Smart Computing Models, Sensors, and Early Diagnostic Speech and Language Deficiencies Indicators in Child Communication (SmartSpeech)

This clinical trial aims to justify a protocol for designing and developing an automated decision-making system to support and enhance screening and early detection procedures for developmental speech/language difficulties in child communication. The system will utilize smart computing models, sensors, and early diagnostic speech and language deficiency indicators. The study participants will be typically and non-typically (neurodevelopmentally atypical) developing children, primarily in preschool and elementary school.

The key research questions the study seeks to address are:

1. Do differences exist between typically developing and neurodevelopmentally atypical children when engaging with the serious game used in the study?
2. If such differences are identified, can the collected data be utilized to train an automated decision-making system capable of accurately distinguishing between typically and neurodevelopmentally atypical children based on speech and language deficiencies?

Researchers will compare typically and non-typically developed children to see if the system can incorporate multiple data points from assessment domains to create a diagnostic profile.

After the parents are informed of the study and provide written consent, they enroll in the system. Participants will be asked to wear a smartwatch and play a serious game on a tablet under the supervision of a clinician. The system will collect data from the gameplay and sensors.

Study Overview

• Status: Completed
• Conditions: Developmental Articulation Disorder; Developmental Language Disorders; Speech Disorder; Neurodevelopmental Disorder (Diagnosis)
• Intervention / Treatment: Diagnostic test: TD Playing the Serious Game APSOU; Diagnostic test: Non-TD Playing Serious Game

Detailed Description

In recent years, integrating digital and mobile technologies in health and education has opened new possibilities for monitoring and assessment. This study will build on these advancements by implementing artificial intelligence (AI) and machine learning models that offer real-time decision-making and personalized feedback. By automating critical aspects of the diagnostic process, this system will help bridge the gap between traditional clinical expertise and cutting-edge technology, ultimately enhancing early intervention efforts.

The SmartSpeech study will be a clinical trial designed to create and validate an automated decision-making system to enhance the screening and early detection of developmental speech and language difficulties in young children. This study will target typically and non-typically (neurodevelopmentally atypical) developing children, primarily focusing on preschool and elementary school-aged participants. The system will employ intelligent computing models, biometric sensors, and early diagnostic indicators of speech and language deficiencies to provide clinicians with a robust tool for identifying potential developmental issues.

Two primary research questions drive the trial:

1. Do differences exist between typically developing and neurodevelopmentally atypical children when engaging with the serious game used in the study?
2. If such differences are identified, can the collected data be utilized to train an automated decision-making system that accurately distinguishes between typically and neurodevelopmentally atypical children based on speech and language deficiencies?

To assess speech and language developmental skills and their manifestations, the interdisciplinary team will design a serious game (SG) based on theories that measure speech and language and psychomotor, cognitive, psychoemotional, and hearing skills. Each domain will assess the performance of specific tasks within the SG activities for the participant.

To address the questions of this study, data from typically developing children will be compared with data from non-typically developing children, analyzing whether the AI system can synthesize multiple data points across assessment domains to generate a comprehensive diagnostic profile.

Participants will be recruited through an open call distributed via private and public health and educational establishments across Greece. The recruitment process will focus on parents of typically and non-typically developing children aged 4 to 12. Parents will be invited to attend informational meetings where the objectives, procedures, and ethical guidelines of the study will be thoroughly explained.

During these sessions, parents will receive comprehensive information about the purpose of the study, including the specific role of the serious game developed in the study. The nature of the game, the expected involvement of their children, and the data collection methods will be discussed in detail to ensure transparency. Parents will be asked to provide informed written consent for their children to participate in full compliance with the General Data Protection Regulation and ethical considerations. These steps are essential to ensure that the rights and privacy of the participants will be protected throughout the study. Parents will also provide their child's developmental history in an online questionnaire embedded in the system.

The study will include typically developing preschool and school-aged children and non-typically developing children. It will not include children with other medical conditions or on medications that could potentially influence the results.

After an informative meeting and written consent from parents, participants will wear a smartwatch and engage in a serious game (designed and developed by the research team). Children will play the SG under the direct supervision of clinicians. This game will be designed to assess various developmental domains, including speech, language, psychomotor skills, cognitive function, psychoemotional behavior, and hearing abilities. The game will capture speech and language data through gameplay interactions alongside sensor data from the wearable device and the tablet camera (eye tracking and heart rate variability).

This game-based approach will naturally and engagingly collect diverse data points, ensuring a comprehensive evaluation of the child's speech and language abilities.

The system will form its datasets for analysis to detect speech and language development, feeding data into the selected artificial intelligence algorithms. The study will measure communication, psychomotor, cognitive, and psychoemotional skills through designed activities. The responses during gameplay will be evaluated, creating the developmental profile of the participant.

This structured and ethical recruitment process will ensure complete adherence to data protection and ethical standards while maintaining active engagement from both parents and children. By incorporating face-to-face supervision and real-time monitoring during gameplay, the study will guarantee the quality and integrity of the data collected for analysis.

The SG, paired with sensors from the smartwatch, will collect real-time data, including:

• Game responses
• Speech responses
• Gaze data from the tablet camera
• Heart Rate Variability (captured through the smartwatch)

This multi-modal data will be processed using neural network/machine learning algorithms to classify speech and language abilities across different developmental stages. The use of these advanced AI models will enable the system to:

• Identify patterns of speech/language deficiencies.
• Differentiate between typically and non-typically developed children.
• Create individualized diagnostic profiles for each participant.

Finally, an evaluation and verification of the intelligent system will be performed.

• Study Type: Interventional
• Enrollment (Actual): 520
• Phase: Not Applicable

Contacts and Locations

Study Locations

• Greece
  • Ioannina, Greece, 45500
    • University of Ioannina

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Child
• Accepts Healthy Volunteers: Yes

Description

Inclusion Criteria:

• Typically developing preschoolers and school-aged children
• Children diagnosed with neurodevelopmental disorders
• Children should be native Greek speakers

Exclusion Criteria:

• Children with other medical conditions
• Children on medications that might affect the results of the study

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Screening
• Allocation: Randomized
• Interventional Model: Parallel Assignment
• Masking: None (Open Label)

Number of Arms

2

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
Experimental: Typically Developed Children PLAYING SERIOUS GAME; The arm describes the data collection process to design an intelligent system that contributes to implementing new innovative tools to assist clinicians and/or other professionals in the complicated screening and diagnostic procedures for communication deficiencies in children. The main objective of this arm is to feed data from collected responses of TD children to the backend of the SmartSpeech system. The data we collect are Game responses, Speech responses, Gaze data from the tablet camera, and Heart Rate Variability (captured through the smartwatch). The backend system will use the data for training, evaluating, and validating the intelligence system.	Diagnostic test: TD Playing the Serious Game APSOU; After parents participate in informative meetings, they proceed with written consent, SmartSpeech enrollment, and child developmental history. Then, the child may interact with the Serious Game. Typically developed (TD) children engage under the clinician's face-to-face supervision with the serious game APSOU, which assesses developmental domains, including speech, language, psychomotor, cognitive, psychoemotional, and hearing abilities. The average time for TD children to play the whole game is estimated to be 45 minutes.
Active Comparator: Non-Typically Developed Children PLAYING SERIOUS GAME; The arm describes the data collection process to design an intelligent system that contributes to implementing new innovative tools to assist clinicians and/or other professionals in the complicated screening and diagnostic procedures for communication deficiencies in children. The main objective of this arm is to feed data from collected responses of Non-TD children to the backend of the SmartSpeech system. The data we collect are Game responses, Speech responses, Gaze data from the tablet camera, and Heart Rate Variability (captured through the smartwatch). The backend system will use the data for training, evaluating, and validating the intelligence system.	Diagnostic test: Non-TD Playing Serious Game; After parents participate in informative meetings, they proceed with written consent, SmartSpeech enrollment, and child developmental history. Then the child may interact with the serious game. Non-typically developed (non-TD) children engage under the clinician's face-to-face supervision with the serious game Apsou, which assesses developmental domains, including speech, language, psychomotor, cognitive, psychoemotional, and hearing abilities. The average time for non-TD children to play the whole game is estimated to be over an hour.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Speech and language for TD children	Score on Serious Game responses in Speech and language activities (0 - 100) by TD children	10-20 minutes
Psychomotor for TD children	Score on Serious Game responses in Psychomotor activities (0 - 100) by TD children	10-20 minutes
Cognitive for TD children	Score on Serious Game responses in Cognitive activities (0 - 100) by TD children	10-20 minutes
Psychoemotional for TD children	Score on Serious Game responses in Psychoemotional activities (0 - 100) by TD children	10-20 minutes
Hearing for TD children	Score on Serious Game responses in Hearing activities (0 - 100) by TD children	10-20 minutes
Eye-tracking fixation duration for TD children	Fixation Duration is the time the gaze stays on a single point in milliseconds (during specific activities in the Serious Game)	1-10 minutes
Eye-tracking trajectory path for TD children	Trajectory Path is the overall movement path of the gaze in millimeters on screen (during specific activities in the Serious Game)	1-10 minutes
Heart rate variability for TD children	Heart rate variability during the game that is calculated by HRV=(HRe-HRs)/HRs, HRs: Heart Rate at the start of the measurement, HRe: Heart Rate at the end of the measurement. The duration between start - end of each activity measurement is a few seconds	1-10 minutes
Speech and language for non-TD children	Score on Serious Game responses in Speech and language activities (0 - 100) by non-TD children	10-20 minutes
Psychomotor for non-TD children	Score on Serious Game responses in Psychomotor activities (0 - 100) by non-TD children	10-20 minutes
Cognitive for non-TD children	Score on Serious Game responses in Cognitive activities (0 - 100) by non-TD children	10-20 minutes
Psychoemotional for non-TD children	Score on Serious Game responses in Psychoemotional activities (0 - 100) by non-TD children	10-20 minutes
Hearing for non-TD children	Score on Serious Game responses in Hearing activities (0 - 100) by non-TD children	10-20 minutes
Eye-tracking fixation duration for non-TD children	Fixation Duration is the time the gaze stays on a single point in milliseconds (during specific activities in the Serious Game)	1-10 minutes
Eye-tracking trajectory path for non-TD children	Trajectory Path is the overall movement path of the gaze in millimeters on screen (during specific activities in the Serious Game)	1-10 minutes
Heart rate variability for non-TD children	Heart rate variability during the game that is calculated by HRV=(HRe-HRs)/HRs, HRs: Heart Rate at the start of the measurement, HRe: Heart Rate at the end of the measurement. The duration between start - end of each activity measurement is a few seconds	1-10 minutes

Collaborators and Investigators

• Sponsor: University of Ioannina
• Collaborators: DOTSOFT SA - GR; TELEGLOBAL LP - GR; Region of Epirus (Regional Operational Programme 'Epirus 2014-2020') - GR; Ioannina University Hospital, GR
• Investigators: Study Chair: Eugenia I Toki, PhD, University of Ioannina, Greece

Study record dates

Study Major Dates

• Study Start (Actual): March 29, 2021
• Primary Completion (Actual): June 30, 2023
• Study Completion (Actual): November 25, 2023

Study Registration Dates

• First Submitted: September 26, 2024
• First Submitted That Met QC Criteria: October 7, 2024
• First Posted (Actual): October 9, 2024

Study Record Updates

• Last Update Posted (Actual): October 15, 2024
• Last Update Submitted That Met QC Criteria: October 8, 2024
• Last Verified: October 1, 2024

More Information

Terms related to this study

• Keywords: screening; Neural Networks; Machine Learning; developmental disorder; Serious Game; language disorder; neurodevelopmental disorder; neurodevelopmental screening; developmental speech and language disorders; speech disorder; developmental deficiencies; Early Diagnostic Speech and Language Deficiencies Indicators; Child Communication
• Additional Relevant MeSH Terms: Mental Disorders; Nervous System Diseases; Neurologic Manifestations; Neurobehavioral Manifestations; Communication Disorders; Neurodevelopmental Disorders; Language Disorders; Language Development Disorders; Speech Disorders; Articulation Disorders
• Other Study ID Numbers: HP1AB-28185 (MIS: 5033088); 18435/15.05.2020 (Registry Identifier: Research Ethics Committee, University of Ioannina)

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO
• IPD Plan Description: The participants of this study did not give written consent for their data to be shared publicly, so due to privacy restrictions and the sensitive nature of this research data sharing is not applicable

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No