Showing papers by "Yaniv Zigel published in 2019"

Speaker diarization during noisy clinical diagnoses of autism

Abstract: Autism Spectrum Disorder (ASD) is characterized by difficulties in social communication, social interactions and repetitive behaviors. Some of these difficulties are apparent in the speech characteristics of ASD children who are verbal. Developing algorithms that can extract and quantify speech features that are unique to ASD children is, therefore, extremely valuable for assessing the initial state of each child and their development over time. An important component of such algorithms is speaker diarization in the noisy clinical environments where ASD children are diagnosed. Here we present a Gaussian Mixture Model (GMM) approach for speaker diarization that was applied to 34 recordings from clinical assessments using the Autism Diagnostic Observation Schedule (ADOS). We used mel-frequency cepstral coefficients (MFCC) and pitch based features to classify segments containing speech of the child, therapist, parent, movement noises (chair, toys, etc.) and simultaneous speech. We achieved an accuracy of 89% in identifying segments with children’s speech and an accuracy of 74.5% in identifying children’s and therapists’ speech segments. These accuracy rates are similar to the diarization accuracy rates reported by previous similar studies, thereby demonstrating a promising route for the automated assessment of speech in children with ASD.

Automatic Audio-Based Classification of Patient Inhaler Use: A Pharmacy Based Study

Abstract: Chronic respiratory diseases may be controlled through the delivery of medication to the airways and lungs using an inhaler. However, adherence to correct inhaler technique is poor, which impedes patients from receiving maximum clinical benefit from their medication. In this study, the Inhaler Compliance Assessment device was employed to record audio of patients using a Diskus dry powder inhaler. An algorithm that classifies inhaler sounds (blister, inhalation, interference) was developed to automatically assess patient adherence from these inhaler audio recordings. The presented algorithm employed audio-based signal processing methods and statistical modeling in the form of quadratic discriminant analysis (QDA). A total of 350 audio recordings were obtained from 70 patients. The acquired audio dataset was split evenly for training and testing. A total accuracy of 85.35% was obtained (testing dataset) for this 3-class classification system. A sensitivity of 89.22% and 70% was obtained for inhalation and blister detection respectively. This approach may have significant clinical impact by providing healthcare professionals with an efficient, objective method of monitoring patient adherence to inhaler treatment.