Objective Recognition of Tinnitus Location Using Electroencephalography Connectivity Features

Abstract

Purpose: Tinnitus is a common but obscure auditory disease to be studied. This study will determine whether the connectivity features in electroencephalography (EEG) signals can be used as the biomarkers for an efficient and fast diagnosis method for chronic tinnitus.Methods: In this study, the resting-state EEG signals of tinnitus patients with different tinnitus locations were recorded. Four connectivity features [including the Phase-locking value (PLV), Phase lag index (PLI), Pearson correlation coefficient (PCC), and Transfer entropy (TE)] and two time-frequency domain features in the EEG signals were extracted, and four machine learning algorithms, included two support vector machine models (SVM), a multi-layer perception network (MLP) and a convolutional neural network (CNN), were used based on the selected features to classify different possible tinnitus sources.Results: Classification accuracy was highest when the SVM algorithm or the MLP algorithm was applied to the PCC feature sets, achieving final average classification accuracies of 99.42 or 99.1%, respectively. And based on the PLV feature, the classification result was also particularly good. And MLP ran the fastest, with an average computing time of only 4.2 s, which was more suitable than other methods when a real-time diagnosis was required.Conclusion: Connectivity features of the resting-state EEG signals could characterize the differentiation of tinnitus location. The connectivity features (PCC and PLV) were more suitable as the biomarkers for the objective diagnosing of tinnitus. And the results were helpful for clinicians in the initial diagnosis of tinnitus.

Use of Some Relevant Parameters for Primary Prediction of Brain Activity in Idiopathic Tinnitus Based on a Machine Learning Application

“…Much research has used machine-learning techniques in tinnitus phenotyping and classifying. They trained data extracted from positron emission tomography and voxel-based morphometry [Schecklmann et al, 2012], functional near-infrared spectroscopy (fNIRS) [Li et al, 2021], big data, and subjective questionnaires [Crowson et al, 2021;Schlee et al, 2021]. All these studies recommended using of machine learning for better understanding of tinnitus either based on objective or subjective data.…”

“…These findings can be appropriately implied, classified, and enriched using machine-learning models. Machine learning is an artificial intelligence method by which the neuroimaging biomarkers and subjective correlates of tinnitus can be implemented properly in computer software to develop a proposed model that can be used for diagnostic and prognostic aspects [Jordan and Mitchell, 2015;Allgaier et al, 2021;Li et al, 2021]. Such a model helps in decision support in the diagnosis and treatment.…”

Tinnitus and distress: an electroencephalography classification study

“…12 , 13 More recent works have implemented machine learning to study tinnitus with EEG and other tools such as near-infrared spectroscopy. 14 , 15 The findings of these studies indicate the potential to differentiate tinnitus patients from healthy controls based on connectivity features and sample entropy. 15 , 16 These preliminary results are promising, and our study expands on them by implementing intensive classification procedures that use a wider range of candidate features and sizable samples of subjects.…”

“… 14 , 15 The findings of these studies indicate the potential to differentiate tinnitus patients from healthy controls based on connectivity features and sample entropy. 15 , 16 These preliminary results are promising, and our study expands on them by implementing intensive classification procedures that use a wider range of candidate features and sizable samples of subjects. We also develop a classifier specific to tinnitus-induced distress, which we use to further classify tinnitus patients as having high or low distress levels.…”