Quantified Assessment of Infant's Gross Motor Abilities Using a Multisensor Wearable

MAIJU is a multisensor wearable, and it looks like a common place with or infant size swimming to it, where the four sensors, one on each limb. MAIJU is used for recording infants movement, basically anywhere, but it's especially for when you want to measure infants movement in an uncontrolled situation like their spontaneous activity at at their homes. MAIJU is used for the assessment of motor development, monitoring the developmental milestones like rolling, sitting, walking is an essential part of the overall developmental follow up.

From the healthcare perspective, issues with the motor development are very common, and we also use motor performance as a key outcome measure in many therapeutic trials, and for both aims or purposes, we need better methods for motor assessment. Compared to the traditional standardized milestones methods, MAIJU is automated, scalable, and more objective method than the previous ones. You can go take it anywhere in the world and it would give you a comparable results.

First, select the correct suit size so that the suit fits snugly yet comfortably, and the child can move freely without interference. Start the MAIJU logger application, which will guide you through the preparations. Equip the sensors with batteries before each recording.

Pair each sensor with MAIJU logger by selecting a limb location on the app and bringing the sensor close to the mobile device. Confirm that the correct sensor number is shown on the app. Also ensure that the battery charge level is over 80%by checking the indicator below the sensor number.

After connecting each sensor with the app, place and fix it into the snap-on mount in the corresponding pocket. When pocketing the sensors follow the position guide on the mount to ensure correct orientation. Incorrect orientation will lead to unusable data.

Proceed to check correct sensor pairing. Shake the sensors one by one and check the app to see that the correct indicator is wiggling. When the checks are complete, proceed to the start page.

If needed, the duration of the recording can be set manually from the settings. You can now start the recording by pressing the record button. Wait until the sensors are ready.

This may take a few moments. For unsupervised home recordings, lock the screen of the mobile device and pack the suit for delivery to the recipient. Use a courier or similar service to deliver the suit to the recipient immediately after it has been prepared.

Ensure that the infant is nursed and feels secure for a natural playtime before you dress the suit on the infant. The suit can be removed for diaper change later if needed. Dress the suit on the infant as you would do with regular overalls.

Check that the sensor pockets are facing outwards and the suit fits snugly on the limbs at each sensor location. Adjust the straps near the sensor pockets if needed. Keep the mobile device in the same room or within 10 meters from the infant to ensure a reliable data transmission over Bluetooth connection.

Parents are not expected to handle the mobile device, so it may be practical to keep the device in the pocket that was used for the courier transport. It is not advisable to pause and then continue the recording as it may disrupt the connection between sensors and the mobile device. Arrange the surroundings to facilitate play with toys and other objects, and encourage the infant to move freely and to his or her capabilities.

The aim is to record the natural movement of the infant, so ensure that the infant feels comfortable and secure enough to engage in spontaneous play without experiencing anxiety due to new people or unfamiliar places Unless otherwise indicated, continue recording for at least a total of one hour of free play. It can consist of multiple epochs. For unsupervised home recordings, the recording is set to stop automatically.

In supervised settings, end the recording by pressing the stop button on the application. The suit can be taken off and packed for return to the laboratory. When the suit has been returned to the laboratory, the recorded data is transferred from the mobile device to the Baba Cloud servers for computational analysis, The recorded data is stored on the mobile device as a zip file, which you can export from the database in the app.

It launches an upload to the Baba Cloud website. During data upload, choose an identification number for the subject. Then add the required information as indicated by the Baba Cloud workflow.

For detailed instructions on how to use the Baba Cloud interface, see the written report. After uploading the automated analysis pipeline will work on the data and generate results in a few minutes. The results will appear on the Baba Cloud website under the given subject ID.The method quantifies infants gross motor performance by classifying postures and movements for every second of the measurement session.

This is done with an automated analysis pipeline based on machine learning algorithms. The results output from the classifier is a text matrix where time is shown in the first column and it's corresponding posture and movement categories are shown in the next columns. There may also be other classifier outputs such as carrying detection in the following columns.

The pipeline then calculates the proportion of time spent in each posture and movement type, and their age dependent variations can be presented, for instance, as violin plots. If the age of the child is given when uploading the data for analysis, the report will depict the current data in relation to the age typical proportions. For some purposes, it is more useful to combine all the movement and posture data into a holistic estimate.

One such measure is the Baba Infant motor Score or BIMS. It models the typical developmental progress of gross motor skills by predicting the child's age from all the recorded data, which is then normalized into a BIM score. For full analytic flexibility, you can always download the raw sensor data and carry out more detailed studies, such as developing abnormality detection or analyzing context dependent movement that focuses on certain postures only.

Taken together, the automated analysis pipeline will give you many alternative and complimentary measures. These measures can be used to track development or an aspect of it. They allow quantitative studying of population differences or can be used as an outcome measure in therapeutic interventions.

The metrics derived from the analysis pipeline are aimed to be generic. They can be used in many ways to answer your study questions and you can develop further measures optimized for your research. MAIJU or MAIJU-like technologies open a new door in scientific and medical research, and in particular it brings the instrumentation from the lab to the world.

It's very important in MAIJU that it's able to record infants in their own environment, which means that they, what we measure is, is their actual performance, not what they would do in a, in a perhaps scary and an odd hospital or lab environment. Taken together, I think MAIJU is a great tool to measure infant gross motor performance. However, MAIJU is, you should think of MAIJU more as an enabler rather than doer and the utility or the the, the benefits from MAIJU depend fully on your creative ways of using it.