a two—way basis to send instructions to the paddles as well as gather the monitoring data. The central rack hub may also monitor, through on-board sensors, various ambient ions such rack humidity, temperature, light, and rack resonant frequency etc.

Whether UWB or other technologies as described above are employed in the rack, which may include dozens of other cages and multiple racks of cages, there is the chance of signal e, interference, or con?ict with regulatory bodies such as FCC on radio emissions levels. This is further compounded by the fact that there may also be several data ission radios on board associated with each sensor pod as well. To address these issues, there are several mitigation strategies available which include on-board algorithms to monitor and schedule physiological signal acquisition and data transmission per unit, rack location (ie, where a given cage is located in a rack of cages)-based algorithms to manage all the cages in a speci?c rack and room—based algorithms that are controlled by the central server-based analytical system (including but not limited to selectively activating the sensors in one or more cages to determine which sensors of which cages interfere with each other and then schedule — or operate — such that they signals from the cages do not interfere with each . Techniques could include dithering so that sensors do not ?re all at the same time, frequency hopping of transmission channels, and adjustments to the transmission power of the radios. In the case of UWB-based sensors, each individual sensor transmission may include an electronic serial number of the c sensor transmitting so that even if there was leakage into another receiving sensor’s circuitry, the received signal would be ignored. In the case of individual self- contained paddles, which are not lly controlled, radio transmission mitigation strategies may include using various techniques as described above except some of them would be stered by the cian directly or by on- board thms in their mobile tablet device as the technician is , removing, initializing, or issioning cage-based s. The self- contained s would also have sensors on-board to monitor the environment that they are operating in to cease certain types of transmissions if they detect a potential con?icting signal and to schedule and prioritize data transmissions based on radio "quiet" times or based on the importance of the data they are The paddle is attached to the cage through attachment methods that will ?t the various cage manufacture’s designs and water delivery systems whether it be c, wire or a combination of the two. Such attachment strategies could include using a type of strong elastic bands similar to what is used in human dental appliances, using vice or compression types of connectors or through the use of suction cups attached to the top, bottom or sides of the cage.

In one embodiment, to obtain vital sign readings, the rodent while drinking, leans against the paddle triggering the device to turn on. The UWB and other types of on-board sensors record a time stamped heart rate, respiration rate, relative blood pressure, and resonate frequency of the animal. Such readings may be obtained by detecting the micro nts of the heart muscles, blood vessels, and chest muscles. As a duct of the UWB’s sensing capability it would also provide an indication of a buildup of ?uids in the animal’s lungs and/or around its heart as well as a derived heart rate ility (HRV) indicator.

The captured data is then spot checked for ability and accuracy using on- board algorithms before being stored temporally in the main memory of the paddle device, stored in the top of the cage located control onics enclosure or stored in central rack-based hub. Such data is then augmented with additional sensor data that may been collected from external sensors along the way such as various RFID implant readings, ambient ature readings, ambient humidity readings, ambient light gs, and the ambient resonant frequency of the cage system. Paddle sensor activations may also be used to trigger external or internal video-based systems that may be used to monitor the animal’s or and/or be used to help identify which animal triggered the sensor array monitoring system.

The video system could be in the form a small camera with special ?lters and capabilities that is placed in the cage over the water delivery system and ed to the top of the cage electronics enclosure. Not only could this camera be used for animal identification by using advanced face recognition software or by reading the ear tag ID markings but it may also be used to measure pain or wellness levels using the image—based rodent grimace scale. Other techniques could include the ring of heart rate through image-based PPG technology looking at the minute color changes in the capillaries in the animals face and nose, looking at pre—cursors of es by examining the retina of the animal using diabetic retinopathy and/or identifying the animal by g at the unique retinal blood vessel pattern of the eyes or the unique resonant frequency pattern of the animal’s rs.

The base paddle obtained data and augmented data is then transmitted to the local server or the cloud by Ethernet cable, WiFi, Bluetooth, cellular, UWB, LoRA, or other types RF logies.

Once the data is at site of the central analytical system it is ?rst scrubbed for accuracy, then coef?cients are applied and the fused data goes into the deep ng engine where it is matched against a validated pre-known set of parameters to determine which rodent to assign the acquired sensory data to all based on their unique digital sensory signature. In any case, if the readings are outside a set of preset normal values, an alarm will be triggered that may be communicated by various methods including email, text, ?ashing the monitor screen and/or g LED’s at the cage paddle level and cage control unit level which may be picked up by authorized personnel.

In another embodiment the UWB technology is replaced or supplemented with other technologies such as ultrasound, mm-wave, bio-impedance, EFM, capacitive sensing, strain gauges, and RFID implants, etc.

In another embodiment, all of the technologies described above may be employed in combination with a common plastic or cardboard rodent t (this is where the animal goes for quiet time) y placed on the ?oor of the cage. Such an employment of the various logies may provide important animal physiological readings as well other ased measurements such as moisture (waste) levels, urine/waste analysis, biomarker analysis or the detection of water delivery system failures. Proximity sensors are used to detect the presence of a rodent in the rodent retreat and then various technologies as described above may be triggered to capture various physiological signs. Such readings may be transferred to data collection points at the cage, rack or room level using technology as described previously associated with the paddle-based sensor by wireless techniques or directly by armored cable, In another embodiment, all of the electronics and sensors are located in a smart pad that ?ts either in the bottom of the cage underneath the bedding or ?ts underneath the cage situated between the bottom of the c cage and the rack that is holding the cage in place. If the pad is placed on the ?oor of the cage, like the rodent retreat as described above, the sensors on the underside of the smart pad have the opportunity to sense and analyze the ?uids that collect below the pad including looking for biomarkers with various spectrum analysis or al reaction-based sensors. As the cage pad covers the entire surface of the bottom of the cage there is an opportunity to count the number of occupants in the cage and track their movement using the UWB technology as well as measuring heart rate, respiration and blood re for each individual . These movement artifacts may be valuable to the researcher or the nary technician to know when a rodent has had a litter or if one of the occupants has deceased as well as measuring activity and behavior.

The animals in question may have RFID implants in them, which among other things may include the [D of the . By placing RFID readers around the perimeter of the cage and at important cage locations such drinking, eating, resting, and elimination etc. various behavioral readings may be ed and eventually transmitted to the central analytical system.

To e water consumption and food consumption, small passive RFID strain gauges or other technologies may be placed under where the water bottle sits against the cage and under a pressure plate that the food is loaded on. When the RFID strain gauge is excited by various radio s located in the cage, the resulting data would be itted and added to the central analytical server for reporting.

Brief Description of the Drawings Figure 1 shows a rodent cage in a laboratory in accordance with one or more aspects of the disclosure.

Figures 2A-2C show various paddles in accordance with one or more aspects of the disclosure.

Figures 3A-3C show illustrative examples of rodents being sensed when drinking or g in accordance with one or more aspects of the disclosure.

Figures 4A-4B show different con?gurations of paddles in accordance with one or more aspects of the sure.

Figures 5A-5D and 6 show various enclosures with one or more s in accordance with one or more aspects of the disclosure.

Figures 7 and 8A-8B show alternative sensing systems in accordance with one or more aspects of the disclosure.

Figure 9 shows a sensor reading sequence in accordance with one or more aspects of the disclosure.

Figure 10 shows a process for determining which animals have been sensed by the sensors in accordance with one or more aspects of the disclosure.

Figure 11 shows an illustrative reporting system in ance with one or more aspects of the disclosure.

Detailed ption Figure 1 shows a rodent cage 100 in laboratory setting that includes two UWB paddles lOl attached to the cage below the water delivery system 102 (possibly including a spigot or tube as known in tional water delivery systems). In this particular depiction, the paddles 101 have a complete makeup of a microcontroller, sensors, battery, radios and various antennas. Therefore, it may act independently on its own or communicate interactively by various RF techniques such as Wi-Fi 103, ar, UWB, LoRA, or other RF techniques 104, and/or through Bluetooth to a nearby laboratory technician with a tablet or other type of mobile device 105. Having the tablet close by while setting up the cage with the various occupants would be useful in ing confirmation of each rodent as they drink from the water delivery system for the 1st time. This event may be correlated with the obtained digital y data to provide a unique digital ?nger print of each animal back at central analytical system 107.

Such raw, augmented and processed data may then be accessed by knowledge workers 108 located either onsite 109 or remotely located to the central analytical system.

Figure 2A shows a detailed view of a paddle version that only includes the UWB antennas, the UWB control circuitry and other sensors such as a 6 axis accelerometer and activation s of various types. The UWB paddle 101 depicted here is free moving in that it is attached to the cage adapter bracket 110 by a pin 111 and is pushed against the paddle stopper 112 by the spring 113.

When the rodent leans up against the paddle 101 it pushes the paddle backwards while the rodent makes full body contact with the paddle. The presence of the animal may be detected by a mercury switch, the on-board accelerometer, or a proximity sensor to ine that a ed event is taking place. The UWB sensor or other comparable technology is then triggered to capture various physiological readings. To indicate that a successful reading is taking place, the LED 114 on the side of the paddle changes color.

The cage adapter 110 may be attached to the rodent cage in various ways. In Figure 2A there is a depiction of small wings 115 attached to the side of the bracket where strong elastic bands may be attached to one side of the adapter and then threaded h the wire cage over to the other side. This design does not require a cage wire mesh of any particular dimension or spacing. In this particular embodiment the UWB paddles 101 communicate to a top of cage central onics controller through an armored cable 116.

In Figure 2B an exploded view of the paddle 101 is ed with the top cover 117, the electronics and antenna package 118, and the back cover 119 shown.

Also shown is the armored cable 116 that ts ly to the paddle electronics package 118. Also shown is the stopping bracket 113 and the two "L" shaped downward facing pins 111 that are used to hold the paddle 101 and springs 113 in place. As indicated, the springs 113 have both an upper coiled on and extended straight long position that also inserts into the paddle just like the "L" shaped pins 111. Included in the electronics package 118 are a matched set of UWB ceramic antennas for RX and TX functions that are either angled or situated on a ?at plane. The onboard 6 axis accelerometer may also be used to take the resonant frequency of the rodent as it leans into the paddle as well as obtaining an ambient background resonant facility frequency readings once the paddle returns to its resting position. Such readings may also be used by the UWB algorithms to reject s readings or to re-calibrate the UWB frequency domain ?lters to better isolate the desired rodent UWB signals.

In Figure 2C there is side view representation of the exploded view of the paddles 101 that is different from what is depicted in Figure 2B. In Figure 2C, the front of the paddle 121, is slightly curved and has two small protrusions 122 on it which allow the UWB antenna’s 123 to be angled to better obtain UWB readings as well as channel the rodent into the center of the paddle’s 121 front housing. This is in comparison to Figure 2C where the antennas 118 are on ?at plane. In this embodiment of the UWB paddle 101, the unit is totally self-contained in that it es all of the hardware and software components to operate and communicate independently including its own battery 124. As most short term logy studies are no longer than three weeks in length, this design would suit study protocols where the researcher has instructed the laboratory that no external wires of other electronic enclosures are to be present in the cage.

In Figure 3 there is a depiction of how the stock UWB paddle 101 may be adapted to the various cage styles and water ry systems. In Figure 3A the water delivery system 125 is at a 450 angle with a ?at nipple 126 and the cage adapter 110 is snug up against side of the cage and water bottle 125. In this case the rodent 127 displaces the paddle 101 producing a gap 128 between the paddle 101 and the ng bracket 112 and the electronic readings commence.

Figure 3B is identical to Figure 3A except that the ?at nipple 126 has been replaced with an extension spout 129. To accommodate this arrangement, the standard cage attachment system is replaced with an extension cage attachment bracket 130 which s the s 101 out further to ct with the rodent 127. In another ment, the cage ment bracket also has an upright protrusion located further along the bracket 130 with a hole in it which allows the spout 129 to pass through it and therefore providing more support for the entire assembly.

In Figure 3C the water delivery systems are at a 90° angle with a ?at nipple 126.

In this case a different cage attachment bracket 130 is used that sits around the opening for the water delivery system 125 and then extends out and down to where the paddle 101 may be attached with the retaining pin 111. Usually these types of water delivery systems 125 allow the rodents 127 to approach the ?at nipple 126 from any angle. With this embodiment, the paddle 101 is positioned just aft of the ?at nipple so that the rodent 127 may only drink from one side which activates the paddle 101 by ing it. In another embodiment, the cage attachment bracket 130 is designed to swivel around the water delivery system 125 opening so that the rodents may drink from any angle by just g the paddle 101 around to a position that suits them.

Figure 3C shows the paddle sensory array reading passive or active RFID chips 131 located inside the animal itself. RFlD readers may also be located around various perimeter positions as well as situated under the drinking and eating stations. Various physiological readings may be obtained such as glucose , core temperature, EMG, EEG, HR, respiration, bio-markers, and pressures in the brain and body etc.

In Figure 4A, the opposing wings 132 of the r are not ?xed but moveable and they may be ed to grab the wire cage, of any dimension or spacing, in a vice like fashion by using an Allen key placed in the socket 133 of a threated bolt 134 to compress or expand the wings 132 position relative to each other.

Figure 4B shows an example for an all—plastic cage top. In this case, the adapter 130 uses n cups 135 ed to the wings 136 of the UWB paddle cage attachment which will in turn attach to the underside of the plastic cage top, bottom, side of the plastic cage. The mounts that the suction cups are on also swivel 137 as well as they may be designed to extend or retract the suction cups 135 to allow for ?exibility in attaching to various surface shapes other than just ?at surfaces.

Figure 5A, 5B, 5C, and 5D show different views of one version of the rodentia electronic cage system which includes a lower unit 138, the top unit 139 that holds the food and water system, the top lid 140 the UWB paddles 101, the water delivery system 125 the connecting armored cables 116 and the top mounted cage control electronic enclosure 141. The top d control electronic enclosure may also contain RFID antennas for interrogating implants inside the animal as well communicating with other types of external RFID sensors. These types of external sensors could include ?at microchips that measure the amount of moisture or liquid in the bottom of the cage and/or e the actual urine for speci?c bio-markers. Such data collected could assist in fying which cages need cleaning attention and/or provide valuable research data.

In Figure 6 what is depicted is a rack 142 that holds several cages with bottoms 138, tops 140, water delivery s 125 and UWB paddles 101. In this embodiment the paddles 101 connect through the cables 116 to a small switch 141 that is located on top of the cage top 140. This switch in turn is attached to a wiring harness 143 that runs along the back of the rack 142 up to a central rack control box 144. The rack control box includes a power management system for supplying power to all of the cages in the rack. It also provides a central place to store all data collected from the paddles 101 and as well as other sensors that are part of the overall system, such as humidity, light levels, RFID readings, and motion artifacts etc., wherever they may be d. The l rack control box also es various ways to communicate to the local or in cloud analytical system 107 such using Ethernet cables, Wi-Fi, LoRA, Bluetooth, cellular or other RF technologies etc.

Figure 7 depicts an alternative way in the cage 138 to employ s of various technologies, such as UWB. In this embodiment, a typical plastic rodent retreat house with enclosure 145 is placed in the cage. The enclosure may be transparent, translucent, or opaque. It may also have various tints — if any — clear, blue, red, yellow, and the like. The enclosure may have various cross sections including square, rectangle, circular, oval, and may be combinations thereof (e.g, a square cross section with rounded comers). Attached to the plastic g by clips 146 or other means, is an electronics enclosure 147. In another embodiment the electronic enclosure actually sits inside the plastic enclosure and not attached to the . less of its position, the enclosure includes a PCB containing a micro-computer, memory, a y 148, proximity sensors and radio antennas 149. Physiological sensors, including RFID readers 150, are located both on the top of the onics enclosure and on the bottom of the enclosure. There are also radio ission antennas 151, LED indicators 152 and an on/off switch 153. As the animal enters the ure 145, a proximity sensor 149 activates the sensors 150 and starts the monitoring process. In another embodiment, the actual electronics pad may be designed and manufactured to be much thinner as the bulky electronic items such as the battery, LED’s and switches 148, 152, 153 etc, may be moved to the side or the end of the unit in a form of a tab, Data from this type of design may be transferred to the central analytical system 107 using various on-board radio technologies or by an armored cable attached to a switch 141 on the top of the cage. All previously mentioned techniques for managing sensor data as described above at the cage, rack, or room level are also applicable.

In Figure 8 there is a depiction of a smart cage-based pad in two different versions. Figure 8A depicts a typical rodent cage 138 where the smart cage pad 154 is situated on the bottom of the cage. The pad has an array of sensors 155, ing UWB sensors, located on the top of the pad and other sensors d on the bottom of the pad 156. The pad has area designated to provide a user interface 157 in the form of LED’s and a switch in this example. This user interface may be situated on the side of the pad as depicted here or it may be on the end facing out for veterinary technicians to review similar to what is depicted in Figure 8B. In Figure 8B, there is a depiction of a smart cage pad that is placed under the cage 138 bottom between the cage bottom and the top a racking system that holds many cages (not shown here). In this depiction, the user ace as described in 8A is replaced with an electronics tab 158. This tab may hold bulky electronic items such as the battery, LED’s switches etc, allowing the pad to be much r. The increased size of the electronics tab allows for the inclusion of a display screen (not shown) versus just relying on LED’s to convey important ional status information of the smart pad. Regardless of location, the smart pad may perform all of the operations of the rodent retreat pad as described in Figure 7, except the when the pad in placed under the cage bottom it of course does not have direct access to the waste material ore it would have to rely on other technologies on the top side of the sensor array to perform analysis of the waste material and liquids. In either case, as the cage pad covers the entire surface of the bottom of the cage there is an unity to count the number of occupants in the cage and track their movement using the UWB technology as well as measuring heart rate, respiration and blood pressure for each individual animal. Either of depictions shown they may also be connected ly to cage- based or rack-based external power supplies and communications modules by an armored cable (not shown).

Figure 9 shows a ?ow chart of the sensor reading sequence that may be used on various embodiments described herein. Once the UWB paddle sensor 101, the electronic rodentia retreat 147 or the smart electronic cage pad 154 has been activated 159, by either the animal ring the on-board accelerometer, mercury switch or proximity switch etc, it checks 160 see if a reading is scheduled. If so, the system then performs mitigation strategies 161 to see if there are any ting RF signals inside the electronic enclosure itself, at the cage level, at the rack level, or at the room level. If all is clear, the sensors in the paddles, retreat pad, smart cage pad begins and/or suspends its UWB sensing or RF communicating sequence 159. If a reading is not scheduled, the sensors return to a resting state 159. Once the UWB reading sequence is finished, the quality of the reading is analyzed locally 163 to determine if there were any gross noise artifacts that would disrupt the reading s such as the paddle 101 not being displaced totally or that the duration of the reading was too short etc. In the case of error conditions, the counter would not be incremented and the sensors would return to a resting state 159. If the readings were deemed successful, the readings would then be stored locally 164 and made available for the next scheduled data transmission 165 and the paddle ed to the resting state 159.

Figure 10 depicts the logic work ?ow for determining what possible readings could be obtained by the sensors when they are activated and by which animal 159. In this case the data from all sensors and sources connected to the system is available to the l analytical system 107. As well as the facility, rack and cage number, this may include animal ID 166 as obtained through RFID implants or video analysis (ear tag etc), blood re 167 HR and HRV readings 168, respiration readings 169, ambient facility/rack/cage temperature 170, core temperature, glucose levels, and pressures levels within the s’ body and brain 171 as ed through RFID implants, image—based sensors and/or infrared (IR) sensors located on the paddles, the electronic rodentia retreat pad, or the smart cage pad, the time of the encounter 172, resonant ncies and activity levels of the animal 173 and the ambient resonant frequencies of the cage itself.

In another embodiment, additional readings may be obtained at the paddle level with the addition of two metal ts which may be designed to make contact with the rodent’s front paws as they lean up to drink. Once contact is made, a small electrical current may be passed through one paw to the other paw and electrical impedance myography (EIM) signals may be ed to derive the animal's body ition 174. Such a reading is more than likely to be unique in comparison to the other cage occupants and therefore provides yet another variable to assist in deriving a unique digital signature per cage rodent.

In another embodiment, small al RFID moisture, urine and other bio- marker waste al analysis sensors 175 may be placed in the bottom of the cage or on the bottom of the rodentia retreat’s electronic pad or the smart cage pad that may provide insight into various reactions that the rodents are having to the protocol or to indicate a leak/?ood in the water delivery system.

As mentioned usly, another embodiment is the inclusion of a small video camera attached to the cage that is operated externally or connected by various means to the top of the cage central electronic enclosure 144. Not only could this image-based analysis provide possible lD’s of which rodents at which time were being monitored, but they could also provide important clinical indicators based on the e scale (GS) 176 to monitor pain levels. Another aspect of video is is using retina scanning looking for early signs of diabetic retinopathy as well as using the unique vein pattern of the retina for animal identi?cation purposes 177.

All of these values are compared 179 on a raw data basis and on a data fusion basis to all prior con?rmed readings located in the veri?ed history database 180.

If the readings 181 provide a high enough con?dence factor that they match a speci?c rodent, they are then transferred 182 to the veri?ed history ?le 180. If the readings look like total outliers, then they are dumped 183 into the reject data ?le 184. For valid but unknown animal readings they are marked as such 185 and recorded in the composite cage health ?le 186. In both the known and n rodent readings, instances of the data are also sent to the alarm sing module 187 which is used to determine whether 188 an alarm should be raised and sent out by various onic or display methods 189 otherwise no alarm is sent 161. Alarms may also be based on water, food, or waste levels being monitored.

Periodically all three ?les will be re—analyzed as new data becomes available, from the s or from the direct ational methods, looking for better ways to derive individual digital signatures of each animal and to re-assign data from one category to another, if necessary, using the deep learning engine.

Figure 11 shows a reporting system that provides both ng and a snapshot of the results obtained, augmented and processed. In the graph the bottom scale is for respiration rate per minute 190 and the top part of scale is representing heart rate per minute 191. Along the bottom of the graph is a time slide bar 192 based on timed readings of 60 times an hour that allows for the researcher to move the slide bar to speci?c time intervals of the day 193 or to the beginning of an alarm al etc. Below the graph, the researcher may select either sensor array A 193 or sensor array B 194 in the cage or a combination of the two 195. Along the top of the graph is the identi?cation row of information related to the speci?c cage being reviewed including the date 196, cage location 197, cage number 198, and rodent ID 199. Situated directly below is an information scroll bar of all of the le physiological signs available such as glucose 200, core ature 202, ambient temperature 203, blood pressure 204, etc. The researcher may scroll 200 through all of the available data that is related to that speci?c animal and time interval selected 192.

A number of embodiments have been described where it is understood that various modi?cations may be made t departing from the spirit and scope of the disclosure.

I/WE

Claims (15)

CLAIM 1.:

1. A system comprising: an enclosure for an ; at least two sensors, one of which is a UWB radar, located relative to the enclosure and the other sensor is a ce sensor; a water delivery system sing at least one of a nipple or a tube; and a paddle located below the water delivery system, wherein the UWB radar is located in the paddle; a processor configured to receive signals from the two sensors; and a memory configured to store information relating to information contained in the signals received at the processor, wherein the sensors are in the enclosure and located at feeding/watering station in the enclosure, wherein the ce sensor indicates whether the animal is in physical contact with the paddle, and wherein the UWB radar is active only while the animal is in al contact with the paddle as indicated by the presence .

2. 2. The system according to claim 1, wherein the enclosure is sized to accommodate two or more rodents.

3. The system according to claim 1, further comprising: a schedule clock; a memory storing a schedule of reading times relative to the le clock, wherein the sor checks a time of the schedule clock and compares the time to the schedule of reading times and operates at least one of the sensors at scheduled time.

4. The system according to claim 3, wherein the processor determines a quality of data from the at least one of the sensors.

5. The system according to claim 4, wherein, when determined to be a sufficient quality, the processor stores the data.

6. The system according to claim 5, AH26(24625011_1):MSD wherein, when determined to be less than sufficient quality, the processor stores the data with information regarding the quality of the data.

7. The system ing to claim 1, wherein the processor is configured to determine, based on a return signal from the UWB radar, a heart rate of the animal in physical contact with the paddle.

8. The system ing to claim 7, wherein the processor is further configured to determine, based on the return signal from the UWB radar, a heart rate variability of the animal in physical contact with the paddle.

9. The system according to claim 1, further comprising an ambient temperature .

10. The system according to claim 1, further comprising an RFID sensor.

11. The system according to claim 1, wherein the presence sensor in the paddle is an accelerometer.

12. The system according to claim 1, further comprising a urine analysis sensor.

13. The system according to claim 1, wherein the processor is further configured to determine, based on the return signal from the UWB radar, a blood pressure of the animal in physical contact with the paddle.

14. The system according to claim 1, further comprising: a first storage in which red data of a threshold level with an identification of a monitored animal are stored; and a second storage in which monitored data of a threshold level but g an identification of a monitored animal are .

15. The system according to claim 1, further comprising: AH26(24625011_1):MSD a first storage in which monitored data of above a threshold level are ; and a second storage in which monitored data below the threshold level are stored. i4C Innovations Inc. By the Attorneys for the Applicant SPRUSON & FERGUSON Per: AH26(24625011_1):MSD