Paula Martins

European Portuguese MRI based speech production studies

Knowledge of the speech production mechanism is essential for the development of speech production models and theories. Magnetic resonance imaging delivers high quality images of soft tissues, has multiplanar capacity and allows for the visualization of the entire vocal tract. To our knowledge, there are no complete and systematic magnetic resonance imaging studies of European Portuguese production. In this study, a recently acquired magnetic resonance imaging database including almost all classes of European Portuguese sounds, excluding taps and trills, is presented and analyzed. Our work contemplated not only image acquisition but also the utilization of image processing techniques to allow the exploration of the entire database in a reasonable time. Contours extracted from 2D images, articulatory measures (2D) and area functions are explored and represent valuable information for articulatory synthesis and articulatory phonetics descriptions. Some European Portuguese distinctive characteristics, such as nasality are addressed in more detail. Results relative to oral vowels, nasal vowels and a comparison between both classes are presented. The more detailed information on tract configuration supports results obtained with other techniques, such as EMMA, and allows the comparison of European Portuguese and French nasal vowels articulation, with differences detected at pharyngeal cavity level and velum port opening quotient. A detailed characterization of the central vowels, particularly the [], is presented and compared with classical descriptions. Results for consonants point to the existence of a single positional dark allophone for [l], a more palato-alveolar place of articulation for [@?], a more anterior place of articulation for [@?] relative to [@?], and the use, by our speaker, of a palatal place of articulation for [k]. Some preliminary results concerning coarticulation are also reported. European Portuguese stops revealed less resistant to coarticulatory effects than fricatives. Among all the sounds studied, [@?] and [@?] present the highest resistance to coarticulation. These results follow the main key features found in other studies performed for different languages.

Real-Time MRI for portuguese

In this paper, we present a database of synchronized audio and Real-Time Magnetic Resonance Imaging (RT-MRI) in order to study dynamic aspects of the production of European Portuguese (EP) sounds. Currently, data have been acquired from one native speaker of European Portuguese. The speech corpus was primarily designed to investigate nasal vowels in a wide range of phonological contexts, but also includes examples of other EP sounds. The RT-MRI protocol developed for the acquisition of the data is detailed. Midsagittal and oblique images were acquired with a frame rate of 14 frames/s, resulting in a temporal resolution of 72 ms. Different image processing tools (automatic and semi-automatic) applied for inspection and analysis of the data are described. We demonstrate the potential of this database and processing techniques with some illustrative examples of Portuguese nasal vowels, taps, trills and laterals.

Segmentation and Analysis of Vocal Tract from MidSagittal Real-Time MRI

The articulatory description of European Portuguese (EP) requires the analysis of different anatomical structures (e.g. tongue dorsum and velum), and the study of dynamic aspects of speech production. The use of real-time magnetic resonance imaging (RT-MRI), with frame rates above 10 frames/s, provides adequate support for these studies and results in a large amount of images that need to be processed to extract relevant data to be analysed by linguists. To tackle the required data processing and analysis this article presents methods to perform segmentation of the vocal tract from midsagittal real-time MR image sequences and provide researchers with visualizations of the relevant extracted data. Examples are provided illustrating the analysis of dynamic aspects of EP nasal vowels.

Segmentation and analysis of the oral and nasal cavities from MR time sequences

The study of dynamic aspects of speech production in Portuguese is very important to characterize vowel nasalization. In this context, the analysis of velum movement remains a challenging task and only a few studies present articulatory descriptions of Portuguese nasal vowels. Advances in real-time MRI (magnetic resonance imaging) allow the acquisition of vocal tract images with reasonable spatial and temporal resolution to enable observation and quantification of articulatory movements. The resulting data consists of large image sequences and the structures of interest (e.g., oral cavity) have to be identified (segmented) throughout to enable analysis which can be a time consuming task. This article presents a segmentation tool for real-time MR image sequences of the oral and nasal cavities. The proposed tool has been implemented using MevisLab and provides features for the analysis of the resulting data.

Generalization and maintenance of treatment gains in primary progressive aphasia (PPA): a systematic review.

BACKGROUND Cognitive-linguistic treatments and interventions targeting communication have been developed within the context of primary progressive aphasia (PPA), however knowledge about the scope of generalization and maintenance of therapy gains considering PPA subtypes remains scarce and awaits systematic investigation. AIMS To analyse the effects of semantic therapy on generalization and maintenance of treatment outcomes in individuals with PPA, considering its different subtypes. METHODS & PROCEDURES Central, PubMed, Medline, Web of Knowledge and Scopus were used to retrieve articles of interest. A total of 25 non-randomized studies published between 2000 and 2016 met the eligibility criteria and therefore were included in this study. MAIN CONTRIBUTION This systematic review provides evidence-based information for clinical practice in PPA. Generalization and maintenance effects post-treatment for each PPA variant are analysed and discussed. Several factors are described as important to maximize the scope for generalization and maintenance of treatment gains. CONCLUSIONS & IMPLICATIONS Generalization is particularly hard to achieve in the semantic variant, as in the face of degraded semantic knowledge learning is rigid and context dependent. In contrast, non-fluent and logopenic variants offer better scope for generalization. Maintenance patterns do not seem to be influenced by PPA subtype, but rather by other factors such as continued practice, treatment length and frequency of sessions. In the future, clinicians should consider the PPA subtype when planning the treatment protocol.

Production and modeling of the european portuguese palatal lateral

In this study, an articulatory characterization of the palatal lateral is provided, using MRI images of the vocal tract acquired during the production of /L/by several speakers of European Portuguese. The production of this sound involves: a complete linguo-alveolopalatal closure; inward lateral compression of the tongue and a convex shape of the posterior tongue body, allowing airflow around the sides of the tongue; large cross-sectional areas in the upper pharyngeal and velar regions. The lengths and area functions derived from MRI are analysed and used to model the articulatory-acoustic relations involved in the production of /L/. The results obtained in the first simulations show that the vocal-tract model (VTAR) is reasonably able to estimate the frequencies of the first formants and zeros. The lateral channels combined with the supralingual cavity create pole-zero clusters around and above F3, in the frequency range of 2-4.5 kHz.

Quantitative Analysis of /l/ Production from RT-MRI: First Results

Lateral consonants are complex and variable sounds. Static MRI provides relevant information regarding /l/ geometry, but does not address dynamic properties. Real-time MRI is a well suited technique for dealing with temporal aspects. However, large amounts of data have to be processed to harness its full potential. The main goal of this paper is to extend a recently proposed quantitative framework to the analysis of real-time MRI data for European Portuguese /l/. Several vocal tract configurations of the alveolar consonant, acquired in different syllable positions and vocalic contexts, were compared. The quantitative framework revealed itself capable of dealing with the data for the /l/, allowing a systematic analysis of the multiple realisations. The results regarding syllable position effects and coarticulation of /l/ with adjacent vowels are in line with previous findings.

Functional Mapping of Inner Speech Areas: A Preliminary Study with Portuguese Speakers

Inner speech can be defined as the act of talking silently with ourselves. Several studies aimed to understand how this process is related to speech organization and language. Despite the advances, some results are still contradictory. Importantly, language dependency is scarcely studied. For this first study of inner speech for Portuguese native speakers using fMRI, we selected a confrontation naming task, consisting of 40 black and white line drawings. Five healthy participants were instructed to name in inner and in overt speech the visually presented image. fMRI data analysis considering the proposed inner speech paradigm identified several brain areas such as the left inferior frontal gyrus, including Broca’s area, supplementary motor area, precentral gyrus and left middle temporal gyrus including Wernicke’s area. Our results also show more pronounced bilateral activations during the overt speech task when compared to inner speech, suggesting that inner and overt speech activate similar areas but stronger activation can be found in the later. However, this difference stems in particular from significant activation differences in the right pre-central gyrus and middle temporal gyrus.

Inner Speech in Portuguese: Acquisition Methods, Database and First Results

In this paper, we present a database developed for studying inner speech brain related areas using functional Magnetic Resonance Imaging (fMRI) in the context of the European Portuguese. First, we addressed the type of stimuli used in inner speech studies. In this sense, considering a preliminary study using a picture naming task, we defined a corpus. The corpus was designed based on cardinal vowels, syllable, disyllabic words and sentences with structure S(ubject)V(erb)O(bject) which were balanced in syllable number (six to ten). All the words used are common words from the Portuguese lexicon and possible ambiguities were excluded. Currently, the dataset includes data from twenty healthy participants native Portuguese speakers. Preliminary, exploratory analysis on the data allowed us to identify the most relevant areas part of the inner speech network, that include inferior frontal gyrus (including Broca’s area), supplementary motor area and precentral gyrus. Ultimately, the better understanding of the inner speech mechanisms will pave way to the development of novel intervention strategies in linguistic disorders.

Narrative review of the psychometric properties of language tests used in anomia treatment for primary progressive aphasia (PPA)

ABSTRACT Background: Primary progressive aphasia (PPA) is a neurodegenerative dementia in which language decline is the first and most prominent symptom. Among several interventions for PPA, language rehabilitation has been the most frequently used. Aims: This narrative review aimed to evaluate existing standardised language tests used in the assessment of PPA, in order to determine whether they are appropriate and psychometrically adequate to detect change over time in the treatment of anomia. Main Contribution: The present findings highlight the scarcity of psychometrically robust instruments used to measure therapy-induced gains in PPA. Additionally, most of these instruments were not validated for use with the PPA population, which consequently might bias the results. There is a need for population-based norms for existing instruments. Conclusions: The accurate assessment of lexico-semantic deficits in PPA should rely on objective, reliable, valid, and responsive outcome measures. Psychometric studies are needed to evaluate and eventually improve the quality of language tests used in clinical practice.