Mild Cognitive Impairment (MCI) is a distinct stage between normal age-related cognitive decline and dementia, defined by poor performance in one or more cognitive areas or domains. MCI affects attention, memory, executive cognitive function, language, and visuospatial abilities, and is positively correlated with both age and cardiovascular diseases, with 58% of stroke survivors suffering from cognitive impairment. Computerized Cognitive Training (CCT) is a cost-effective and easily accessible type of cognitive training that provides immediate feedback and makes the training interactive and motivating, enhancing long-term training feasibility. Thus, the aim of the thesis is the development of a novel solution based on a Skill for Amazon Alexa devices to support the Cognitive Training of people affected by MCI. The implemented cognitive training protocol follows the exercises of the MoCA Blind, a vocal version of the Montreal Cognitive Assessment (MoCA), the recommended cognitive screening tool for MCI. A solution was developed based on a three-tier architecture. The Data Tier stores information about the user and their training results in a cloud NoSQL Database using MongoDB. The Application Tier manages the interaction between the Alexa Skill and the MongoDB through a REST API built using the NestJS framework. The Presentation Tier handles the interaction between the user and the solution, handled through a Skill developed for Amazon Alexa devices. A preliminary test was conducted with 5 participants (1 male, 4 females), of median age 27 (27; 26,5). Volunteers were asked to try the Skill on their own Alexa devices for 7 days and were asked to complete the System Usability Scale Questionnaire and a semi-structured interview at the end of the experiment. Results were good, with a median SUS of 92,5, and overall positive results from the semi-structured interview. The proposed solution is found usable to propose the MoCA test in a direct and effective way. Further studies will include a larger population, together with involvement of psychologists and psychiatrists to evaluate the capabilities of the solution to reduce MCI in affected patients.
Il Mild Cognitive Impairment (MCI) è una fase distinta tra il normale declino cognitivo legato all'età e la demenza, definita da una scarsa prestazione in una o più aree cognitive. L’MCI influisce su attenzione, memoria, funzioni cognitive esecutive, linguaggio e abilità visivo-spaziali, ed è correlato positivamente sia con l'età che con le malattie cardiovascolari, con il 58% dei sopravvissuti a ictus che soffre di MCI. Il Computerized Cognitive Training (CCT) è un tipo di training cognitivo economico e facilmente accessibile che fornisce un feedback immediato e rende l’allenamento interattivo e motivante, incentivando il training a lungo termine. Pertanto, lo scopo della tesi è sviluppare una nuova soluzione basata su una Skill per dispositivi Amazon Alexa per supportare l'Allenamento Cognitivo delle persone affette da MCI. Il protocollo di allenamento cognitivo implementato segue gli esercizi del MoCA Blind, una versione vocale del Montreal Cognitive Assessment (MoCA), lo strumento di screening cognitivo raccomandato per l’MCI. La soluzione sviluppata si basa su un'architettura di tipo three-tier. Il Data Tier salva informazioni sugli utenti e i loro risultati di training in un cloud NoSQL Database usando MongoDB. L’Application Tier gestisce l'interazione tra l’Alexa Skill e MongoDB attraverso una REST API costruita utilizzando il framework NestJS. Il Presentation Tier coordina l'interazione tra l'utente e la piattaforma, gestita tramite una Skill per dispositivi Amazon Alexa. È stato condotto uno studio preliminare con 5 partecipanti (1 maschio, 4 femmine), di età mediana 27 (27; 26,5). Ai volontari è stato chiesto di provare la Skill sui propri dispositivi Alexa per 7 giorni e di completare il questionario System Usability Scale e un'intervista semi-strutturata alla fine dell'esperimento. I risultati sono stati buoni, con una mediana SUS di 92,5 e esiti globalmente positivi dall'intervista. La soluzione risulta usufruibile per proporre il test MoCA in modo diretto ed efficace. Studi futuri includeranno una popolazione più ampia, insieme al coinvolgimento di psicologi e psichiatri per valutare le capacità della soluzione di ridurre l’MCI nei pazienti affetti.
Development of a vocal interactive Amazon Alexa skill for cognitive training
Koteva, Lubov
2022/2023
Abstract
Mild Cognitive Impairment (MCI) is a distinct stage between normal age-related cognitive decline and dementia, defined by poor performance in one or more cognitive areas or domains. MCI affects attention, memory, executive cognitive function, language, and visuospatial abilities, and is positively correlated with both age and cardiovascular diseases, with 58% of stroke survivors suffering from cognitive impairment. Computerized Cognitive Training (CCT) is a cost-effective and easily accessible type of cognitive training that provides immediate feedback and makes the training interactive and motivating, enhancing long-term training feasibility. Thus, the aim of the thesis is the development of a novel solution based on a Skill for Amazon Alexa devices to support the Cognitive Training of people affected by MCI. The implemented cognitive training protocol follows the exercises of the MoCA Blind, a vocal version of the Montreal Cognitive Assessment (MoCA), the recommended cognitive screening tool for MCI. A solution was developed based on a three-tier architecture. The Data Tier stores information about the user and their training results in a cloud NoSQL Database using MongoDB. The Application Tier manages the interaction between the Alexa Skill and the MongoDB through a REST API built using the NestJS framework. The Presentation Tier handles the interaction between the user and the solution, handled through a Skill developed for Amazon Alexa devices. A preliminary test was conducted with 5 participants (1 male, 4 females), of median age 27 (27; 26,5). Volunteers were asked to try the Skill on their own Alexa devices for 7 days and were asked to complete the System Usability Scale Questionnaire and a semi-structured interview at the end of the experiment. Results were good, with a median SUS of 92,5, and overall positive results from the semi-structured interview. The proposed solution is found usable to propose the MoCA test in a direct and effective way. Further studies will include a larger population, together with involvement of psychologists and psychiatrists to evaluate the capabilities of the solution to reduce MCI in affected patients.| File | Dimensione | Formato | |
|---|---|---|---|
|
2024_04_Koteva.pdf
accessibile in internet per tutti
Descrizione: Testo della tesi
Dimensione
7.47 MB
Formato
Adobe PDF
|
7.47 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/218996