Background: Alzheimer’s disease (AD) has paramount importance due to its rising prevalence, the impact on the patient and society, and the related healthcare costs. However, current diagnostic techniques are not designed for frequent mass screening, delaying therapeutic intervention and worsening prognoses. To be able to detect AD at an early stage, ideally at a pre-clinical stage, speech analysis emerges as a simple low-cost non-invasive procedure. Objectives: In this work it is our objective to do a systematic review about speech-based detection and classification of Alzheimer’s Disease with the purpose of identifying the most effective algorithms and best practices. Methods: A systematic literature search was performed from Jan 2015 up to May 2020 using ScienceDirect, PubMed and DBLP. Articles were screened by title, abstract and full text as needed. A manual complementary search among the references of the included papers was also performed. Inclusion criteria and search strategies were defined a priori. Results: We were able: to identify the main resources that can support the development of decision support systems for AD, to list speech features that are correlated with the linguistic and acoustic footprint of the disease, to recognize the data models that can provide robust results and to observe the performance indicators that were reported. Discussion: A computational system with the adequate elements combination, based on the identified best-practices, can point to a whole new diagnostic approach, leading to better insights about AD symptoms and its disease patterns, creating conditions to promote a longer life span as well as an improvement in patient quality of life. The clinically relevant results that were identified can be used to establish a reference system and help to define research guidelines for future developments.
Background: Amyotrophic lateral sclerosis (ALS) is a fatal progressive motor neuron disease. People with ALS demonstrate various speech problems. Summary: We aim to provide an overview of studies concerning the diagnosis of ALS based on the analysis of voice samples. The main focus is on the feasibility of the use of voice and speech assessment as an effective method to diagnose the disease, either in clinical or pre-clinical conditions, and to monitor the disease progression. Specifically, we aim to examine current knowledge on: (a) voice parameters and the data models that can, most effectively, provide robust results; (b) the feasibility of a semi-automatic or automatic diagnosis and outcomes; and (c) the factors that can improve or restrict the use of such systems in a real-world context. Key Messages: The studies already carried out on the possibility of diagnosis of ALS using the voice signal are still sparse but all point to the importance, feasibility and simplicity of this approach. Most co-horts are small which limits the statistical relevance and makes it difficult to infer broader conclusions. The set of features used, although diverse, is quite circumscribed. ALS is difficult to diagnose early because it may mimic several other neurological diseases. Promising results were found for the automatic detection of ALS from speech samples and this can be a feasible process even in pre-symptomatic stages. Improved guidelines must be set in order to establish a robust decision model.
In the last few years, the number of systems and devices that use voice based interaction has grown significantly. For a continued use of these systems, the interface must be reliable and pleasant in order to provide an optimal user experience. However there are currently very few studies that try to evaluate how pleasant is a voice from a perceptual point of view when the final application is a speech based interface. In this paper we present an objective definition for voice pleasantness based on the composition of a representative feature subset and a new automatic voice pleasantness classification and intensity estimation system. Our study is based on a database composed by European Portuguese female voices but the methodology can be extended to male voices or to other languages. In the objective performance evaluation the system achieved a 9.1% error rate for voice pleasantness classification and a 15.7% error rate for voice pleasantness intensity estimation.
Diabetic peripheral neuropathy is a major complication of diabetes mellitus, and it is the leading cause of foot ulceration and amputations. The Semmes–Weinstein monofilament examination (SWME) is a widely used, low-cost, evidence-based tool for predicting the prognosis of diabetic foot patients. The examination can be quick, but due to the high prevalence of the disease, many healthcare professionals can be assigned to this task several days per month. In an ongoing project, it is our objective to minimize the intervention of humans in the SWME by using an automated testing system relying on computer vision. In this paper we present the project’s first part, constituting a system for automatically identifying the SWME testing sites from digital images. For this, we have created a database of plantar images and developed a segmentation system, based on image processing and deep learning—both of which are novelties. From the 9 testing sites, the system was able to correctly identify most 8 in more than 80% of the images, and 3 of the testing sites were correctly identified in more than 97.8% of the images.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.