Objective measurement of tinnitus using functional near-infrared spectroscopy and machine learning

Abstract

Chronic tinnitus is a debilitating condition which affects 10–20% of adults and can severely impact their quality of life. Currently there is no objective measure of tinnitus that can be used clinically. Clinical assessment of the condition uses subjective feedback from individuals which is not always reliable. We investigated the sensitivity of functional near-infrared spectroscopy (fNIRS) to differentiate individuals with and without tinnitus and to identify fNIRS features associated with subjective ratings of tinnitus severity. We recorded fNIRS signals in the resting state and in response to auditory or visual stimuli from 25 individuals with chronic tinnitus and 21 controls matched for age and hearing loss. Severity of tinnitus was rated using the Tinnitus Handicap Inventory and subjective ratings of tinnitus loudness and annoyance were measured on a visual analogue scale. Following statistical group comparisons, machine learning methods including feature extraction and classification were applied to the fNIRS features to classify patients with tinnitus and controls and differentiate tinnitus at different severity levels. Resting state measures of connectivity between temporal regions and frontal and occipital regions were significantly higher in patients with tinnitus compared to controls. In the tinnitus group, temporal-occipital connectivity showed a significant increase with subject ratings of loudness. Also in this group, both visual and auditory evoked responses were significantly reduced in the visual and auditory regions of interest respectively. Naïve Bayes classifiers were able to classify patients with tinnitus from controls with an accuracy of 78.3%. An accuracy of 87.32% was achieved using Neural Networks to differentiate patients with slight/ mild versus moderate/ severe tinnitus. Our findings show the feasibility of using fNIRS and machine learning to develop an objective measure of tinnitus. Such a measure would greatly benefit clinicians and patients by providing a tool to objectively assess new treatments and patients’ treatment progress.

Analysis methods for measuring passive auditory fNIRS responses generated by a block-design paradigm

“…[5][6][7][8][9][10][11][12][13][14] Research questions relating to the development of auditory cortical function 15 and cortical reorganization following impaired sensory input and subsequent rehabilitation 16,17 have been investigated using fNIRS, as have outcomes related to cochlear implantation 5 and auditory pathologies such as tinnitus. 18,19 Of particular clinical relevance within the auditory research community is the field of objective measures. Objective measures utilize passive experimental designs-in which the participant is not required to perform any tasks throughout the measurement-and are routinely utilized to evaluate hearing performance in populations who are unable to provide reliable behavioral responses.…”

A State-of-Art Review of Digital Technologies for the Next Generation of Tinnitus Therapeutics

“…ML is a field of study that applies the principles of computer and mathematical science and statistics to create computational models, which are used for future predictions (based on past data) and identifying patterns in data ( 159 ). The use of ML has increased in healthcare applications and has been applied to behavioral, EEG, functional Magnetic resonance imaging fMRI, and Functional near-infrared spectroscopy (fNIRS) data ( 163 – 165 ). The learning algorithms in ML can be divided into two main groups: supervised and unsupervised ( 166 ).…”

“…The results suggested a combination of 13 cortical/subcortical brain regions had the highest classification accuracy for effectively differentiating patients with tinnitus from healthy subjects ( 163 ). In addition to EEG data, Shoushtarian et al ( 165 ) collected fNIRS data to differentiate tinnitus patients from control participants and to identify fNIRS features associated with tinnitus severity. The Naïve Bayes classifiers (a mathematical formula for determining probability of an outcome occurring, based on a previous outcomes) were used to classify patients with tinnitus from controls.…”

“…The Naïve Bayes classifiers (a mathematical formula for determining probability of an outcome occurring, based on a previous outcomes) were used to classify patients with tinnitus from controls. An accuracy of 87.32% was obtained to distinguish patients with slight/ mild vs. moderate/ severe tinnitus ( 165 ). These findings show the feasibility of using fNIRS and ML to develop an objective measure of tinnitus that might enable clinicians to provide new treatment plans.…”

Handling of uncertainty in medical data using machine learning and probability theory techniques: a review of 30 years (1991–2020)

“…Although perception tasks such as object detection need human intelligence, a subsequent task that needs inference and reasoning needs more human intelligence. Nowadays, new algorithms like deep learning could extremely improve perception tasks (Alizadehsani et al, 2021;Asgharnezhad et al, 2020;Ghassemi et al, 2019Ghassemi et al, , 2020Khodatars et al, 2020;Mohammadpoor et al, 2016;Sharifrazi et al, 2020;Shoeibi et al, 2020a, b;Shoeibi, 2021;Shoushtarian et al, 2020). However, for higher-level inference, probabilistic graphical models are more powerful than other algorithms.…”