RESUMO
Objective:Subjective tinnitus is an auditory phantom perceptual disorder without an objective biomarker. Fast and efficient diagnostic tools will advance clinical practice by detecting or confirming the condition, tracking change in severity, and monitoring treatment response. Motivated by evidence of subtle anatomical, morphological, or functional information in magnetic resonance images of the brain, we examine data-driven machine learning methods for joint tinnitus classification (tinnitus or no tinnitus) and tinnitus severity prediction.Approach:We propose a deep multi-task multimodal framework for tinnitus classification and severity prediction using structural MRI (sMRI) data. To leverage complementary information multimodal neuroimaging data, we integrate two modalities of three-dimensional sMRI-T1 weighted (T1w) and T2 weighted (T2w) images. To explore the key components in the MR images that drove task performance, we segment both T1w and T2w images into three different components-cerebrospinal fluid, grey matter and white matter, and evaluate performance of each segmented image.Main results:Results demonstrate that our multimodal framework capitalizes on the information across both modalities (T1w and T2w) for the joint task of tinnitus classification and severity prediction.Significance:Our model outperforms existing learning-based and conventional methods in terms of accuracy, sensitivity, specificity, and negative predictive value.
Assuntos
Zumbido , Humanos , Zumbido/diagnóstico por imagem , Imageamento por Ressonância Magnética/métodos , Encéfalo/diagnóstico por imagem , Neuroimagem , Substância CinzentaRESUMO
In order to examine the effect of inhibition on processing auditory temporal information, responses of single neurons in the inferior colliculus of the chinchilla to sinusoidally amplitude-modulated (SAM) tones alone and the presence of a steady-state tone were obtained. The carrier frequency of the SAM tone was either the characteristic frequency (CF) or a frequency in the inhibitory response area of a studied neuron. When the carrier frequency was set to the neuron's CF, neurons responded in synchrony to the SAM-tone envelope, as expected. When the carrier frequency was set to a frequency at which pure tones produced inhibition, SAM tones elicited little or no response, also as expected. However, when the same SAM tone was paired with a pure tone whose frequency was set to the neuron's CF, responses synchronized to the SAM tone envelope were obtained. These modulated responses were typically one-half cycle out-of-phase with the response to the SAM tone at CF, suggesting that they arose from cyclic inhibition and release from inhibition by the SAM tone. The results demonstrate that the representation of temporal information by inferior colliculus neurons is influenced by temporally-patterned inhibition arising from locations remote from CF.