Introduction and Purpose: Dysarthria affects 80% of people with Amyotrophic Lateral Sclerosis (ALS), significantly impairing communication. In response to the growing need to monitor disease progression and intervene promptly, this research aims to identify technological tools and outcomes for measuring acoustic parameters related to speech in dysarthric subjects. The analysis focuses on the strengths and limitations of the available tools, evaluating their applicability in hospital, outpatient, and home settings, while identifying the clinical and technical implications for speech therapy practice. Materials and Methods: The research was conducted independently by two authors, who initially reviewed the titles and abstracts of the selected studies, followed by a full-text analysis to verify adherence to inclusion and exclusion criteria. The discrepancies were resolved through a discussion among the individuals involved in the research (including the authors and the supervisors). Articles published between 2020 and 2024 were reviewed from the SCOPUS, Medline, and CINAHL databases. Data extraction focused on clinical information. Specifically, detailed information was collected on the characteristics of participants with ALS and, where applicable, those in the control group. Additionally, the inclusion and exclusion criteria for each study, as well as the screening and clinical assessment tests, were recorded. Furthermore, the inclusion and exclusion criteria of each study, as well as screening and clinical assessment tests, were noted. The specific features of each study were analyzed: the linguistic domains investigated, tasks presented, and the acoustic and kinematic signals and databases. Finally, the clinical and technical results were discussed and argued through thematic analysis. Results: After the selection process, 43 articles were included: 21 focused on acoustics, 11 on kinematics, and 11 on both acoustic and kinematic signals. The articles relevant to this study pertain to acoustic parameters (32 articles). Recent studies on artificial intelligence in the assessment of ALS-related dysarthria have introduced innovative tools for monitoring and early diagnosis, such as decision tree models and deep learning systems that analyze audio and video signals. Significant research has validated diagnostic tools, such as the Winterlight pipeline, which classifies patients and monitors disease progression, as well as innovative methods like vocal tremor tracking and a multimodal framework. From the analysis of the articles, it emerged that dysarthria is characterized by deficits in the speed and accuracy of articulation, affecting the intelligibility of speech, particularly in the production of high and mid vowels. The diadochokinesis (DDK) task proved crucial for diagnosis, allowing for the distinction between dysarthria and verbal apraxia, while sustained vowels revealed specific difficulties in production.However, many studies are limited by small sample sizes, making it difficult to generalize the results. Conclusions: This research highlights the importance of developing advanced technologies and the opportunity to adopt existing methodologies for monitoring dysarthric profiles in ALS patients by measuring acoustic parameters related to speech articulation. However, the studies are limited by small and non-diversified samples, requiring further experimental investigation. A broader and more inclusive approach could improve the level of evidence, leading to crucial interventions for communication abilities and, consequently, the quality of life for individuals affected by ALS. Keywords: ALS, Dysarthria, acoustic signals, intelligibility, diadochokinesis, monitoring
Introduzione e scopo: La disartria colpisce l'80% delle persone affette da Sclerosi Laterale Amiotrofica (SLA), compromettendo significativamente la comunicazione. In risposta alla crescente necessità di monitorare il deterioramento delle condizioni e intervenire tempestivamente, questa ricerca mira a identificare strumenti tecnologici e outcome per misurare i parametri acustici inerenti allo speech in soggetti disartrici. L'analisi si concentra sui punti di forza e le limitazioni degli strumenti disponibili, valutando la loro applicabilità in setting ospedaliero-ambulatoriale e domiciliare, e individuando le implicazioni cliniche e tecniche per la pratica logopedica. Materiali e metodi: La ricerca è stata condotta in modo indipendente da due autrici, che hanno inizialmente esaminato titoli e abstract degli studi selezionati, seguiti dall'analisi dei testi completi per verificare il rispetto dei criteri di inclusione ed esclusione. Le discrepanze sono state risolte con un confronto tra i soggetti coinvolti nella ricerca (incluse le autrici e le relatrici). Sono stati esaminati articoli pubblicati tra il 2020 e il 2024 nelle banche dati SCOPUS, Medline e CINHAL. L'estrazione dei dati si è concentrata sulle informazioni cliniche. Nello specifico sono state raccolte informazioni dettagliate sulle caratteristiche dei partecipanti affetti da SLA e dei soggetti del gruppo di controllo, ove presente. Inoltre, sono stati rilevati i criteri di inclusione ed esclusione di ogni studio, i test di screening e di inquadramento clinico. Sono state analizzate le caratteristiche specifiche di ogni studio: i domini linguistici indagati; i task presentati; i segnali e i database acustici e cinematici. Infine, sono stati discussi i risultati clinici e tecnici rilevati, ed argomentati attraverso un’analisi tematica. Risultati: Al termine del processo di selezione, sono stati inclusi 43 articoli: 21 acustici, 11 cinematici e 11 con segnali sia acustici che cinematici. Gli articoli di interesse della presente trattazione riguardano i parametri acustici (32 articoli). È stato rilevato come recenti studi sull'intelligenza artificiale nella valutazione della disartria legata alla SLA introducono strumenti innovativi per il monitoraggio e la diagnosi precoce, come modelli di albero decisionale e sistemi basati su deep learning che analizzano segnali audio e video. Ricerche significative hanno validato strumenti diagnostici, come la pipeline di Winterlight, che classifica i pazienti e monitora l'evoluzione della malattia, e metodi innovativi come il tracciatore di tremore vocale e un framework multimodale. Dall'analisi degli articoli è emerso che la disartria si caratterizza per deficit nella velocità e nell'accuratezza articolatoria, influenzando l'intelligibilità del linguaggio, in particolare nella produzione delle vocali alte e medie. Il compito di diadococinesia (DDK) è risultato cruciale per la diagnosi, permettendo di distinguere tra disartria e aprassia verbale, mentre le vocali sostenute hanno evidenziato difficoltà specifiche nella produzione. Tuttavia, molti studi presentano la limitazione di campioni ridotti, rendendo difficile generalizzare i risultati. Conclusioni: La ricerca evidenzia l'importanza di sviluppare tecnologie avanzate e l’opportunità di adottare metodologie esistenti nel monitoraggio del quadro disartrico dei soggetti affetti da SLA, attraverso la misurazione di parametri acustici relativi all’articolazione dello speech. Tuttavia, gli studi presentano la limitazione di campioni ridotti e non diversificati, richiedendo ulteriori approfondimenti sperimentali. Un approccio più ampio e inclusivo potrebbe migliorare il livello delle evidenze, a favore di un intervento cruciale sulla capacità comunicativa e, quindi, sulla qualità di vita dei soggetti affetti da SLA. Parole chiave: SLA, Disartria, segnali acustici, intelligibilità, diadococinesia, monitoraggio
Valutazione strumentale dell'intelligibilità e dei parametri acustici nella disartria secondaria a Sclerosi Laterale Amiotrofica
PERTICARINI, LAURA
2023/2024
Abstract
Introduction and Purpose: Dysarthria affects 80% of people with Amyotrophic Lateral Sclerosis (ALS), significantly impairing communication. In response to the growing need to monitor disease progression and intervene promptly, this research aims to identify technological tools and outcomes for measuring acoustic parameters related to speech in dysarthric subjects. The analysis focuses on the strengths and limitations of the available tools, evaluating their applicability in hospital, outpatient, and home settings, while identifying the clinical and technical implications for speech therapy practice. Materials and Methods: The research was conducted independently by two authors, who initially reviewed the titles and abstracts of the selected studies, followed by a full-text analysis to verify adherence to inclusion and exclusion criteria. The discrepancies were resolved through a discussion among the individuals involved in the research (including the authors and the supervisors). Articles published between 2020 and 2024 were reviewed from the SCOPUS, Medline, and CINAHL databases. Data extraction focused on clinical information. Specifically, detailed information was collected on the characteristics of participants with ALS and, where applicable, those in the control group. Additionally, the inclusion and exclusion criteria for each study, as well as the screening and clinical assessment tests, were recorded. Furthermore, the inclusion and exclusion criteria of each study, as well as screening and clinical assessment tests, were noted. The specific features of each study were analyzed: the linguistic domains investigated, tasks presented, and the acoustic and kinematic signals and databases. Finally, the clinical and technical results were discussed and argued through thematic analysis. Results: After the selection process, 43 articles were included: 21 focused on acoustics, 11 on kinematics, and 11 on both acoustic and kinematic signals. The articles relevant to this study pertain to acoustic parameters (32 articles). Recent studies on artificial intelligence in the assessment of ALS-related dysarthria have introduced innovative tools for monitoring and early diagnosis, such as decision tree models and deep learning systems that analyze audio and video signals. Significant research has validated diagnostic tools, such as the Winterlight pipeline, which classifies patients and monitors disease progression, as well as innovative methods like vocal tremor tracking and a multimodal framework. From the analysis of the articles, it emerged that dysarthria is characterized by deficits in the speed and accuracy of articulation, affecting the intelligibility of speech, particularly in the production of high and mid vowels. The diadochokinesis (DDK) task proved crucial for diagnosis, allowing for the distinction between dysarthria and verbal apraxia, while sustained vowels revealed specific difficulties in production.However, many studies are limited by small sample sizes, making it difficult to generalize the results. Conclusions: This research highlights the importance of developing advanced technologies and the opportunity to adopt existing methodologies for monitoring dysarthric profiles in ALS patients by measuring acoustic parameters related to speech articulation. However, the studies are limited by small and non-diversified samples, requiring further experimental investigation. A broader and more inclusive approach could improve the level of evidence, leading to crucial interventions for communication abilities and, consequently, the quality of life for individuals affected by ALS. Keywords: ALS, Dysarthria, acoustic signals, intelligibility, diadochokinesis, monitoring| File | Dimensione | Formato | |
|---|---|---|---|
|
TESI_pdfA.pdf
embargo fino al 08/11/2027
Dimensione
801.4 kB
Formato
Adobe PDF
|
801.4 kB | Adobe PDF |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.12075/19656