Side-Aware Meta-Learning for Cross-Dataset Listener Diagnosis With Subjective Tinnitus

Abstract

With the development of digital technology, machine learning has paved the way for the next generation of tinnitus diagnoses. Although machine learning has been widely applied in EEG-based tinnitus analysis, most current models are dataset-specific. Each dataset may be limited to a specific range of symptoms, overall disease severity, and demographic attributes; further, dataset formats may differ, impacting model performance. This paper proposes a side-aware meta-learning for cross-dataset tinnitus diagnosis, which can effectively classify tinnitus in subjects of divergent ages and genders from different data collection processes. Owing to the superiority of meta-learning, our method does not rely on large-scale datasets like conventional deep learning models. Moreover, we design a subject-specific training process to assist the model in fitting the data pattern of different patients or healthy people. Our method achieves a high accuracy of 73.8% in the cross-dataset classification. We conduct an extensive analysis to show the effectiveness of side information of ears in enhancing model performance and side-aware metalearning in improving the quality of the learned features.

Deep learning-based electroencephalic diagnosis of tinnitus symptom

“…Due to many possible causes, such as abnormality in top-down or bottom-up processes, and different symptoms, such as hearing loss or noise trauma, there is currently no universally effective clinical method for tinnitus diagnosis ( Hall et al, 2016 ; Liu et al, 2022 ). At present, the diagnosis battery for tinnitus relies mainly on subjective assessments and self-reports, such as case history, audiometric tests, detailed tinnitus inquiry, tinnitus matching, and neuropsychological assessment ( Basile et al, 2013 ; Tang et al, 2019 ).…”

Cross-Subject Tinnitus Diagnosis Based on Multi-Band EEG Contrastive Representation Learning

“…ACC AUC Precision Recall F1-score ACC improvement v-SVM [37] 0.5817(0.025) 0.6123(0.018) 0.5933(0.020) 0.5968(0.027) 0.6007(0.026) 57.023% MLP [38] 0.6546(0.031) 0.6763(0.036) 0.6849(0.025) 0.6618(0.036) 0.6753(0.034) 39.536% EEGNet [39] 0.7366(0.067) 0.7427(0.071) 0.7582(0.073) 0.7285(0.064) 0.7423(0.083) 24.002% SiameseAE [14] 0.7972(0.042) 0.8018(0.048) 0.8241(0.052) 0.8068(0.051) 0.8193(0.048) 14.576% SMeta-SAE [40] methods can hardly handle complex EEG signals, Power Spectrum Density (PSD) of each electrode is calculated as the input, which is a common way of hard-encoding the EEG data in disease diagnosis. 2) MLP [38] stands for Multilayer Perceptron, which is a basic deep learning model and has strong nonlinear fitting ability.…”

“…However, the auto-encoder structure is too simple and does not take the spatial and frequency information of the EEG data into account, which limits the further improvement of the diagnosis performance. 5) SMeta-SAE [40] is an ABR-based cross-dataset tinnitus diagnosis method using SiameseAE as the backbone model. The introduction of meta-learning alleviates the problem of missing large-scale datasets.…”

Explainable AI Method for Tinnitus Diagnosis via Neighbor-Augmented Knowledge Graph and Traditional Chinese Medicine: Development and Validation Study

“…S n = G(f 20 (score(X,Y))) (7) where G denotes a frequency-counting method in which X and Y represent sets of patient nodes. f 20 () was used to obtain the top 20 patient syndromes based on the scores.…”

“…Previous studies have focused on the use of artificial intelligence (AI) to assist doctors in diagnosing tinnitus and improving diagnostic accuracy. Liu et al [ 7 ] proposed a meta-learning method based on lateral perception for cross–data set tinnitus diagnosis. Sun et al [ 8 ] used a support vector machine classifier to distinguish between patients with tinnitus and healthy individuals.…”