Neuroscience and Biobehavioral Reviews 107 (2019) 104–114

Contents lists available at ScienceDirect

Neuroscience and Biobehavioral Reviews

journal homepage: www.elsevier.com/locate/neubiorev

Review article

Neural architectures of music – Insights from acquired amusia☆ T

a,b,⁎ b c d
Aleksi J. Sihvonen , Teppo Särkämö , Antoni Rodríguez-Fornells , Pablo Ripollés ,
Thomas F. Müntee, Seppo Soinilaf
a
Department of Neurosciences, University of Helsinki, Finland
b
Cognitive Brain Research Unit, Department of Psychology and Logopedics, University of Helsinki, Finland
c
Department of Cognition, University of Barcelona, Cognition & Brain Plasticity Unit, Bellvitge Biomedical Research Institute (IDIBELL), Institució Catalana de recerca i
Estudis Avançats (ICREA), Barcelona, Spain
d
Department of Psychology, New York University and Music and Audio Research Laboratory, New York University, USA
e
Department of Neurology and Institute of Psychology II, University of Lübeck, Germany
f
Division of Clinical Neurosciences, Turku University Hospital, Department of Neurology, University of Turku, Finland

A R T I C LE I N FO A B S T R A C T

Keywords: The ability to perceive and produce music is a quintessential element of human life, present in all known cul-
Acquired amusia tures. Modern functional neuroimaging has revealed that music listening activates a large-scale bilateral network
Music of cortical and subcortical regions in the healthy brain. Even the most accurate structural studies do not reveal
Stroke which brain areas are critical and causally linked to music processing. Such questions may be answered by
Lesions
analysing the effects of focal brain lesions in patients´ ability to perceive music. In this sense, acquired amusia
after stroke provides a unique opportunity to investigate the neural architectures crucial for normal music
processing. Based on the first large-scale longitudinal studies on stroke-induced amusia using modern multi-
modal magnetic resonance imaging (MRI) techniques, such as advanced lesion-symptom mapping, grey and
white matter morphometry, tractography and functional connectivity, we discuss neural structures critical for
music processing, consider music processing in light of the dual-stream model in the right hemisphere, and
propose a neural model for acquired amusia.

1. Introduction 2004; Adolphs, 2016).

A specific neurological deficit known as amusia (the severe im-
The ability to perceive, enjoy, and produce music is a core element pairment of music perception and/or production not caused by general
of human civilization, present in all cultures throughout history. cognitive, perceptual, or motor disability) was first reported at the end
Attempts to unravel the neural structures underlying music processing of the 19th century (Graziano and Johnson, 2015). There are essentially
have been booming during the past two decades and studies have re- two types of amusia: congenital amusia (CA), which is a neurodeve-
vealed a large-scale music processing network comprising bilateral lopmental disorder, and acquired amusia (AA), which is caused by
temporal, frontal, parietal, and subcortical regions (for recent meta- brain damage resulting from a neurological disorder. While both CA
analyses and reviews, see Zatorre and Salimpoor, 2013; Koelsch, 2014; and AA share the core symptoms of amusia, their neural mechanisms
Gordon et al., 2018; Freitas et al., 2018). While distinct brain areas are likely to differ: CA is a developmental deficit in acquiring musical
involved in music processing have been identified in healthy subjects, syntax and tonal representations that hampers the early development of
the evidence derived from functional neuroimaging studies is largely the music processing network and leads to lifelong, often heterogeneous
correlational. Thus, the question of which brain areas are critical for impairments with music (Stewart, 2008; Omigie et al., 2012). In con-
music perception has remained elusive. Throughout the history of trast, AA reflects a clear-cut transition from a previously normal to a
neuroscience and neurology, lesion-based studies have provided the deficient music processing system caused by a brain lesion (e.g. stroke).
opportunity to explore the relationships between the lesioned brain This transition creates a unique opportunity to examine and pin down
areas and specific cognitive or motor functions (Rorden and Karnath, the neural structures that are most critical – and causally linked – to

☆
All authors have been involved in drafting or revising the article critically for intellectual content and have read and approved the final version of the manuscript.
⁎
Corresponding author at: Cognitive Brain Research Unit, Department of Psychology and Logopedics, Haartmaninkatu 3, P.O. Box 21, FI-00014 University of
Helsinki, Finland.
E-mail address: aleksi.sihvonen@helsinki.fi (A.J. Sihvonen).

https://doi.org/10.1016/j.neubiorev.2019.08.023
Received 13 March 2019; Received in revised form 27 August 2019; Accepted 29 August 2019
Available online 31 August 2019
0149-7634/ © 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
A.J. Sihvonen, et al. Neuroscience and Biobehavioral Reviews 107 (2019) 104–114

music perception (Price and Friston, 2002; Rorden and Karnath, 2004). 2. Structural and functional neural correlates of acquired amusia
The neural basis of CA has been relatively well mapped using
modern structural neuroimaging methods, such as voxel-based mor- 2.1. Voxel-based lesion-symptom mapping
phometry (VBM) and diffusion tensor imaging (DTI), as well as func-
tional and neurophysiological neuroimaging methods, including elec- Lesion-symptom relationships can be explored with VLSM, which
troencephalography (EEG) and magnetoencephalography (MEG) and uses the same voxel-based approach as other neuroimaging techniques
functional MRI (fMRI; for recent reviews, see Tillmann et al., 2015; (Bates et al., 2003). Since 2004, VLSM has been successfully used in
Peretz, 2016). Studies on CA have provided evidence of reduced white uncovering the lesion patterns of aphasia (Dronkers et al., 2004; Geva
matter concentration and abnormalities in grey matter volume (GMV) et al., 2012; Mirman et al., 2015), but it has been applied in AA only
or cortical thickness in the inferior frontal gyrus (IFG) and the superior recently. In two studies of stroke patients (Sihvonen et al., 2016, 2017a)
temporal gyrus (STG), predominantly in the right (Hyde et al., 2006, VLSM was used in conjunction with Scale and Rhythm subtests of the
2007; Albouy et al., 2013) but also in the left (Mandell et al., 2007) Montreal Battery of Evaluation of Amusia (MBEA; Peretz et al., 2003),
hemisphere. In addition, CA has been associated with reduced volume gold standard assessment for amusia, and in one study (Hirel et al.,
of the arcuate fasciculus (AF) interconnecting these frontotemporal 2017) with a music (melodic) short-term memory task similar to the
structures (Loui et al., 2009). However, conflicting DTI results have MBEA. In their study of 20 chronic stroke patients, Hirel et al. (2017)
been published leaving the tract deficit evidence in CA unsettled (Chen found that lesions in the insula and frontoparietal operculum were as-
et al., 2015). There is also some controversy regarding the functional sociated with deficits in performing the music task, but the lesion la-
role of the auditory cortex in CA, with fMRI and MEG studies indicating terality was not determined, as the left and right lesions were com-
either normal (Hyde et al., 2011; Norman-Haignere et al., 2016) or bined. Using a larger cohort (N = 90), Sihvonen et al. (2016, 2017a)
abnormal (Albouy et al., 2013, 2015; Albouy et al., 2019) processing of followed stroke patients with left (N = 43) and right (N = 47) lesions
pitch changes or tonal memory in the right STG. However, current from the acute to 6-month post-stroke stage. At the acute stage, AA was
evidence suggests that in CA conscious attention to and cognitive associated with a specific lesion pattern comprising right temporal
analysis of musical pitch information in right frontal regions, especially [STG, middle temporal gyrus (MTG)] and subcortical (striatum, globus
in the IFG, fails due to reduced functional connectivity between the IFG pallidus) regions as well as the right insula, IFG, and hippocampus
and STG (Peretz et al., 2005, 2009; Hyde et al., 2011; Albouy et al., (Fig. 2A). The lesion patterns for pitch-amusia and rhythm-amusia were
2015; Zendel et al., 2015; Leveque et al., 2016; Norman-Haignere et al., largely overlapping, though lesions in the right dorsal striatum were the
2016). CA is currently considered as a disconnection syndrome caused most significant in rhythm-amusia. In order to compare acute lesion
by anomalous recurrent processing in the right frontotemporal (dorsal) patterns predicting later recovery of AA, the amusic group was further
network (Peretz, 2016). divided based on MBEA performance at the 3-month and 6-month
In contrast to CA, the exploration of the neural basis of AA has stages, with patients scoring above the cut-off value of the MBEA
previously been limited to descriptive symptom-led studies of in- (Peretz et al., 2003) defined as showing good recovery (recovered
dividual cases and lesion-led studies of patient groups (for a review, see amusics) and patients scoring below the cut-off as showing poor re-
Stewart et al., 2006). Methodologically, these studies have been limited covery (non-recovered amusics). Lesions in the left IFG were associated
by small sample sizes and by both low temporal (data collected only with a better recovery of amusia whereas lesions in the right anterior
from a single time point) and spatial accuracy, with lesion areas re- STG, insula, and IFG were associated with a poor recovery (Fig. 2B).
ported at gross (lobar, hemispheric) anatomical level and not quanti- These results are similar to the VBM findings in CA (Hyde et al., 2006,
fied. Consequently, the conclusions on the location and laterality of the 2007; Albouy et al., 2013) and suggest that the right STG and IFG are
lesion and the type of consequent musical deficit have been coarse and the core areas underlying severe, persistent AA. However, these results
inconsistent. As shown in Fig. 1, which is an updated figure of the also indicate that damage to the left IFG may result in less severe,
previous distinguished review on AA by Stewart et al. (2006), the lesion transient AA.
sites reported in previous case and group studies include multiple Interestingly, there was virtually no overlap between the lesion
temporal, frontal, parietal, insular, and striatal areas in both left and patterns of AA and aphasia, the latter being located in the left STG and
right hemispheres (see also Supplementary Tables 1 and 2). The le- insula (Sihvonen et al., 2016), supporting the dissociation of AA and
sioned areas largely overlap for deficits in spectral (e.g., pitch, timbre) aphasia found in earlier case studies (for a recent review, see Peretz and
and temporal (e.g., rhythm, tempo) musical features, although right Coltheart, 2003). Nonetheless, around 45% of the amusic patients had
temporal lesions are predominant, especially for spectral deficits. The at least minor aphasia (Sihvonen et al., 2016; Sihvonen, Ripollés et al.,
few EEG and MEG studies of AA stroke patients have used auditory 2017a), similar to the percentages reported in earlier studies (Schuppert
oddball paradigms and shown reduced evoked responses to novel et al., 2000; Stewart et al., 2006), suggesting that musical and language
sounds reflecting impaired attentional processing (P3a, P3b; Münte deficits often occur together. In the healthy brain, speech and music
et al., 1998; Johannes et al., 1998) and to pitch and duration changes share neural processing resources in auditory cortical and frontal re-
reflecting impaired auditory sensory memory (MMN; Kohlmetz et al., gions (Koelsch and Siebel, 2005; Koelsch, 2011). However, as the two
2001; Särkämö et al., 2010), but have not localized the neural sources deficits can occur separately, it has been proposed that speech and
of these aberrant responses. music utilize overlapping neural resources but also have unique neural
Modern neuroimaging methods, including voxel-based lesion- representations in the brain (Patel, 2003). Especially the IFG has been
symptom mapping (VLSM; a method that goes beyond single case stu- suggested to be a crucial node for the shared syntactic integration re-
dies to identify brain regions that cause specific symptoms when le- sources between music and speech (Patel, 2003) whereas their pro-
sioned), VBM, DTI, and fMRI have been utilized to map the structural cessing in non-primary temporal lobe areas shows regional selectivity,
and functional neural correlates of AA in studies published only very with speech being processed in more ventrolateral areas and music in
recently. In this article, we review the current knowledge about the more dorsomedial areas, extending into the inferior parietal lobule (IPL;
neural basis of AA and its recovery, and we briefly outline a neural Tervaniemi et al., 2006; Rogalsky et al., 2011; Norman-Haignere et al.,
model of the key structural and functional changes in AA and discuss 2015). The abovementioned VLSM results also provide some support
their role in the recovery process and in light of normal music pro- for the shared resource hypothesis as they delineated lesions in the left
cessing. IFG in association with transient AA (Sihvonen, Ripollés et al., 2017a).
There are, however, some notable methodological limitations to
VLSM and to the causality and reliability of the conclusions that can be
drawn from it. First of all, most VLSM studies lack pre-injury data on

105
A.J. Sihvonen, et al. Neuroscience and Biobehavioral Reviews 107 (2019) 104–114

Fig. 1. Single case A) and group B) lesion studies regarding AA. The number of single cases / group studies which reported AA after damage to a particular brain
region is shown for both spectral (left) and temporal (right) deficits. The same number is reported in different slices (for example, three case studies reported spectral
deficits after damage to left frontal regions, so “3″ is shown in the red ROI in both axial and sagittal slices). Areas in which no deficit was reported have no number.
The different brain regions were identified using well-known atlases (Tzourio-Mazoyer et al., 2002; Frazier et al., 2005; Bürgel et al., 2006; Desikan et al., 2006;
Makris et al., 2006; Goldstein et al., 2007; Wakana et al., 2007; Hua et al., 2008). Neurological convention is used with brain regions shown over a canonical template
in MNI space. The figure is based on the data reported by Stewart et al. (2006; Supplementary Tables 1 & 2) and updated for studies from 2006 to 2018. See
Supplementary Tables 1 and 2 for details. AA = Acquired amusia; MNI = Montreal Neurological Institution; ROI = Region of interest.

performance in the function or skill being assessed, often relying on both local and more distant connections between brain areas (Catani
self-report or other indirect indices to assess the premorbid level. and ffytche, 2005), VLSM results need to be integrated also with in-
Second, VLSM results do not take diaschisis into account and therefore formation on the integrity of white matter pathways.
do not assess the potential effects of functional changes outside the
initial lesion site. Third, naturally occurring stroke lesions follow the
2.2. Diffusion tensor imaging
vascular territories of the brain and therefore are not typically focal and
restricted to one brain structure, but are generally larger and en-
Studies utilizing DTI in AA were lacking until 2016, when a sub-
compass multiple areas. Large lesions have been associated with poor
sample of the patients (N = 42) of the study by Sihvonen et al. (2016)
outcomes in aphasia (Samson et al., 1999; Benghanem et al., 2019), and
was used to determine white matter pathway deficits underlying AA
a similar observation was made also in the recent AA studies, with
(Sihvonen, Ripollés et al., 2017b). Converging results from determi-
larger lesion volumes in those amusics who did not show recovery over
nistic tractography (tract level) and tract-based spatial statistics (voxel
6 months compared to those who did (Sihvonen et al., 2016; Sihvonen,
level) analyses of acute stage DTI data showed that non-recovered
Ripollés et al., 2017a). Importantly, given that larger lesions can disrupt
amusic patients had reduced volume or fractional anisotropy (FA) in

106
A.J. Sihvonen, et al. Neuroscience and Biobehavioral Reviews 107 (2019) 104–114

Fig. 2. Acute changes in AA. A) Stroke lesion pattern associated with lower MBEA total scores (N = 90); B) VLSM analyses comparing non-recovered amusic vs. non-
amusic patients (red: N = 74), and recovered amusic vs. non-amusic patients (blue; N = 53); C) TBSS analysis comparing non-recovered amusic vs. non-amusic
patients (N = 42); D) Main DT findings in AA at the acute stage (N = 42); E) fMRI activation pattern comparison between the non-amusic and amusic patients during
instrumental music listening at the acute stage (N = 41), and right frontoparietal network functional connectivity engagement comparison between recovered and
non-recovered amusic patients (N = 24). CAU = Caudate; DT = Deterministic tractography; GP = Globus pallidus; HIP = Hippocampus; INS = Insula;
PUT = Putamen; R = Right.

the right AF (long segment), inferior fronto-occipital fasciculus (IFOF), presented using a block-design with six blocks of vocal music, six blocks
and uncinate fasciculus (UF), as well as in the corpus callosum (CC) of instrumental music and 12 blocks of rest (no-stimuli) in between the
compared to the non-amusics (Fig. 2C-D). In contrast, volume of the left music blocks. The duration of each block was 15 s. At the acute post-
AF (posterior segment) was associated with recovery of amusia. Im- stroke stage, amusic patients showed clearly reduced activation to in-
portantly, all these effects emerged after controlling for potential con- strumental music in the right STG/MTG compared to non-amusics
founding factors, including education, lesion size, and verbal memory (Fig. 2E). Functional connectivity analyses indicated that the amusics
performance, increasing their specificity for musical deficits. Overall, with poor recovery showed reduced engagement of the right fronto-
the recent DTI results in AA converge with the right dorsal stream parietal network compared to amusics with good recovery, suggesting
deficit observed in CA (Loui et al., 2009), but also implicate other (i.e. impaired allocation of attention to instrumental music, already at the
ventral) right frontotemporal tracts (IFOF, UF) as well as interhemi- acute stage. Interestingly, amusics did not show decreased activations
spheric tracts (CC) in severe and persistent AA. These findings in or functional connectivity deficits in the vocal music condition. In fact,
amusia parallel with recent findings on aphasia indicating that not only compared to non-amusics, the amusics showed enhanced activation in
the left dorsal (AF) but also left ventral tracts (IFOF, UF) underlie right frontal [IFG; middle frontal gyrus (MFG)], temporal (Heschl’s
deficits in speech processing, especially language comprehension gyrus, HG), and parietal [IPL, post central gyrus (PCG)] areas, the left
(Kummerer et al., 2013; Harvey et al., 2013; Ivanova et al., 2016). supplementary motor area (SMA), and bilateral medial parieto-occipital
(cuneus, precuneus) and frontal (anterior cingulate) areas to vocal vs.
2.3. Functional magnetic resonance imaging instrumental music at the 3-month stage (Fig. 3C). As there were no
significant differences between the non-amusic and amusic groups or
Using fMRI in stroke patients (N = 41), Sihvonen and colleagues amusics with good or poor recovery in aphasia occurrence (Sihvonen
also explored changes in activation patterns and functional connectivity et al., 2017d), this effect of preserved vocal music processing in AA is
in AA during passive listening to vocal (sung) and instrumental versions unlikely to be explained by post-stroke aphasia.
of familiar songs (Sihvonen et al., 2017d). Music excerpts were

107
A.J. Sihvonen, et al. Neuroscience and Biobehavioral Reviews 107 (2019) 104–114

Fig. 3. Longitudinal changes in AA. A) Grey (red) and white matter (blue) VBM analyses comparing recovered and non-recovered amusic patients (N = 53); B) TBSS
analysis comparing non-recovered amusic vs. non-amusic patients (N = 25); C) fMRI activation pattern comparison between the non-amusic and amusic patients
during instrumental music listening and amusic vs. non-music patients during vocal music > instrumental music listening at the 3-month stage (N = 41); D) Left
frontoparietal network functional connectivity engagement comparison between recovered and non-recovered amusic patients (N = 24). CIN = Cinculate gyrus;
CUN = Cuneus; HG = Heschl’s gyrus; HIP = Hippocampus; INS = Insula; IPL = Inferior parietal lobule; L = Left; MFG = Middle frontal gyrus; PCG = Postcentral
gyrus; PUT = Putamen; R = Right.

3. Longitudinal neural changes underlying the recovery of greater sensitivity to changes in the temporal domain and posterior
acquired amusia temporal regions show great sensitivity to changes in the spectral do-
main. In addition to the right posterior STG/MTG, the right IPL has an
3.1. Voxel-based morphometry important role in melodic perception, especially is maintaining tonal
pitch structure in working memory (Royal et al., 2016).
Sihvonen and colleagues utilized VBM in stroke patients (N = 90) to
explore GMV and white matter volume (WMV) changes associated with 3.2. Diffusion tensor imaging
the longitudinal (acute to 6-month post-stroke stage) outcome of AA
(Sihvonen et al., 2016; Sihvonen, Ripollés et al., 2017a). While con- The longitudinal DTI data linked persistent amusia to a progressive
trolling for education and lesion size, poor vs. good amusia recovery white matter damage, indicated by increased mean or radial diffusivity
was associated with GMV decrease in the right STG/MTG and IFG, and (MD/RD), in the right IFOF, AF, and UF as well as in the CC (Fig. 3B;
with WMV decrease in the right MTG, inferior temporal gyrus (ITG), Sihvonen, Ripollés et al., 2017b). This degenerative pattern was similar
striatum, and hippocampus (Fig. 3A), indicating that impaired amusia in rhythm- and pitch-amusia for the right frontotemporal tracts (IFOF,
recovery is associated with atrophy in areas adjacent to the initial le- AF, UF), but there were also some differences. In rhythm-amusia, poor
sion. This atrophy pattern was somewhat different depending on recovery was additionally associated with MD/RD increase in the left
amusia type: poor recovery was linked to right posterior STG/MTG and AF. In pitch-amusia, MD/RD increase in the posterior part of the CC,
IPL atrophy in pitch-amusia and to right anterior MTG/ITG and hip- which connects the left and right temporal lobes, was linked to poor
pocampus atrophy in rhythm-amusia. The anterior-posterior distribu- recovery whereas MD/RD decrease in the right anterior AF was linked
tion of the rhythm- and pitch-amusia atrophy patterns along the tem- to good recovery, suggesting that preservation of this dorsal tract
poral plane is similar to the findings reported in both animals (Bendor connecting the right IFG and IPL plays a role in the recovery of pitch-
and Wang, 2008) and humans (Warren et al., 2005; Jamison et al., amusia. Interestingly, and in contrast with the deficient dorsal tract
2006; Samson et al., 2011) where anterior temporal regions show thought to underlie CA, the ventral tract (IFOF) deficiency was the

108
A.J. Sihvonen, et al. Neuroscience and Biobehavioral Reviews 107 (2019) 104–114

strongest predictor of AA in two regression analyses. leading to AA. From ontogenetic and phylogenetic standpoints, the
results linking the right IFOF to music perception is particularly inter-
3.3. Functional magnetic resonance imaging esting since in humans, the IFOF is known to be present already at birth
(Perani et al., 2011), but it is clearly less developed in monkeys
Longitudinal fMRI data showed that in the instrumental music (Thiebaut de Schotten et al., 2012). While the prime neural archi-
condition, good vs. poor amusia recovery was linked to increased ac- tectures in AA were right-lateralised, amusia was also associated with
tivation in left and right MFG and IPL, left superior parietal lobule damage to interhemispheric connectivity (i.e. the CC and tapetum),
(SPL), and right precentral gyrus (PreCG) at the 3-month stage and in more pronouncedly in rhythm-amusia (Sihvonen, Ripollés et al.,
right IFG and MFG at the 6-month stage as well as to increased func- 2017b). In contrast, damage to left frontal areas (IFG) or left tempor-
tional connectivity in the left frontoparietal network, especially during oparietal pathways were associated with better later recovery (i.e.
the first 3 months (Fig. 3C-D; Sihvonen et al., 2017d). Lesion size did transient AA). Moreover, the lesion pattern in AA functionally inhibits
not differ between the recovered and non-recovered amusics. These the engagement of the right STG/MTG in music listening (Sihvonen
results suggest that amusia recovery is underpinned by a dynamic shift et al., 2017d). Altogether, these findings support the rationale of ad-
characterized by a widespread recruitment of bilateral frontoparietal ditional left hemispheric subsystems needed in music perception and
regions at the early stage and more focal right prefrontal regions at later processing, although with right hemisphere dominance.
stage. This pattern was instrumental music specific as similar changes The longitudinal recovery process of AA is driven by a complex set
were not observed in the vocal condition, which also suggests that the of both structural and functional neural changes. Impaired recovery of
observed effects cannot be due to deficits in general attentional or- AA was associated with atrophy in the right cortical (STG, MTG, ITG,
ientation. In the vocal condition, increased bilateral cerebellar activa- IFG) and subcortical (striatum, hippocampus) areas adjacent to the le-
tion at 6 months was seen in the non-recovered compared to recovered sion as well as degeneration of the right frontotemporal tracts (IFOF,
amusics, potentially indicating that AA recovery is associated with less AF, UF), the left AF (in rhythm-amusics), and the posterior CC (in pitch-
neural effort in the sensory-motor processing and possible covert motor amusics). Poor recovery of amusia was associated with large lesion size
articulation of singing (Callan et al., 2007; Brattico et al., 2011). and predicted by the extent of damage to the right frontotemporal
Overall, the pattern of increased bilateral activation to music in AA pathways (mainly ventral). In addition, the amusics who did not re-
recovery closely mirrors that of language recovery: aphasics who re- cover over the following 6 months had reduced functional connectivity
cover show an upregulation of bilateral frontal regions, especially IFG in the right frontoparietal network at the acute stage compared to the
and SMA, at subacute stage, followed by normalization of activation recovered amusics. In contrast, good recovery of AA was promoted by
with a re-shift to the left hemisphere in the chronic stage (Saur et al., increased recruitment of both left and right dorsal streams as well as the
2006). interhemispheric pathways. Amusia recovery was also facilitated by
preservation of the right anterior AF (in pitch-amusia) as well as in-
4. Towards a neural model of acquired amusia creased functional recruitment of largely bilateral frontal (IFG, MFG,
PreCG) and parietal (IPL, SPL) regions and the left frontoparietal net-
Functional neuroimaging studies on healthy subjects have shown work.
that music perception involves a wide-spread network comprising bi- In language research, two processing streams, dorsal and ventral,
lateral temporal, frontal, parietal, and subcortical brain regions are broadly accepted to underlie speech processing (Rauschecker and
(Schmithorst, 2005; Brattico et al., 2011; Alluri et al., 2012). Studies on Tian, 2000; Hickok and Poeppel, 2007; Rauschecker, 2014). The lan-
stroke patients have provided similar evidence of cross-hemispheric guage dual-stream model is largely left dominant, but, for example,
network underlying music perception. However, as suggested first by spectrotemporal analysis and phonological processing are thought to be
Peretz (1990) and later by Schuppert and colleagues (2000), music largely bilateral, and overall the model assumes that the ventral stream
perception relies on initial recognition of global musical structures in is largely bilaterally organized (Hickok and Poeppel, 2007). A similar
the right hemisphere, supported by the left hemisphere subsystems that dual-stream model has been proposed to act in parallel in music pro-
are subordinate to the right hemisphere (Peretz, 1990; Schuppert et al., cessing, transferring crucial musical auditory information between the
2000). In rhythm perception for example, these different systems in- temporal, inferior parietal, and inferior frontal regions in the right
terpret metre (global) or discriminate the durational values of notes hemisphere (Zatorre et al., 2002; Rauschecker, 2014; Sammler et al.,
(local) to process the temporal information of auditory stimuli (Peretz, 2015; Loui, 2015; Musso et al., 2015). Of the two streams, the dorsal
1990). In general, spatially distributed brain areas subserving a cogni- stream (“where” or “how”) connecting temporal and inferior parietal
tive function are connected through white matter tracts which form a regions with frontal areas is hypothesized to be important for evalua-
network to maximize the processing, storage, and manipulation of in- tion of audio-motor movement and spatial information, whereas the
formation (Ross, 2010). Disruption of the neural network and its con- ventral stream (“what”) connecting parieto-occipital, temporal and in-
nections can lead to a disconnection syndrome and a cognitive-beha- ferior frontal areas is involved in categorizing sound to auditory ob-
vioural deficit (Catani and Mesulam, 2008; Thiebaut de Schotten et al., jects, pitch class and melodic contour (Rauschecker and Tian, 2000;
2008). In musical domain, analysing musical structures demand various Griffiths et al., 2007; Rauschecker and Scott, 2009; Rauschecker, 2014;
cognitive architectures which can be selectively impaired leading to a Sammler et al., 2015). In aphasia, damage to the dorsal stream is as-
fine-grained forms of AA (Clark et al., 2015). sociated with productive impairments, while comprehension deficits
Based on the recent multimodal MRI results discussed above, which are associated with injury of the ventral stream (extreme capsule, or
extend earlier symptom-led and lesion-led explorations of AA (Stewart rather the IFOF; Kummerer et al., 2013). In the musical domain, it is
et al., 2006) as well as provide a more precise, spatially accurate, and likewise possible that damage to individual pathways (dorsal or ven-
comprehensive picture of the structural and functional neural changes tral) would manifest in different musical impairments (production
underpinning AA and its recovery, the critical connections for music versus perception; Loui et al., 2008; Loui, 2015; Sammler et al., 2015).
perception seem to be located in the right hemisphere, in contrast with If both the ventral and dorsal streams are damaged, due to larger lesion
the bilateral large-scale music network observed in healthy subjects. size or location, it is unlikely that AA recovers. Instead, acquired
The core of the lesion area causing AA is in the right insula and striatum amusics with at least one preserved right hemispheric music-related
from where the lesion extends to temporal (STG/MTG), frontal (IFG), pathway interconnecting frontal, parietal and temporal regions could
and limbic (hippocampus) areas. This lesion area affects crucial fron- engage recovery as the two streams have been found to share func-
totemporal pathways, especially the ventral stream, but also the dorsal tionalities and mediate compensatory mechanisms in the language
pathway, linking the right temporal and inferior frontal regions, thus domain (Lopez-Barroso et al., 2011). This rationale is supported by the

109
A.J. Sihvonen, et al. Neuroscience and Biobehavioral Reviews 107 (2019) 104–114

recent results (Sihvonen et al., 2017a, 2017b). speaking. Both functions engage a large network including sensor-
While the structural findings in AA were largely right-lateralised, imotor areas and inferior frontal regions, but compared to speaking,
functional anomalies were observed in both hemispheres during in- singing induces greater activations in the right temporoparietal regions
strumental music listening. In addition to the global (i.e. metre) versus as well as in right frontal regions (Callan et al., 2006; Özdemir et al.,
local (i.e. rhythm) dissociation of auditory information processing dis- 2006). The recent AA studies provide evidence on both shared and
cussed previously (Peretz, 1990; Schuppert et al., 2000), an additional distinct resources underlying music and language processing in the
explanation is that the critical hubs, meaning brain areas crucial for brain. Post-stroke language and musical deficits can often occur to-
enabling neuronal signalling and communication (van den Heuvel and gether (Sihvonen et al., 2016; Sihvonen, Ripollés et al., 2017a), yet
Sporns, 2013), in the music processing network are located in the right these domains can be selectively impaired (Sidtis and Volpe, 1988;
hemisphere (Schmithorst, 2005; Brattico et al., 2011; Alluri et al., Peretz et al., 1997; Tzortzis et al., 2000; Piccirilli et al., 2000; Mendez,
2012) and damage to these neural structures manifests in wide-spread 2001; Uetsuki et al., 2016). The lesion patterns giving rise to aphasia
processing deficits during music perception. Collectively, the previously and persistent AA showed clear lateralization (Sihvonen et al., 2016),
published studies suggest that these critical hubs are the right STG/ clustering in left and right hemispheres, respectively. In contrast, the
MTG and IFG, and moreover comprise the white matter pathways in- stroke lesions associated with recoverable AA were localized primarily
terconnecting the two regions, similar to CA (Peretz, 2016). However, in the left IFG, a crucial node of shared resources between music and
the critical connection damaged in AA seems to be the right ventral speech (Patel, 2003). In AA, the right ventral and dorsal streams were
stream as opposed to the dorsal stream. As AA can occur also after left the most critically damaged white matter pathways, with the degree of
hemisphere damage (Schuppert et al., 2000) lesions affecting the cru- initial damage to the right ventral stream being the most significant
cial tracts interconnecting music processing areas in the left hemisphere predictor of severe AA (Sihvonen, Ripollés et al., 2017b). As the right
to the crucial right hemispheric music-related brain regions (i.e. critical ventral stream is implicated in both language (Hickok and Poeppel,
hubs) might lead to AA. Moreover, intact right hemisphere has been 2007) and music processing (Sihvonen, Ripollés et al., 2017b), based on
suggested to compensate for the music perception deficits after left the recent results, this suggests that this crucial pathway is a part of the
hemisphere damage (Schuppert et al., 2000) which underlines the im- shared neural network between these two domains with a role espe-
portance of the right hemisphere in music processing. One other pos- cially in the encoding and analysis melodic features important for music
sible explanation is that during naturalistic music listening, more global and speech perception. In contrast, the processing of syntactic in-
auditory information processing is needed, in contrast to local proces- formation in music and language have been shown to share a left
sing (Peretz, 1990; Schuppert et al., 2000). hemispheric network centred upon left IFG and on left dorsal and
Moreover, the disparity between the lateralisation of lesion and ventral pathways (Musso et al., 2015). Interestingly, and in contrast
functional anomalies could arise from stimulus complexity. In language with syntactic processing, the perception of prosody relies on both
domain, the lateralisation of prosodic emotion processing is dependent dorsal and ventral pathways in the right hemisphere (Wildgruber et al.,
on the verbal complexity: As the complexity of the presented prosodic 2006; Sammler et al., 2015; Frühholz et al., 2015). While post-stroke
information increases from sentences with asyllabic to monosyllabic aprosodia has also been linked to right hemisphere damage (Ross and
and to multi-syllable words, the fMRI patterns shifts from being pre- Monnot, 2008; Ross, 2010; Jafari et al., 2017), recent evidence suggests
dominantly right-lateralized to bilateral activity (Mitchell and Ross, a cross-talk between the hemispheres via the posterior corpus callosum
2008). Similarly, music contains complex acoustic components as well (i.e. tapetum) is needed to combine prosodic and syntactic information
as a language component, and therefore it is reasonable to expect bi- in language (Sammler et al., 2018). These posterior transcallosal fibers
lateral wide-spread brain activations during music listening (Toiviainen were also implicated in AA, especially in pitch-amusia (Sihvonen,
et al., 2014) than during listening to single tones or melodies (Hyde Ripollés et al., 2017b), and similar to language processing, it is possible
et al., 2008; Jerde et al., 2011). Taken together, right hemisphere le- that the pitch information in music is transferred to the left hemisphere
sions leading to AA might manifest in wide-spread global music pro- for syntactic integration via this pathway, as music-syntactic processing
cessing deficits whereas left hemisphere damage might affect only local has been shown to engage the IFG bilaterally, but with right-hemi-
processing and thus lead to small-scale activation deficits. spheric weighting (Maess et al., 2001; Koelsch and Siebel, 2005;
Music with vocals combines characteristics of both language (e.g. Tillmann et al., 2006). Overall, acting in conjunction, the two streams
linguistic syntax, semantics) and music (e.g. melody, harmony, transform complex acoustic feature combinations into abstract re-
rhythm). Interestingly, amusics showed enhanced activation to vocal presentations and analyse sensorimotor information to be integrated
vs. instrumental music in an extensive network of right temporal (HG) with these representations (Rauschecker and Scott, 2009). The dis-
and frontoparietal (IFG, MFG, IPL, PCG) cortical areas, left SMA, and cussed data suggest that normal music perception relies on this dual
medial frontoparietal (e.g., cingulate, precuneus) areas (Sihvonen et al., route, especially in the right hemisphere. Crucially, the recent results
2017d). This suggests that the processing of the vocal component of enable modelling the cascade of lesion-induced structural changes
music is relatively well preserved in AA and mediated by spared au- giving rise to AA (Sihvonen et al., 2016; Hirel et al., 2017; Sihvonen,
ditory, vocal-motor, and attention- and memory-related regions, similar Ripollés et al., 2017a) and the structural (Sihvonen et al., 2016;
to those being reported in fMRI studies of healthy subjects comparing Sihvonen, Ripollés et al., 2017a, b) and functional (Sihvonen et al.,
vocal and instrumental music processing (Brattico et al., 2011; Alluri 2017d) changes that either impair or facilitate recovery from AA
et al., 2013; Lévêque and Schön, 2015). Moreover, as there were no (Fig. 4).
significant differences between the groups in the occurrence of post-
stroke aphasia, the observed results are unlikely to be explained by 5. Conclusions, clinical considerations and future directions
dissimilarities in language processing between the amusic and non-
amusic patients. However, this finding is based on one study and it While AA is a common deficit after a middle cerebral artery stroke,
therefore needs to be replicated and further studied in the future for with prevalence ranging between 35% and 69% (Ayotte et al., 2000;
more details. Schuppert et al., 2000; Särkämö et al., 2009), it is not routinely eval-
Overall, neuroimaging studies on healthy subjects have shown that uated in clinical practice and therefore remains underdiagnosed. The
the processing of music and speech share resources in the brain, but novel findings reported in this review enable an accurate identification
preferentially engage distinct cortical networks: listening to music ac- of AA, based on recognition of amusia lesion pattern reported by VLSM
tivates the insula and superior temporal regions bilaterally as well as studies and whether it damages the ventral or dorsal streams or both in
the right IFG more than listening to speech (LaCroix et al., 2015). Si- the right hemisphere, and subsequent evaluation of the patient´s music
milarly, both shared and distinct neural networks facilitate singing and abilities (i.e. perception and production) using, for example, the MBEA

110
A.J. Sihvonen, et al. Neuroscience and Biobehavioral Reviews 107 (2019) 104–114

Fig. 4. Dual-stream model of acquired amusia. Both the right ventral and the dorsal stream play an important role in amusia recovery, and if both ventral and dorsal
streams in the right hemisphere are damaged, recovery from acquired amusia is unlikely. However, if one of these streams in the right hemisphere is preserved,
recovery is possible due to the shared functionalities of the streams.

(Peretz et al., 2003) as a screening test for amusia. The results cited here in a music emotional recognition task, the intensity of the emotional
also provide a prognostic tool for AA outcome: initially amusic patients experience of music was intact when compared to a matched control
with smaller lesions and preserved dorsal stream are more likely to sample (Leveque et al., 2018). This study highlights the importance of
recover. While the current studies focused on cognitive (pitch and further studying the impact of amusia in emotional processing, as for
rhythm processing) aspects of music, the hedonic qualities of music example the possible dissociation between emotional recognition and
should not be neglected as patients with right hemisphere damage have emotional impact.
shown to have impaired emotional, rewarding, vegetative and senti- Given the quintessential role of music in human life, diagnosing AA
mental aspects of music (Behne, 1997). In healthy subjects, musical is especially important in patients who are vocationally engaged with
anhedonia has been associated mainly with selective reduction of ac- music, such as musicians and music teachers, but also in those patients
tivity in the nucleus accumbens, as well as its decreased functional and who have musical hobbies or for whom music-based rehabilitation is
structural connectivity to the right auditory cortex (Martinez-Molina considered (Sihvonen et al., 2017b). Based on the recent fMRI study on
et al., 2016; Hernandez et al., 2019; Martinez-Molina et al., 2019). The AA (Sihvonen et al., 2017d), if music-based interventions are con-
first case describing acquired musical anhedonia was reported over 25 sidered for amusic patients, vocal music listening and singing-based
years ago (Mazzoni et al., 1993). This patient with a right tempor- intervention could be promising as vocal music processing seems to be,
oparietal lesion was observed to have lost the capacity to evaluate at least partly, preserved in amusic brain. More particularly, if AA is
qualitative aspects of music experiences, and specially the emotional accompanied with aphasia, Melodic Intonation Therapy (MIT; Albert
involvement of music, in spite of performing normally in all cognitive et al., 1973) might still be a viable method of language rehabilitation as
domains (see also following case reports of musical anhedonia after the processing of vocal aspects in music are preserved (Sihvonen et al.,
brain lesions Griffiths et al., 2004; Satoh et al., 2011, 2016). Belfi et al. 2017d). However, as the proposed pathway to language recovery using
(2017) investigated a large group of patients with focal brain damage MIT is through recruitment of homologous language and speech-motor
with the Musical Anhedonia Questionnaire and the Barcelona Music regions in the right hemisphere (Schlaug et al., 2010) utilizing the right
Reward Questionnaire (Mas-Herrero et al., 2013) to assess whether the dorsal stream (Schlaug et al., 2009; Wan et al., 2014), the gains ob-
ability to feel musical enjoyment and reward is impaired after brain tained through MIT are likely to be dependent on the preservation of
injury (Belfi et al., 2017). However, no pattern of brain lesions emerged the right dorsal stream. Amusia is often coupled also with problems in
that could reliably predict musical anhedonia. Overall, musical anhe- linguistic (Liu et al., 2010) or emotional (Thompson et al., 2012; Jafari
donia was rare among the patients, suggesting that musical reward may et al., 2017) prosody, which rely on the perception subtle pitch, timbre,
be highly resistant to neural damage or that damage in several parts on and intensity variations in speech. This may lead to difficulties in ev-
the auditory-reward decoding network might induce similar deficits eryday communication and social interaction. While intervention stu-
(Loui et al., 2017; Martinez-Molina et al., 2019). Some authors have dies on AA are lacking, singing interventions have been found pro-
reported cases in which the loss of aesthetic pleasure was concurrent mising in CA (Anderson et al., 2012; Wilbiks et al., 2016) and in light of
with AA (Mazzucchi et al., 1982; Hirel et al., 2014). In this regard, a the preserved vocal music processing observed with fMRI in AA
recent study has shown that while subjects with CA showed impairment (Sihvonen et al., 2017d) could provide a tool also in the rehabilitation

111
A.J. Sihvonen, et al. Neuroscience and Biobehavioral Reviews 107 (2019) 104–114

of AA. Ann. N. Y. Acad. Sci. 1252, 345–353.

As a whole, music processing entails multiple elements, ranging Ayotte, J., Peretz, I., Rousseau, I., Bard, C., Bojanowski, M., 2000. Patterns of music
agnosia associated with middle cerebral artery infarcts. Brain 123, 1926–1938.
from the perceptual analysis of different musical components, including Bates, E., Wilson, S.M., Saygin, A.P., Dick, F., Sereno, M.I., Knight, R.T., et al., 2003.
tonality, melodic intervals and contours, rhythm, and meter, to pro- Voxel-based lesion-symptom mapping. Nat. Neurosci. 6, 448–450.
ductive elements, such as singing and moving to music, the impairment Behne, K.-, 1997. The development of "musikerleben" in adolescence: how and why young
people listen to music. In: Deliège, I., Sloboda, J. (Eds.), Perception and Cognition of
of which can give rise to a spectrum of more fine-grained forms of AA Music. Hove. Psychology Press/Erlbaum (UK) Taylor & Francis, England, pp.
(Peretz and Coltheart, 2003). Although most recent studies discussed 143–159.
above have utilized only MBEA in testing specific musical abilities (i.e. Belfi, A.M., Evans, E., Heskje, J., Bruss, J., Tranel, D., 2017. Musical anhedonia after focal
brain damage. Neuropsychologia 97, 29–37.
pitch and rhythm perception), their results have revealed functionally Bendor, D., Wang, X., 2008. Neural response properties of primary, rostral, and ros-
crucial neural components of the music network. In future, further re- trotemporal core fields in the auditory cortex of marmoset monkeys. J. Neurophysiol.
search on both perceptual and productive musical deficits is needed to 100, 888–906.
Benghanem, S., Rosso, C., Arbizu, C., Moulton, E., Dormont, D., Leger, A., et al., 2019.
determine in depth how the music processing structures function at
Aphasia outcome: the interactions between initial severity, lesion size and location. J.
network level. This would be promoted by studying AA that stems from Neurol. 266, 1303–1309.
various neurological diseases with different symptoms, neural bases Brattico, E., Alluri, V., Bogert, B., Jacobsen, T., Vartiainen, N., Nieminen, S., et al., 2011.
and trajectories. Although this review focuses on the recent advances in A functional MRI study of happy and sad emotions in music with and without lyrics.
Front. Psychol. 2, 308.
understanding stroke-induced amusia, studying musical abilities in Bürgel, U., Amunts, K., Hoemke, L., Mohlberg, H., Gilsbach, J.M., Zilles, K., 2006. White
persons with dementia is encouraged as these patients have been shown matter fiber tracts of the human brain: three-dimensional mapping at microscopic
to have pitch processing deficits compared to healthy age-matched resolution, topography and intersubject variability. Neuroimage 29, 1092–1105.
Callan, D.E., Kawato, M., Parsons, L., Turner, R., 2007. Speech and song: the role of the
controls, while the function of various other musical modules (e.g. cerebellum. Cerebellum 6, 321–327.
temporal and timbre processing) remain intact (Golden et al., 2017). Callan, D.E., Tsytsarev, V., Hanakawa, T., Callan, A.M., Katsuhara, M., Fukuyama, H.,
Moreover, studying music processing, both intact and defective, across et al., 2006. Song and speech: brain regions involved with perception and covert
production. Neuroimage 31, 1327–1342.
different musical cultures would provide information on universally Catani, M., Mesulam, M., 2008. The arcuate fasciculus and the disconnection theme in
shared musical neural structures. Methodologically, precise modelling language and aphasia: history and current state. Cortex 44, 953–961.
of structural and functional connections of the individual musical Catani, M., ffytche, D.H., 2005. The rises and falls of disconnection syndromes. Brain 128,
2224–2239.
modules is needed, which requires improved MRI methods, compre- Chen, J.L., Kumar, S., Williamson, V.J., Scholz, J., Griffiths, T.D., Stewart, L., 2015.
hensive testing of the subjects’ musical abilities as well as larger sample Detection of the arcuate fasciculus in congenital amusia depends on the tractography
sizes and even combination of data sets. In addition, longitudinal algorithm. Front. Psychol. 6, 9.
Clark, C.N., Golden, H.L., Warren, J.D., 2015. Acquired amusia. Handb. Clin. Neurol. 129,
analyses of lesion sizes and their relation to AA recovery are needed.
607–631.
Application of state-of-the-art multivoxel pattern analysis to the VLSM Desikan, R.S., Segonne, F., Fischl, B., Quinn, B.T., Dickerson, B.C., Blacker, D., et al.,
(Sperber et al., 2018) and functional data (Zavaglia et al., 2015) ob- 2006. An automated labeling system for subdividing the human cerebral cortex on
tained from patients with AA would provide even more precise view of MRI scans into gyral based regions of interest. Neuroimage 31, 968–980.
Dronkers, N.F., Wilkins, D.P., Van Valin Jr, R.D., Redfern, B.B., Jaeger, J.J., 2004. Lesion
the essential neural structures involved in music perception. analysis of the brain areas involved in language comprehension. Cognition 92,
145–177.
Funding Frazier, J.A., Chiu, S., Breeze, J.L., Makris, N., Lange, N., Kennedy, D.N., et al., 2005.
Structural brain magnetic resonance imaging of limbic and thalamic volumes in pe-
diatric bipolar disorder. Am. J. Psychiatry 162, 1256–1265.
Financial support for the work was provided by the Academy of Freitas, C., Manzato, E., Burini, A., Taylor, M.J., Lerch, J.P., Anagnostou, E., 2018. Neural
Finland (grants 257077, 277693, 299044), Tyks Research Funding correlates of familiarity in music listening: a systematic review and a neuroimaging
meta-analysis. Front. Neurosci. 12, 686.
(grant 13944), Finnish Brain Research and Rehabilitation Foundation, Frühholz, S., Gschwind, M., Grandjean, D., 2015. Bilateral dorsal and ventral fiber
Finnish Brain Foundation, Finnish Cultural Foundation, Ella and Georg pathways for the processing of affective prosody identified by probabilistic fiber
Ehrnrooth Foundation, Signe and Ane Gyllenberg Foundation, Maire tracking. Neuroimage 109, 27–34.
Geva, S., Baron, J.C., Jones, P.S., Price, C.J., Warburton, E.A., 2012. A comparison of
Taponen Foundation, and Jenny and Antti Wihuri Foundation.
VLSM and VBM in a cohort of patients with post-stroke aphasia. Neuroimage Clin. 1,
37–47.
Appendix A. Supplementary data Golden, H.L., Clark, C.N., Nicholas, J.M., Cohen, M.H., Slattery, C.F., Paterson, R.W.,
et al., 2017. Music perception in dementia. J. Alzheimers Dis. 55, 933–949.
Goldstein, J.M., Seidman, L.J., Makris, N., Ahern, T., O’Brien, L.M., Caviness Jr, V.S.,
Supplementary material related to this article can be found, in the et al., 2007. Hypothalamic abnormalities in schizophrenia: sex effects and genetic
online version, at doi:https://doi.org/10.1016/j.neubiorev.2019.08. vulnerability. Biol. Psychiatry 61, 935–945.
023. Gordon, C.L., Cobb, P.R., Balasubramaniam, R., 2018. Recruitment of the motor system
during music listening: an ALE meta-analysis of fMRI data. PLoS One 13, e0207213.
Graziano, A.B., Johnson, J.K., 2015. Music, neurology, and psychology in the nineteenth
References century. Prog. Brain Res. 216, 33–49.
Griffiths, T.D., Warren, J.D., Dean, J.L., Howard, D., 2004. "When the feeling’s gone": a
selective loss of musical emotion. J Neurol Neurosurg Psychiatry 75, 344–345.
Adolphs, R., 2016. Human lesion studies in the 21st century. Neuron 90, 1151–1153. Griffiths, T.D., Kumar, S., Warren, J.D., Stewart, L., Stephan, K.E., Friston, K.J., 2007.
Albert, M.L., Sparks, R.W., Helm, N.A., 1973. Melodic intonation therapy for aphasia. Approaches to the cortical analysis of auditory objects. Hear. Res. 229, 46–53.
Arch. Neurol. 29, 130–131. Harvey, D.Y., Wei, T., Ellmore, T.M., Hamilton, A.C., Schnur, T.T., 2013.
Albouy, P., Mattout, J., Sanchez, G., Tillmann, B., Caclin, A., 2015. Altered retrieval of Neuropsychological evidence for the functional role of the uncinate fasciculus in
melodic information in congenital amusia: insights from dynamic causal modeling of semantic control. Neuropsychologia 51, 789–801.
MEG data. Front. Hum. Neurosci. 9, 20. Hernandez, M., Palomar-Garcia, M.A., Nohales-Nieto, B., Olcina-Sempere, G., Villar-
Albouy, P., Peretz, I., Bermudez, P., Zatorre, R.J., Tillmann, B., Caclin, A., 2019. Rodriguez, E., Pastor, R., et al., 2019. Separate contribution of striatum volume and
Specialized neural dynamics for verbal and tonal memory: fMRI evidence in con- pitch discrimination to individual differences in music reward. Psychol. Sci.,
genital amusia. Hum. Brain Mapp. 40, 855–867. 956797619859339.
Albouy, P., Mattout, J., Bouet, R., Maby, E., Sanchez, G., Aguera, P.E., et al., 2013. Hickok, G., Poeppel, D., 2007. The cortical organization of speech processing. Nat. Rev.
Impaired pitch perception and memory in congenital amusia: the deficit starts in the Neurosci. 8, 393–402.
auditory cortex. Brain 136, 1639–1661. Hirel, C., Nighoghossian, N., Leveque, Y., Hannoun, S., Fornoni, L., Daligault, S., et al.,
Alluri, V., Toiviainen, P., Jääskeläinen, I.P., Glerean, E., Sams, M., Brattico, E., 2012. 2017. Verbal and musical short-term memory: variety of auditory disorders after
Large-scale brain networks emerge from dynamic processing of musical timbre, key stroke. Brain Cogn. 113, 10–22.
and rhythm. Neuroimage 59, 3677–3689. Hirel, C., Leveque, Y., Deiana, G., Richard, N., Cho, T.H., Mechtouff, L., et al., 2014.
Alluri, V., Toiviainen, P., Lund, T.E., Wallentin, M., Vuust, P., Nandi, A.K., et al., 2013. Acquired amusia and musical anhedonia. Rev. Neurol. (Paris) 170, 536–540.
From Vivaldi to Beatles and back: predicting lateralized brain responses to music. Hua, K., Zhang, J., Wakana, S., Jiang, H., Li, X., Reich, D.S., et al., 2008. Tract probability
Neuroimage 83, 627–636. maps in stereotaxic spaces: analyses of white matter anatomy and tract-specific
Anderson, S., Himonides, E., Wise, K., Welch, G., Stewart, L., 2012. Is there potential for quantification. Neuroimage 39, 336–347.
learning in amusia? A study of the effect of singing intervention in congenital amusia. Hyde, K.L., Zatorre, R.J., Peretz, I., 2011. Functional MRI evidence of an abnormal neural

112
A.J. Sihvonen, et al. Neuroscience and Biobehavioral Reviews 107 (2019) 104–114

network for pitch processing in congenital amusia. Cereb. Cortex 21, 292–299. Münte, T.F., Schuppert, M., Johannes, S., Wieringa, B.M., Kohlmetz, C., Altenmüller, E.,
Hyde, K.L., Peretz, I., Zatorre, R.J., 2008. Evidence for the role of the right auditory cortex 1998. Brain potentials in patients with music perception deficits: evidence for an
in fine pitch resolution. Neuropsychologia 46, 632–639. early locus. Neurosci. Lett. 256, 85–88.
Hyde, K.L., Zatorre, R.J., Griffiths, T.D., Lerch, J.P., Peretz, I., 2006. Morphometry of the Musso, M., Weiller, C., Horn, A., Glauche, V., Umarova, R., Hennig, J., et al., 2015. A
amusic brain: a two-site study. Brain 129, 2562–2570. single dual-stream framework for syntactic computations in music and language.
Hyde, K.L., Lerch, J.P., Zatorre, R.J., Griffiths, T.D., Evans, A.C., Peretz, I., 2007. Cortical Neuroimage 117, 267–283.
thickness in congenital Amusia: when less is better than more. J. Neurosci. 27, Norman-Haignere, S., Kanwisher, N.G., McDermott, J.H., 2015. Distinct cortical pathways
13028–13032. for music and speech revealed by Hypothesis-Free voxel decomposition. Neuron 88,
Ivanova, M.V., Isaev, D.Y., Dragoy, O.V., Akinina, Y.S., Petrushevskiy, A.G., Fedina, O.N., 1281–1296.
et al., 2016. Diffusion-tensor imaging of major white matter tracts and their role in Norman-Haignere, S.V., Albouy, P., Caclin, A., McDermott, J.H., Kanwisher, N.G.,
language processing in aphasia. Cortex 85, 165–181. Tillmann, B., 2016. Pitch-responsive cortical regions in congenital Amusia. J.
Jafari, Z., Esmaili, M., Delbari, A., Mehrpour, M., Mohajerani, M.H., 2017. Post-stroke Neurosci. 36, 2986–2994.
acquired amusia: a comparison between right- and left-brain hemispheric damages. Omigie, D., Müllensiefen, D., Stewart, L., 2012. The experience of music in congenital
NeuroRehabilitation 40, 233–241. Amusia. Music Percept. 30, 1–18.
Jamison, H.L., Watkins, K.E., Bishop, D.V., Matthews, P.M., 2006. Hemispheric specia- Özdemir, E., Norton, A., Schlaug, G., 2006. Shared and distinct neural correlates of
lization for processing auditory nonspeech stimuli. Cereb. Cortex 16, 1266–1275. singing and speaking. Neuroimage 33, 628–635.
Jerde, T.A., Childs, S.K., Handy, S.T., Nagode, J.C., Pardo, J.V., 2011. Dissociable systems Patel, A.D., 2003. Language, music, syntax and the brain. Nat. Neurosci. 6, 674–681.
of working memory for rhythm and melody. Neuroimage 57, 1572–1579. Perani, D., Saccuman, M.C., Scifo, P., Anwander, A., Spada, D., Baldoli, C., et al., 2011.
Johannes, S., Jobges, M.E., Dengler, R., Münte, T.F., 1998. Cortical auditory disorders: a Neural language networks at birth. Proc. Natl. Acad. Sci. U. S. A. 108, 16056–16061.
case of non-verbal disturbances assessed with event-related brain potentials. Behav. Peretz, I., 2016. Neurobiology of congenital Amusia. Trends Cogn. Sci. 20, 857–867.
Neurol. 11, 55–73. Peretz, I., 1990. Processing of local and global musical information by unilateral brain-
Koelsch, S., 2014. Brain correlates of music-evoked emotions. Nat. Rev. Neurosci. 15, damaged patients. Brain 113 (Pt 4), 1185–1205.
170–180. Peretz, I., Coltheart, M., 2003. Modularity of music processing. Nat. Neurosci. 6,
Koelsch, S., 2011. Toward a neural basis of music perception - a review and updated 688–691.
model. Front. Psychol. 2, 110. Peretz, I., Brattico, E., Tervaniemi, M., 2005. Abnormal electrical brain responses to pitch
Koelsch, S., Siebel, W.A., 2005. Towards a neural basis of music perception. Trends Cogn. in congenital amusia. Ann. Neurol. 58, 478–482.
Sci. (Regul. Ed.) 9, 578–584. Peretz, I., Champod, A.S., Hyde, K., 2003. Varieties of musical disorders. The montreal
Kohlmetz, C., Altenmüller, E., Schuppert, M., Wieringa, B.M., Münte, T.F., 2001. Deficit in battery of evaluation of Amusia. Ann. N. Y. Acad. Sci. 999, 58–75.
automatic sound-change detection may underlie some music perception deficits after Peretz, I., Belleville, S., Fontaine, S., 1997. Dissociations between music and language
acute hemispheric stroke. Neuropsychologia 39, 1121–1124. functions after cerebral resection: a new case of amusia without aphasia. Can. J. Exp.
Kummerer, D., Hartwigsen, G., Kellmeyer, P., Glauche, V., Mader, I., Kloppel, S., et al., Psychol. 51, 354–368.
2013. Damage to ventral and dorsal language pathways in acute aphasia. Brain 136, Peretz, I., Brattico, E., Järvenpää, M., Tervaniemi, M., 2009. The amusic brain: in tune,
619–629. out of key, and unaware. Brain 132, 1277–1286.
LaCroix, A.N., Diaz, A.F., Rogalsky, C., 2015. The relationship between the neural com- Piccirilli, M., Sciarma, T., Luzzi, S., 2000. Modularity of music: evidence from a case of
putations for speech and music perception is context-dependent: an activation like- pure amusia. J. Neurol. Neurosurg. Psychiatry 69, 541–545.
lihood estimate study. Front. Psychol. 6, 1138. Price, C.J., Friston, K.J., 2002. Degeneracy and cognitive anatomy. Trends Cogn. Sci. 6,
Leveque, Y., Teyssier, P., Bouchet, P., Bigand, E., Caclin, A., Tillmann, B., 2018. Musical 416–421.
emotions in congenital amusia: impaired recognition, but preserved emotional in- Rauschecker, J.P., 2014. Is there a tape recorder in your head? How the brain stores and
tensity. Neuropsychology 32, 880–894. retrieves musical melodies. Front. Syst. Neurosci. 8, 149.
Leveque, Y., Fauvel, B., Groussard, M., Caclin, A., Albouy, P., Platel, H., et al., 2016. Rauschecker, J.P., Scott, S.K., 2009. Maps and streams in the auditory cortex: nonhuman
Altered intrinsic connectivity of the auditory cortex in congenital amusia. J. primates illuminate human speech processing. Nat. Neurosci. 12, 718–724.
Neurophysiol. 116, 88–97. Rauschecker, J.P., Tian, B., 2000. Mechanisms and streams for processing of "what" and
Lévêque, Y., Schön, D., 2015. Modulation of the motor cortex during singing-voice per- "where" in auditory cortex. Proc. Natl. Acad. Sci. U. S. A. 97, 11800–11806.
ception. Neuropsychologia 70, 58–63. Rogalsky, C., Rong, F., Saberi, K., Hickok, G., 2011. Functional anatomy of language and
Liu, F., Patel, A.D., Fourcin, A., Stewart, L., 2010. Intonation processing in congenital music perception: temporal and structural factors investigated using functional
amusia: discrimination, identification and imitation. Brain 133, 1682–1693. magnetic resonance imaging. J. Neurosci. 31, 3843–3852.
Lopez-Barroso, D., de Diego-Balaguer, R., Cunillera, T., Camara, E., Münte, T.F., Rorden, C., Karnath, H.O., 2004. Using human brain lesions to infer function: a relic from
Rodriguez-Fornells, A., 2011. Language learning under working memory constraints a past era in the fMRI age? Nat. Rev. Neurosci. 5, 813–819.
correlates with microstructural differences in the ventral language pathway. Cereb. Ross, E.D., 2010. Cerebral localization of functions and the neurology of language: fact
Cortex 21, 2742–2750. versus fiction or is it something else? Neuroscientist 16, 222–243.
Loui, P., 2015. A dual-stream neuroanatomy of singing. Music Percept. 32, 232–241. Ross, E.D., Monnot, M., 2008. Neurology of affective prosody and its functional-anatomic
Loui, P., Alsop, D., Schlaug, G., 2009. Tone deafness: a new disconnection syndrome? J. organization in right hemisphere. Brain Lang. 104, 51–74.
Neurosci. 29, 10215–10220. Royal, I., Vuvan, D.T., Zendel, B.R., Robitaille, N., Schonwiesner, M., Peretz, I., 2016.
Loui, P., Guenther, F.H., Mathys, C., Schlaug, G., 2008. Action-perception mismatch in Activation in the right inferior parietal lobule reflects the representation of musical
tone-deafness. Curr. Biol. 18, R331–2. structure beyond simple pitch discrimination. PLoS One 11, e0155291.
Loui, P., Patterson, S., Sachs, M.E., Leung, Y., Zeng, T., Przysinda, E., 2017. White matter Sammler, D., Grosbras, M.H., Anwander, A., Bestelmeyer, P.E., Belin, P., 2015. Dorsal and
correlates of musical anhedonia: implications for evolution of music. Front. Psychol. ventral pathways for prosody. Curr. Biol. 25, 3079–3085.
8, 1664. Sammler, D., Cunitz, K., Gierhan, S.M.E., Anwander, A., Adermann, J., Meixensberger, J.,
Maess, B., Koelsch, S., Gunter, T.C., Friederici, A.D., 2001. Musical syntax is processed in et al., 2018. White matter pathways for prosodic structure building: a case study.
Broca’s area: an MEG study. Nat. Neurosci. 4, 540–545. Brain Lang. 183, 1–10.
Makris, N., Goldstein, J.M., Kennedy, D., Hodge, S.M., Caviness, V.S., Faraone, S.V., et al., Samson, F., Zeffiro, T.A., Toussaint, A., Belin, P., 2011. Stimulus complexity and cate-
2006. Decreased volume of left and total anterior insular lobule in schizophrenia. gorical effects in human auditory cortex: an activation likelihood estimation meta-
Schizophr. Res. 83, 155–171. analysis. Front. Psychol. 1, 241.
Mandell, J., Schulze, K., Schlaug, G., 2007. Congenital amusia: an auditory-motor feed- Samson, Y., Belin, P., Zilbovicius, M., Remy, P., van Eeckhout, P., Rancurel, G., 1999.
back disorder? Restor. Neurol. Neurosci. 25, 323–334. Mechanisms of aphasia recovery and brain imaging. Rev. Neurol. (Paris) 155,
Martinez-Molina, N., Mas-Herrero, E., Rodriguez-Fornells, A., Zatorre, R.J., Marco- 725–730.
Pallares, J., 2019. White matter microstructure reflects individual differences in Särkämö, T., Tervaniemi, M., Soinila, S., Autti, T., Silvennoinen, H.M., Laine, M., et al.,
music reward sensitivity. J. Neurosci. 39, 5018–5027. 2009. Cognitive deficits associated with acquired amusia after stroke: a neu-
Martinez-Molina, N., Mas-Herrero, E., Rodriguez-Fornells, A., Zatorre, R.J., Marco- ropsychological follow-up study. Neuropsychologia 47, 2642–2651.
Pallares, J., 2016. Neural correlates of specific musical anhedonia. Proc. Natl. Acad. Särkämö, T., Tervaniemi, M., Soinila, S., Autti, T., Silvennoinen, H.M., Laine, M., et al.,
Sci. U. S. A. 113, E7337–E7345. 2010. Auditory and cognitive deficits associated with acquired amusia after stroke: a
Mas-Herrero, E., Marco-Pallares, J., Lorenzo-Seva, U., Zatorre, R.J., Rodriguez-Fornells, magnetoencephalography and neuropsychological follow-up study. PLoS One 5,
A., 2013. Individual differences in music reward experiences. Music Perception 31, e15157.
118–138. Satoh, M., Nakase, T., Nagata, K., Tomimoto, H., 2011. Musical anhedonia: selective loss
Mazzoni, M., Moretti, P., Pardossi, L., Vista, M., Muratorio, A., Puglioli, M., 1993. A case of emotional experience in listening to music. Neurocase 17, 410–417.
of music imperception. J. Neurol. Neurosurg. Psychiatry 56, 322. Satoh, M., Kato, N., Tabei, K.I., Nakano, C., Abe, M., Fujita, R., et al., 2016. A case of
Mazzucchi, A., Marchini, C., Budai, R., Parma, M., 1982. A case of receptive amusia with musical anhedonia due to right putaminal hemorrhage: a disconnection syndrome
prominent timbre perception defect. J. Neurol. Neurosurg. Psychiatry 45, 644–647. between the auditory cortex and insula. Neurocase 22, 518–525.
Mendez, M.F., 2001. Generalized auditory agnosia with spared music recognition in a left- Saur, D., Lange, R., Baumgaertner, A., Schraknepper, V., Willmes, K., Rijntjes, M., et al.,
hander. Analysis of a case with a right temporal stroke. Cortex 37, 139–150. 2006. Dynamics of language reorganization after stroke. Brain 129, 1371–1384.
Mirman, D., Chen, Q., Zhang, Y., Wang, Z., Faseyitan, O.K., Coslett, H.B., et al., 2015. Schlaug, G., Marchina, S., Norton, A., 2009. Evidence for plasticity in white-matter tracts
Neural organization of spoken language revealed by lesion-symptom mapping. Nat. of patients with chronic Broca’s aphasia undergoing intense intonation-based speech
Commun. 6, 6762. therapy. Ann. N. Y. Acad. Sci. 1169, 385–394.
Mitchell, R.L., Ross, E.D., 2008. fMRI evidence for the effect of verbal complexity on Schlaug, G., Norton, A., Marchina, S., Zipse, L., Wan, C.Y., 2010. From singing to
lateralisation of the neural response associated with decoding prosodic emotion. speaking: facilitating recovery from nonfluent aphasia. Future Neurol. 5, 657–665.
Neuropsychologia 46, 2880–2887. Schmithorst, V.J., 2005. Separate cortical networks involved in music perception:

113
A.J. Sihvonen, et al. Neuroscience and Biobehavioral Reviews 107 (2019) 104–114

preliminary functional MRI evidence for modularity of music processing. Neuroimage 589–605.
25, 444–451. Tillmann, B., Koelsch, S., Escoffier, N., Bigand, E., Lalitte, P., Friederici, A.D., et al., 2006.
Schuppert, M., Münte, T.F., Wieringa, B.M., Altenmüller, E., 2000. Receptive amusia: Cognitive priming in sung and instrumental music: activation of inferior frontal
evidence for cross-hemispheric neural networks underlying music processing strate- cortex. Neuroimage 31, 1771–1782.
gies. Brain 123 (Pt 3), 546–559. Toiviainen, P., Alluri, V., Brattico, E., Wallentin, M., Vuust, P., 2014. Capturing the
Sidtis, J.J., Volpe, B.T., 1988. Selective loss of complex-pitch or speech discrimination musical brain with Lasso: dynamic decoding of musical features from fMRI data.
after unilateral lesion. Brain Lang. 34, 235–245. Neuroimage 88, 170–180.
Sihvonen, A.J., Ripollés, P., Rodríguez-Fornells, A., Soinila, S., Särkämö, T., 2017a. Tzortzis, C., Goldblum, M.C., Dang, M., Forette, F., Boller, F., 2000. Absence of amusia
Revisiting the neural basis of acquired amusia: structural changes underlying amusia and preserved naming of musical instruments in an aphasic composer. Cortex 36,
recovery. Front. Neurosci. 11. 227–242.
Sihvonen, A.J., Särkämö, T., Leo, V., Tervaniemi, M., Altenmüller, E., Soinila, S., 2017b. Tzourio-Mazoyer, N., Landeau, B., Papathanassiou, D., Crivello, F., Etard, O., Delcroix, N.,
Music-based interventions in neurological rehabilitation. Lancet Neurol. 16, et al., 2002. Automated anatomical labeling of activations in SPM using a macro-
648–660. scopic anatomical parcellation of the MNI MRI single-subject brain. Neuroimage 15,
Sihvonen, A.J., Ripollés, P., Leo, V., Rodríguez-Fornells, A., Soinila, S., Särkämö, T., 2016. 273–289.
Neural basis of acquired Amusia and its recovery after stroke. J. Neurosci. 36, Uetsuki, S., Kinoshita, H., Takahashi, R., Obata, S., Kakigi, T., Wada, Y., et al., 2016. A
8872–8881. case of expressive-vocal amusia in a right-handed patient with left hemispheric cer-
Sihvonen, A.J., Ripollés, P., Särkämö, T., Leo, V., Rodriguez-Fornells, A., Saunavaara, J., ebral infarction. Brain Cogn. 103, 23–29.
et al., 2017c. Tracting the neural basis of music: deficient structural connectivity van den Heuvel, M.P., Sporns, O., 2013. Network hubs in the human brain. Trends Cogn.
underlying acquired amusia. Cortex. Sci. 17, 683–696.
Sihvonen, A.J., Särkämö, T., Ripollés, P., Leo, V., Saunavaara, J., Parkkola, R., et al., Wakana, S., Caprihan, A., Panzenboeck, M.M., Fallon, J.H., Perry, M., Gollub, R.L., et al.,
2017d. Functional neural changes associated with acquired amusia across different 2007. Reproducibility of quantitative tractography methods applied to cerebral white
stages of recovery after stroke. Sci. Rep. 7 (11390) 017-11841-6. matter. Neuroimage 36, 630–644.
Sperber, C., Wiesen, D., Karnath, H.O., 2018. An empirical evaluation of multivariate Wan, C.Y., Zheng, X., Marchina, S., Norton, A., Schlaug, G., 2014. Intensive therapy in-
lesion behaviour mapping using support vector regression. Hum. Brain Mapp. duces contralateral white matter changes in chronic stroke patients with Broca’s
Stewart, L., 2008. Fractionating the musical mind: insights from congenital amusia. Curr. aphasia. Brain Lang. 136, 1–7.
Opin. Neurobiol. 18, 127–130. Warren, J.D., Jennings, A.R., Griffiths, T.D., 2005. Analysis of the spectral envelope of
Stewart, L., von Kriegstein, K., Warren, J.D., Griffiths, T.D., 2006. Music and the brain: sounds by the human brain. Neuroimage 24, 1052–1057.
disorders of musical listening. Brain 129, 2533–2553. Wilbiks, J.M., Vuvan, D.T., Girard, P.Y., Peretz, I., Russo, F.A., 2016. Effects of vocal
Tervaniemi, M., Szameitat, A.J., Kruck, S., Schroger, E., Alter, K., De Baene, W., et al., training in a musicophile with congenital amusia. Neurocase 22, 526–537.
2006. From air oscillations to music and speech: functional magnetic resonance Wildgruber, D., Ackermann, H., Kreifelts, B., Ethofer, T., 2006. Cerebral processing of
imaging evidence for fine-tuned neural networks in audition. J. Neurosci. 26, linguistic and emotional prosody: fMRI studies. Prog. Brain Res. 156, 249–268.
8647–8652. Zatorre, R.J., Salimpoor, V.N., 2013. From perception to pleasure: music and its neural
Thiebaut de Schotten, M., Dell’Acqua, F., Valabregue, R., Catani, M., 2012. Monkey to substrates. Proc. Natl. Acad. Sci. U. S. A. 110 (Suppl 2), 10430–10437.
human comparative anatomy of the frontal lobe association tracts. Cortex 48, 82–96. Zatorre, R.J., Belin, P., Penhune, V.B., 2002. Structure and function of auditory cortex:
Thiebaut de Schotten, M., Kinkingnehun, S., Delmaire, C., Lehericy, S., Duffau, H., music and speech. Trends Cogn. Sci. 6, 37–46.
Thivard, L., et al., 2008. Visualization of disconnection syndromes in humans. Cortex Zavaglia, M., Forkert, N.D., Cheng, B., Gerloff, C., Thomalla, G., Hilgetag, C.C., 2015.
44, 1097–1103. Mapping causal functional contributions derived from the clinical assessment of brain
Thompson, W.F., Marin, M.M., Stewart, L., 2012. Reduced sensitivity to emotional pro- damage after stroke. Neuroimage Clin. 9, 83–94.
sody in congenital amusia rekindles the musical protolanguage hypothesis. Proc. Zendel, B.R., Lagrois, M.E., Robitaille, N., Peretz, I., 2015. Attending to pitch information
Natl. Acad. Sci. U. S. A. 109, 19027–19032. inhibits processing of pitch information: the curious case of amusia. J. Neurosci. 35,
Tillmann, B., Albouy, P., Caclin, A., 2015. Congenital amusias. Handb. Clin. Neurol. 129, 3815–3824.

114