ADVANCED PLAY ENVIRONMENT FOR SCREENING AND EARLY DIAGNOSIS OF INFANT DEVELOPMENTAL DELAYS AND NEUROLOGICAL IMPAIRMENTS

Abstract

The presently disclosed subject matter relates to a play environment for children and infants for detecting motor delays or impairments, evaluating neurological development, and for diagnosing developmental disorders, and methods and systems of using the same.

Claims

1. A play environment for evaluating neurological development in a child comprising: a. a mat or a toy, and b. one or more sensors, c. wherein said sensors are adapted for measuring at least one type of movement or cognitive data.

2. The play environment of claim 1, wherein said movement or cognitive data is selected from the group consisting of: touch, grasp, force, acceleration, velocity, orientation, position, pressure, attention, stimulus-response, or play behavior.

3. The play environment of claim 1, comprising a toy, wherein said toy is embedded with one or more sensors, said one or more sensors are adapted for measuring at least one type of movement or cognitive data.

4. The play environment of claim 3, wherein said toy further comprises a feedback mechanism that generates sound, vibration or light.

5. The play environment of claim 1, comprising a mat, wherein said mat is embedded with one or more sensors, said one or more sensors are adapted for measuring at least one type of movement or cognitive data.

6. The play environment of claim 1, further comprising a framework.

7. The play environment of claim 6, wherein said framework comprises a cross connection, wherein said cross connection comprises two to four flexible bars.

8. The play environment of claim 6, wherein said framework is collapsible.

9. The play environment of claim 7, wherein said framework further comprises one or more adjustable parallel bars attached to the flexible bars of the cross connection, wherein said parallel bars are detachable from the flexible bars.

10. The play environment of claim 6, comprising a toy, wherein said toy is attached to said framework.

11. The play environment of claim 1, wherein said sensor is an imaging sensor.

12. The play environment of claim 11, wherein said imaging sensor comprises a 3D Kinect imaging system.

13. The play environment of claim 11, wherein said imagining sensor comprises a GoPro.

14. The play environment of claim 1, wherein said child is an infant.

15. The play environment of claim 14, wherein said child is an infant between the ages of 2 months and 11 months old.

16. A method for detecting motor delays or impairments in a child comprising a. placing a child in a play environment which comprises a mat or a toy and one or more sensors, wherein said sensors are adapted for measuring at least one type of movement or cognitive data and b. measuring at least one type of movement or cognitive data.

17. The method of claim 16, further comprising measuring a plurality of movement data points or a plurality of cognitive data points.

18. The method of claim 16, comprising measuring at least one type of movement data and at least one type of cognitive data.

19. A method for monitoring infant motion comprising a. placing an infant in a play environment which comprises a mat and one or more sensors, wherein said sensors are adapted for measuring at least one type of movement or cognitive data and b. measuring at least one type of movement or cognitive data.

20. The method of claim 19, wherein said at least one type of movement or cognitive data is pressure data.

21. The play environment of claim 15, wherein said mat is located on at least four load cells, wherein the at least four load cells are configured to measure ground reaction force, location of the ground reaction force, movements of the infant on the at least four load cells, or combinations thereof, wherein the ground reaction force is a sum of forces on the at least four load cells.

22. The play environment of claim 21, further comprising at least one camera that record and transmit actions of the child for analysis, wherein the actions are selected from the group consisting of head actions, left and right arm actions, left and right leg actions, torso actions, and a combination thereof.

23. The play environment of claim 22, wherein the play environment is configured to track the at least one type of movement on the mat by detecting movements of arm, body, legs, head or combinations thereof.

24. The method of claim 19, wherein the at least four load cells measure the location of the ground reaction force and the movements of the infant by tracking a center of pressure generated by the infant.

25. The method of claim 19, wherein a toy is hung within arm's reach of the infant to assess an upper limb impairment.

26. The method of claim 19, wherein a toy is hung within leg's reach of the infant to assess a lower limb impairment.

27. The method of claim 19, wherein said movement or cognitive data is selected from the group consisting of touch, grasp, force, acceleration, velocity, orientation, position, pressure, attention, stimulus-response, play behavior, pressure, haptic interaction, kinetic interaction, frequency of a toy stimulus event, frequency of a toy feedback event, response time, frequency of touch response, and frequency of look response and a combination thereof.

28. The method of claim 27, wherein the haptic interaction comprises frequency and mean force of the grasp.

29. The method of claim 27, wherein the kinetic interaction comprise frequency of arm's reach, frequency of leg kick, time to toy contact, toy movement, maximum toy displacement, toy contact duration, or a combination thereof.

30. The method of claim 19, wherein the measuring comprises measuring degree, entropy, and/or frequency of the at least one type of movement, wherein the at least one type of movement is a repeated movement.

31. The method of claim 28, further comprising identifying atypical movements and/or typical movements based on the measured data.

32. The method of claim 31, wherein the atypical movements are symptoms of cerebral palsy.

Description

BRIEF DESCRIPTION OF THE FIGURES

[0017] FIG. 1. shows an example play environment for observing and evaluating neurological development.

[0018] FIGS. 2A-2C. show example toys that are sensor-equipped.

[0019] FIG. 3. shows an example play environment that includes a framework from which toys can be hung and imaging sensors can be attached.

[0020] FIGS. 4A-4F show example toys that can be utilized in the play environments or the methods of the disclosed inventions are depicted. Example sensors and feedback mechanisms that can be included in toys are illustrated.

[0021] FIG. 5. shows an example system for data collection and information dissemination.

[0022] FIG. 6. shows representative data displaying the X, Y and Z displacement of the elephant toy during a specific infant trial.

[0023] FIG. 7. shows an example plot displaying how the elephant toy moved in the XY plane during a specific infant trial. The star marks the original position

[0024] FIG. 8. shows a diagram illustrating how example mat data can be collected.

[0025] FIG. 9. shows a plot of example mat position data.

[0026] FIGS. 10. A-C. shows an example play environment utilizing multiple cameras.

DETAILED DESCRIPTION

[0027] The present disclosure is directed to a play environment comprising sensor-equipped toys and/or a sensor-equipped mat to evaluate neurological development in infants, as well as methods of using the same in the early detection of developmental disorders or warning signs thereof.

A. Definitions

[0028] According to the present disclosure, a child is a human under the age of 18 years old. An infant is a child under the age of one-year old.

[0029] As used herein, the term framework refers to any structure which can be placed above or around an infant or small child, and to which toys or other objects can be attached.

B. Play Environment for Measuring Motor Responses

[0030] In certain embodiments, the present disclosure is directed to a play environment comprising one or more sensor-equipped toys (3), which can be used to measure motor responses in an infant or small child and to detect early signs of motor delay. The sensor-equipped toy can be any toy which is embedded with one or more sensors. In certain embodiments, the sensor can measure movement or cognitive data. In certain embodiments, the sensor can measure contact, touch, grasp, acceleration, toy position and/or orientation. In preferred embodiments, the sensor-equipped toys are age-appropriate and visually similar to common toys for infants and small children. In certain embodiments, the sensor-equipped toys are attached to a framework (1) comprising one or more bars, which may be made of flexible or inflexible material. The framework may be constructed using any materials and techniques used to build existing play environments. In certain embodiments, the play environment may be constructed from industrial parts and/or using a 3D printer. An example play environment framework (1) is depicted in FIG. 1. The framework (1) includes a cross connection to stabilize the structure and allow for easy break-down and portability, as well as adjustable parallel bars attached to the cross connection which can be used to hang toys above the infant. The surrounding mat area around the structure allows for the placement of peripheral toys that the child can interact with as they begin to develop motor skills. Sensor-equipped toys hang from the parallel bars within arm's reach of the infant, and within foot's reach of the infant while in the supine position. Another example play environment framework is depicted in FIG. 10C.

[0031] Sensors used in certain embodiments are adapted for measuring and/or collecting movement or cognitive data. Non-limiting types of movement data include: contact, touch, grasp, force, acceleration, velocity, orientation, position and pressure data. Movement data, for example, can include physical interactions. Certain physical interactions can be sub-divided into kinematic and haptic interactions. Non-limiting examples of kinematic interactions include: frequency of arm reach, frequency of leg kick, time to toy contact, toy movement, maximum toy displacement, or toy contact duration. Non-limiting examples of haptic interactions include: frequency of grasps or mean grasp force. In certain embodiments the sensors are adapted for measuring at least one type of cognitive data. Non-limiting types of cognitive data include attention, stimulus-response and play behavior data. Non-limiting examples of measuring cognitive data include measuring: frequency of toy stimulus event, frequency of toy feedback event, response time, frequency of touch response, frequency of look response.

[0032] In certain embodiments, the play environment comprises a framework with a cross connection or a truss connection designed to stabilize the structure and allow for easy break down and portability. Adjustable parallel bars may be attached to the framework to allow for the hanging of toys within arms' reach of the infant, and within foot's reach of the infant while in the supine position. In certain embodiments, the play environment also comprises a mat (2) which is placed below or around the framework. The surrounding mat area around the structure allows for the strategic placement of peripheral toys with which the infant or child can interact as they begin to develop motor skills. In certain embodiments, the mat is a pressure-tracking mat which is equipped with pressure sensors to detect movement.

[0033] In certain embodiments, the infant or small child is placed underneath and within the perimeter of the play environment. Alternatively, the play environment may be placed above and/or around an infant or small child such that the sensor-equipped toys are above and about the child to elicit responses such as touching, kicking, hitting, reaching, squeezing, or grasping.

[0034] In certain embodiments, the play environment further comprises an eye-tracker to monitor the baby's gaze and reaction to visual stimuli. In certain embodiments, sensor-equipped toys that make sounds may be used with the play environment to detect deafness as well as motor disorders.

[0035] In certain embodiments, wireless data transmission can be integrated with the invention such that an infant's play session can be wirelessly transmitted to a pediatrician or other medical professional for analysis. Example embodiments are depicted in FIG. 5. Data is collected from sensors in the play environment (13) and transmitted to a central hub (14). From the central hub, the data is further transmitted to either or both of an activity app that can be monitored on a cell phone or other device to be accessed by a parent (15) or to a computer to be accessed by a physician (16).

[0036] In certain embodiments structure of the proposed system will have a framework, sufficiently sturdy to accommodate infants of different sizes, to allow an adjustable mount for cameras and to allow adjustable mounting for hanging toys (see H bar, FIG. 3 and FIG. 10C). In certain embodiments the square footage will be about 9 square feet to about 16 square feet and will accommodate toys at the periphery (FIG. 3, see duck). A pressure sensorized mat such as those made by Tekscan or Xsensor or a sensorized mat with four load cells at each corner placed under a 4 ft4 ft dragon plate will be placed between a blanket and the base of the gym; this mat will be used to measure postural changes to detect rolling, crawling and other movements away from the initial posturereasonable specifications would be sensor density 0.25-1 sensor/cm.sup.2 with a spatial resolution about 1 cm.sup.2, adapted for measuring vertical reaction forces up to 120 Newton (Donati 2013). In certain embodiments, an imaging sensor may be used, such as a 3D motion capture system. In certain embodiments, the 3D motion capture system will include stereo cameras mounted to the framework structure to enable collect video data and the infants' reaching and body kinematics in response to the toys. In certain embodiments, a 3D Kinect camera may be used as the motion capture system and if used, the Kinect camera will be mounted at least 1.2 m to get good resolution.

[0037] In certain embodiments, a computer will collect data from the wireless sensor network (WSN) via one or more microprocessor as well as the data from the pressure mat and the camera. In certain embodiments, a wireless chip such as Xbee can be used in connection with the toy-based platform; this chip can communicate with a microprocessor via another wireless unit located off the gym. In certain embodiments, data can be temporally synchronized using a digital pulse generated by a go-button and spatially synchronized to the inertial frame of the camera location. In certain embodiments, a visual parsing and tracking program can analyze the sensor data including the camera data; e.g., to detect the infant torso, each hand and foot of the infant and extract position over time.

TABLE-US-00001 TABLE 1 Examples of Toys to be used. Hanging Toy Action-Stimulus *light or Toys location sound will be reward Infant Posture Target Age 1. Toy animal for At feet Shake, Sound/Music Supine, Seated 2-5 leg reach (Lion) (FIG. 4C) 2. Ring Toy: soft Side or Grasp, Shake, Squeeze Supine, Seated 3-5 or hard Center 3. Plush Side or Squeeze Tips and nose Supine, Seated 3-5 Elephanttoy Center (FIG. 4A) 4. Organtan Side or Pull apart/put together Supine, Seated 7-10+ (two part toy) Center (FIG. 4B) Peripheral Toys Action-Stimulus Target Posture Target Age 1. Monkey Toy Outer 1 Grasp, Hit Seated, Crawl 8-10+ (FIG. 4D) 2. Tree Toy Outer 2 Grasp, Squeeze Seated, Crawl 8-10+ (FIG. 4E) 3. ball Outer 3 Grasp, Roll, Shake Seated, Crawl 8-10+ 4. Alligator toy Outer 4 Grasp, Squeeze Seated, Crawl 8-10+ (FIG. 4F)

[0038] In certain embodiments these toys will be modified commercial toys. In certain embodiments these toys will be custom-made toys. In certain embodiments, each toy can have a unique identifier so that its position will be known with respect to a designated inertial frame. In certain embodiments, each toy can be sensorized appropriately. Toys can be equipped with one or more pressure and force sensors to detect some level of infant grasping forces from gross and fine motor manipulation. In certain embodiments, inertial sensing units (IMU) (gyroscope, accelerometers, and magnetometer) can be used to detect twisting, turning, shaking, actions experienced by the toy. In certain embodiments, the toys provide feedback, such a sound, vibration or light, depending on whether the infant grasps, shakes or touches it. In certain embodiments, and for safety, short tethers can be used with the toys to allow handling of the toy without loss during handling. In certain embodiments, the outer parts of toys can be washable or wipeable. In certain embodiments, electronics can be removable from the outer part to enable ease of washing. In certain embodiments, toys should enable grasping of a key part.

[0039] Examples of toys with that can be used in play environments are depicted in FIGS. 2A-2C and 4A-F. In FIG. 4A, the example toy, which is an elephant, has a pressure sensor (5) and a feedback mechanism, which produces vibration (6) in each ear. The elephant toy also has a pressure sensor embedded in its trunk and an IMU (7) as well as a feedback mechanism that produces sound (8) in the elephant's head.

[0040] In FIG. 4B, the example toy, which is an orangutan, has two pressure sensors (5) embedded in each arm. The arms are attached to a feedback mechanism that produces sound (8) when the arms are apart and separated. The orangutan toy also has an IMU (7) as well as a feedback mechanism that produces sound (8) in the orangutan's head.

[0041] In FIG. 4C, the example toy, which is a lion, has an IMU (7) and a feedback mechanism that produces light (10) in the lion's head

[0042] In FIG. 4D, the example toy, which is a monkey, has embedded pressure sensors (5), feedback mechanisms that produce vibration (6) and sound (8), and magnets in the arms (12) that are attached to a reed switch (11).

[0043] In FIG. 4E, the example toy, which is a tree, has embedded pressure sensors (5) and feedback mechanisms that produce vibration (6), sound (8), and light (10).

[0044] In FIG. 4F, the example toy, which is a crocodile, has embedded pressure sensors (5) and feedback mechanisms that produce vibration (6), sound (8), and light (10).

[0045] TOYS FOR YOUNGER INFANTS (2 months-7 months). In certain embodiments, toys above head can be accessible to the baby supine or seated and toys placed at the periphery can be accessible to the crawling baby. In certain embodiments, five toys can be hanging (1 at each foot, 1 over left side, 1 over right side, 1 center) (Table 1). In certain embodiments, and to identify reaching behaviors with the feet, a toy can be hung by the left and right foot of the infant. In certain embodiments, the toy can provide sound when the infant kicks it. To identify reach and/or grasp behaviors with the left hand, right hand or both, toys can be hung to the center, right, and left of the baby. Kicking behaviors should be evident at 2 months. Reaching and gross grasping behaviors should be evident starting at 3 months and finer motor grasping should be evident at about 4 months (Ho 2010). In certain embodiments, and to identify any differentiated bimanual actions where the right and left arm has different roles in a bimanual task, one toy can be a two part toy that the infant can put it together or take it apart. This insert and remove behavior should be very evident starting at 7 months (Kimmerle 2010). In certain embodiments, the toy can provide feedback, such as sound, vibration or light for the removal or insertion action. In certain embodiments, the other toys can be single part toys that cannot come apart such as a ring toy.

[0046] TOYS for OLDER INFANTS (7 months-10+ months). In certain embodiments, and to identify intentional crawl and roll movements toward a toy, three toys can be placed on the periphery to engage the infants who begin to crawl (Ho 2010). These toys should encourage baby to go after thembut they should not move to far away e.g., a hit, a squeeze causes a movementcrawling, rolling etc. but doesn't move too far away; and encourages baby to repeat action.

C. Methods for Evaluating Neurological Responses and Detecting Developmental Disorders

[0047] In certain embodiments, the present disclosure is directed to a method of evaluating neurological development using the play environments of the present disclosure.

[0048] In certain embodiments, the present disclosure is directed to a method of diagnosing a developmental disorder using the data collected from the use of the play environments of the present disclosure.

[0049] In certain embodiments, the play environments can be used by hospitals and private pediatrician practices. In other embodiments, parents can buy or rent the play environments of the present disclosure from their hospital or pediatrician, and use it for evaluation at home. The play environment of the present disclosure can also be deployed to remote areas, including third world countries.

TABLE-US-00002 TABLE 2 Primary and Secondary Measures (*collected daily for 5 days): Y = Young (3-5 months) Variable Ex. Construct Measure Measure Ref. Sample Item(s) Ex. Averages (STD) Unimanual (L, R) % Frequency Sgandurra 2012 % grasping actions % freq grasp to center and Bimanual Arm Corbetta 2008 % reaching actions toy: TYPICAL: L; use (B, D-B) Kimmerle 2010 Y: 16.67(4.16)%; 0: 35(7.26)% Leg Use (L, R) % Frequency % kicking actions to elicit stimuli Grasp Forces Mean Force Cecchi 2011 Magnitude of squeezing or TYP: L (Newtons) or Pressure Guzetta 2014 pinching action Y: 8.07 N, 0: 12.32 N Sgandurra 2012 Arm Use: Position Mean Bhat 2006 Distance of each wrist to R distance to toy: TYP Distance toy Y: 250 mm, 0: 185 mm Arm use: Speed Mean Speed Bhat 2006 Mean hand speed from R speed to toy: TYPY: baseline 170 mm/s; 0: 220 mm/s

EXAMPLES

[0050] The following examples are put forth so as to provide those of ordinary skill in the art with a complete disclosure and description of how to make and use the presently disclosed subject matter, including the use of the play environments of the present disclosure to measure motor responses and evaluate neurological development. The following examples are not intended to limit the scope of what the inventors regard as their presently disclosed subject matter. It is understood that various other embodiments may be practiced, given the general description provided above.

Example 1

[0051] FIG. 1 depicts a prototype of a play environment of the present disclosure, called a SmarToyGym. The play environment visually appears similar to existing toys for infants. Sensor-equipped toys dangle down in front of the child, while the baby's movements are captured using a Microsoft Kinect Sensor and a mat equipped with pressure sensors. The framework for the play environment was built using industrial parts and a 3D printer. A 3D Kinect imaging system is currently embedded into the prototype and software has been developed. The Kinect can be a component of the advanced play environment and it can discriminate kinematic movements of atypical vs. typical infants in response to the presence of a toy. These toys can be networked and embedded with sensors to measure touch, grasp, acceleration, toy position and orientation. They can provide light and sound stimuli and the infant's head turning, leg kicking, hand touch and grasp responses can be collected by the embedded processor and analyzed by a custom GUI (this may be on a PC or via an app). The data from the child's movements can be compared with past datasets (for example, the previously past data sets compiled in Table 2) to provide an early warning of developmental disorders, and datasets can be built for identifying problems. The device is not specific to any one disorder, for example, toys that make a sound could be used to detect deafness as well as motor disorders.

[0052] FIG. 10C depicts another version prototype of the play environment of the present disclosure where 3D stereo cameras are used instead of the Kinect camera.

Example 2: SmarToyGym

[0053] A study was performed to quantify the physical and cognitive interactions of low-risk and high-risk infants in a sensorized play environment in a cross-sectional study to investigate whether there are differences between the two groups, and if these differences are indicative of possible motor delays and impairments. A goal is that the results from this study can highlight the specific metrics that are important to focus on when separating the low-risk and at-risk infants.

[0054] To analyze the data, two different types of metrics are measured: physical and cognitive infant interactions with the toys. Physical interactions can be split up into two subcategories: kinematic and haptic interactions. Table 2 consolidates and displays the wide range of variables that can be collected.

TABLE-US-00003 TABLE 2 Variables analyzed in the SmarToyGym project. Type of Interaction Variable Description Kinematic Frequency of arm reach The number of reaches, split into left, right, and bimanual reaches Frequency ofleg kick The number of kicks, split into left, right, and bimanual kicks Time to toy contact (s) Time in seconds of first infant-toy contact Maximum toy displacement The maximum displacement in the XYZ direction Toy contact duration (s) Time in seconds of infant interaction with toy Haptic Frequency of grasps The number of grasps, split into left, right, and bimanual grasping Mean grasp force (psi) Grip force in psi Cognitive Frequency of toy stimulus event Frequency of events to prompt the infant's attention Frequency oftoy feedback event Frequency of key toy events triggered after infant interaction Response times (s) Time in seconds of infant's response to events Frequency of touch response The number of times the infant touched the toy after an event if previously not engaged Frequency of look response The number of times the infant looked at the toy after an event if previously not engaged

[0055] Two sensors are positioned inside the toys: an IMU and a pressure sensor. Each toy is rigidly connected to an electronic box and is able to rotate about the XYZ axes about one fixed point. To quantify toy displacement, a robotic model based on Denavit-Hartenberg parameters was used. The robotic model used is similar to a rigid pendulum suspended from a spherical joint with 3 degrees of freedom (revolute joints). A zero reference frame was set in the system to graphically display toy movement based on these calculated displacements. Image data analysis confirmed these variables.

[0056] From this study, representative data displaying XYZ displacement of the elephant toy during a specific infant trial is shown in FIGS. 6 and 7. FIG. 6 depicts the displacement of the toy in the X, Y, and Z directions on separate 2D plots. FIG. 7 shows the XY displacement in a single 2D plot, with a star denoting the toy's starting position. In the depicted example, the infant moved the toy at a maximum displacement of x=9.9 cm, y=23.5 cm, and z=10.3 cm.

Example 3: SmarToy Vision System

[0057] A play environment was developed with imaging sensors, referred to herein as the SmarToy Vision System. This involved developing a imaging system that would allow for multiple view 3D tracking of the infant under natural play conditions. A stereoscopic imaging framework that can support off the shelf camera devices, such as a GoPro or Kinect, but still generate accurate depth maps, can be utilized. The example SmarToy Vision System depicted in FIG. 10 utilizes 4 GoPro stereo cameras (4). The cameras in the example system are attached to the framework (1) and aimed at the mat (2) as depicted in FIGS. 10B and 10C.

[0058] Given the size constraints of the play environment of the present example, a 3D stereo camera framework was employed over depth cameras such as the Kinect because of the ability to manually specify the stereo baseline to accommodate depth estimation of nearby objects. To become robust to interaction based occlusion a dual stereo camera setup can be used (as depicted in FIG. 10A) with each setup positioned almost orthogonally to each other (as depicted in FIG. 10C). This allows the play environment to track the infant from almost complementary viewpoints, giving it the ability to view the infant's limbs at any given point of time free of severe occlusion.

Example 4: Mat Data

[0059] In this example, data was collected from a play environment which had sensors embedded in a mat. The raw mat data is collected in the way depicted in FIG. 8. The force sensor data collected from each comer of the mat is resolved into center of pressure data as shown in this figure. Each x denotes the babies center of pressure from the beginning of the trial to the end of the trial. This information indicates how the infant moved.

[0060] This raw data, centered around 280 with a span of 560 units (roughly 1.8-3.6V converted to analog), is saved by the GUI into a .txt file. This file is then processed by the matlab script SmarToyMATpostProcessing_vl.m, which takes the data and performs a linear mapping to x,y coordinates 24 to 24 along each dimension. The red comer (bottom left when viewed from above) is 24, 24, and the blue comer (top right) is 24,24. Alongside the raw data, calibrations are taken before each trial by placing a known mass in each mat comer and then in the center to ensure consistency across data sets. This is factored into the post-processing script to determine comer values and locations. The output of this script is an array of position data and a plot of baby motion on the mat, as can be seen in FIG. 9.

[0061] Processing can occur via a standard Matlab script, which focuses on the y-axis motion, also known as caudal-cephalic motion (along the axis of the spine). Center-of-pressure (COP) movement along this axis has been found to be distinct between healthy and non-healthy infants (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2794478/ which uses preterm vs. full term infants). The script can measure root mean square (RMS) displacement and approximate entropy (ApEn) along the y-axis. Larger RMS and a smaller ApEn is expected in healthy infants. RMS can be measured as the sum of deviations from the mean COP. ApEn is a measurement of the predictability of a data set, and is significantly more complicated in its calculation. It is used commonly in measuring repetitive biological signals such as EKG, as well as motion data. An overview is found at https://www.physionet.org/physiotools/ApEn/ and the aforementioned paper gives a number of evaluation constants which would also be applicable to the present data set.

[0062] Overall, the analysis can confirm the occurrence different events based on the play environment data collected, including vision-based baby tracking (Center of Pressure and limb tracking) and mat pressure (Center of Pressure).