The Role of Data Analytics in the Assessment of Pathological Speech—A Critical Appraisal

Abstract

Pathological voice characterization has received increasing attention over the last 20 years. Hundreds of studies have been published showing inventive approaches with very promising findings. Nevertheless, methodological issues might hamper performance assessment trustworthiness. This study reviews some critical aspects regarding data collection and processing, machine learning-oriented methods, and grounding analytical approaches, with a view to embedding developed clinical decision support tools into the diagnosis decision-making process. A set of 26 relevant studies published since 2010 was selected through critical selection criteria and evaluated. The model-driven (MD) or data-driven (DD) character of the selected approaches is deeply examined considering novelty, originality, statistical robustness, trustworthiness, and clinical relevance. It has been found that before 2020 most of the works examined were more aligned with MD approaches, whereas over the last two years a balanced proportion of DD and MD-based studies was found. A total of 15 studies presented MD characters, whereas seven were mainly DD-oriented, and four shared both profiles. Fifteen studies showed exploratory or prospective advanced statistical analysis. Eighteen included some statistical validation to avail claims. Twenty-two reported original work, whereas the remaining four were systematic reviews of others’ work. Clinical relevance and acceptability by voice specialists were found in 14 out of the 26 works commented on. Methodological issues such as detection and classification performance, training and generalization capability, explainability, preservation of semantic load, clinical acceptance, robustness, and development expenses have been identified as major issues in applying machine learning to clinical support systems. Other important aspects to be taken into consideration are trustworthiness, gender-balance issues, and statistical relevance.

Robust and language-independent acoustic features in Parkinson's disease

“…Despite this impairment, which can be mainly ascribable to the heterogeneity of the dataset, the result confirms the feasibility of a classification algorithm trained on a unified dataset. Indeed, it is well known that one of the main problems encountered in the design of automatic tools for vocal pathology assessment concerns the database size, which can lead to feature selection and classification results excessively fitted on the population at hand ( 38 ). Moreover, although the use of highly homogeneous corpora would achieve better results, the same conditions can be hardly repeatable, thus limiting the actual usefulness in real-world scenarios.…”

Exploring the Impact of Fine-Tuning the Wav2vec2 Model in Database-Independent Detection of Dysarthric Speech

“…DL-based techniques for dysarthria detection include mapping from handcrafted acoustic features to output labels, as well as modern end-to-end systems in which the raw speech signal or time-frequency spectrogram is directly used by a DL model to compute the output labels. However, even though the modern end-to-end DL techniques have shown significant progress, they can still be criticized for the following two major issues: 1) large amounts of speech data are needed in the system training [36], and 2) interpretability of results provided by DL approaches is difficult and therefore the clinical relevance of the technology might be questioned by specialists [37].…”

Artificial Intelligence-Based Voice Assessment of Patients with Parkinson’s Disease Off and On Treatment: Machine vs. Deep-Learning Comparison

“…A recent work [ 37 ] reports a critical review of pathological voice characterization approaches, evidencing the methodological issues potentially hampering performance assessment trustworthiness, including the database dimension and a stratified corpora (either among classes or genders). Controversy over which specific Artificial Intelligence (AI) approach to employ, namely Machine Learning (ML) vs.…”