Brain Networks Involved in Sensory Perception in Parkinson’s Disease: A Scoping Review

Parkinson’s Disease (PD) has historically been considered a disorder of motor dysfunction. However, a growing number of studies have demonstrated sensory abnormalities in PD across the modalities of proprioceptive, tactile, visual, auditory and temporal perception. A better understanding of these may inform future drug and neuromodulation therapy. We analysed these studies using a scoping review. In total, 101 studies comprising 2853 human participants (88 studies) and 125 animals (13 studies), published between 1982 and 2022, were included. These highlighted the importance of the basal ganglia in sensory perception across all modalities, with an additional role for the integration of multiple simultaneous sensation types. Numerous studies concluded that sensory abnormalities in PD result from increased noise in the basal ganglia and increased neuronal receptive field size. There is evidence that sensory changes in PD and impaired sensorimotor integration may contribute to motor abnormalities.


Introduction
Parkinson's disease (PD) was first described by James Parkinson over 200 years ago [1] and is diagnosed by ascertaining the presence of movement abnormalities, namely bradykinesia and at least one additional symptom: tremor, muscle rigidity, or postural/balance instability not attributed to primary visual, cerebellar, or proprioceptive dysfunction [2].The core pathology is believed to originate in the basal ganglia, which is crucial for motor function, beginning with the loss of dopamine-secreting neurons in the substantia nigra (SN) pars compacta.This in turn leads to significant changes in other basal ganglia nuclei, resulting in excessive inhibition of thalamocortical and brainstem motor systems [3].Application of deep brain stimulation (DBS) to suppress globus pallidus interna (GPi) or subthalamic nucleus (STN) activity has been demonstrated to alleviate PD motor symptoms and enhance quality of life [4].
People with Parkinson's Disease (PwPD) typically experience a decline in independence in daily activities around 7 years post-diagnosis [5].This initial decline in function is not attributed to a loss of motor response to levodopa, nor motor fluctuations [6], but may be at least partially ascribed to non-motor mechanisms that gradually accumulate and contribute to the burden of disease over time.
The pathology in PD has historically been associated predominantly with motor domain dysfunction; however, there is a growing body of evidence that abnormalities in sensory function also occur e.g., [7][8][9][10][11][12].Sensory abnormalities occur across a range of modalities, including proprioceptive, tactile, visual, auditory and temporal perception functions.These contribute directly to functional disability in PwPD [13].Furthermore, Summary images of activated brain regions were generated solely from controlled for confounding by motor activation or other sensory modalities areas activated by a sensory stimulus and replicated across two or more s included in the summary images.An exception was made when no brain particular sensory stimulus were found to be activated across multiple stu cases, areas activated by a particular sensory stimulus in a single study were the summary images.

Study Selection
The implemented search strategy yielded 776 studies for screening (Figu which met the inclusion criteria.After full-text review, 101 articles were inclusion in the review, all being cross-sectional in nature.

Breakdown of Sample Data
The included articles comprised 89 human and 12 animal studies, publis 1982 and 2022.Animal studies included rats, cats, and monkeys.In total,

Breakdown of Sample Data
The included articles comprised 89 human and 12 animal studies, published between 1982 and 2022.Animal studies included rats, cats, and monkeys.In total, 2853 human subjects and 125 animals were studied.Specific details on the divisions between healthy control (HC) vs. PD, animals vs. human, and the brain interrogation modality used are reported in Figures 2 and 3, and the Supplementary Materials.subjects and 125 animals were studied.Specific details on the divisions between healthy control (HC) vs. PD, animals vs. human, and the brain interrogation modality used are reported in Figures 2 and 3, and the Supplementary Materials.Forty-eight studies evaluated the brain areas activated during sensory stimulation.Only one study assessed multisensory integration [20].Thirty studies investigated evoked potentials and were not included in the assessment of localization of sensory response due to the mixed proprioceptive-tactile nature of the stimulus.Thirteen studies assessed the somatotopic organisation of the basal ganglia.Seven studies evaluated sensorimotor integration and nine studies investigated increased noise or decreased neuronal activation specificity in the basal ganglia in PD.See the Supplementary Materials for further details of the sample characteristics.
In the review of the literature presented below, studies were undertaken in humans unless otherwise specified.Forty-eight studies evaluated the brain areas activated during sensory stimulation.Only one study assessed multisensory integration [20].Thirty studies investigated evoked potentials and were not included in the assessment of localization of sensory response due to the mixed proprioceptive-tactile nature of the stimulus.Thirteen studies assessed the somatotopic organisation of the basal ganglia.Seven studies evaluated sensorimotor integration and nine studies investigated increased noise or decreased neuronal activation specificity in the basal ganglia in PD.See the Supplementary Materials for further details of the sample characteristics.
In the review of the literature presented below, studies were undertaken in humans unless otherwise specified.

Sensory Testing Methodologies
Human sensory testing methods were as follows: (i) Temporal sensory-system testing utilized temporal (tactile and auditory) discrimination thresholds, temporal order judgement, temporal (auditory and visual) duration discrimination and time reproduction tasks.(ii) Auditory sensory-system testing utilized frequency discrimination (deviant tones) and spatial discrimination (localizing site of tone).(iii) Visual sensory-system testing utilized light intensity discrimination, pattern (moving or static) discrimination, shape recognition, face (emotion and gender) recognition and flickering checkerboard.(iv) Tactile system testing utilized tactile shape discrimination, tactile roughness discrimination, tactile grating orientation discrimination, tactile amplitude (of ring electrode impulse) discrimination, tactile spatial discrimination (defining site of arm stimulated), tactile vibratory-frequency discrimination.(v) Proprioception was stimulated with passive movement of a joint; however, the acuity of proprioception was not tested in the studies which were identified by our search strategy.

Sensory Testing Methodologies
Human sensory testing methods were as follows: (i) Temporal sensory-system testing utilized temporal (tactile and auditory) discrimination thresholds, temporal order judgement, temporal (auditory and visual) duration discrimination and time reproduction tasks.(ii) Auditory sensory-system testing utilized frequency discrimination (deviant tones) and spatial discrimination (localizing site of tone).(iii) Visual sensory-system testing utilized light intensity discrimination, pattern (moving or static) discrimination, shape recognition, face (emotion and gender) recognition and flickering checkerboard.(iv) Tactile system testing utilized tactile shape discrimination, tactile roughness discrimination, tactile grating orientation discrimination, tactile amplitude (of ring electrode impulse) discrimination, tactile spatial discrimination (defining site of arm stimulated), tactile vibratory-frequency discrimination.(v) Proprioception was stimulated with passive movement of a joint; however, the acuity of proprioception was not tested in the studies which were identified by our search strategy.
Animal sensory systems were stimulated with light spots on a high contrast background, moving vertical gratings (visual), monaural clicks or white noise (auditory), whisker deflection, light touch to shaved skin, brushing of hair, punctate stimulation (tactile), passive joint movement (proprioception) and muscle palpation/tendon tap (mixed tactile and proprioception).

What Brain Areas Are Involved in Sensory Processing?
We present the evidence from the literature for each sensory modality, as follows: temporal processing, auditory, visual, tactile and proprioception.The findings are subcategorised as evidence from healthy controls (HC) first and then from PwPD.The putamen, posterior parietal cortex (PPC), supplementary motor area (SMA), temporal gyri, pre-SMA and anterior cingulate cortex (ACC) are active in temporal perception (the perception of the magnitude of time interval) (Figure 4).The SMA is involved in tactile [21,22], visual [23] and auditory [24] forms of temporal discrimination.The pre-supplementary motor area (pre-SMA) is shown to be activated in tactile [21,25] and auditory [26] forms of temporal discrimination.The temporal lobe (including the superior temporal gyrus and temporal gyri) is involved in tactile [21,27] and visual [23] forms of temporal discrimination.The PPC is involved in visual and auditory [28] forms of temporal discrimination.The PPC is also involved in tactile forms of temporal discrimination [22,29] although neither of these studies controlled for brain-area activation due to the tactile aspect of the response.However, in light of the Bueti et al. [28] finding that found PPC involvement during temporal discrimination assessed by two different sensory modalities (visual and auditory), these results are likely reflective of temporal processing.

What Brain Areas Are Involved in Sensory Processing?
We present the evidence from the literature for each sensory modality, as follows: temporal processing, auditory, visual, tactile and proprioception.The findings are subcategorised as evidence from healthy controls (HC) first and then from PwPD.

What Brain Areas Are Involved in Temporal Processing in Healthy Cohorts?
The putamen, posterior parietal cortex (PPC), supplementary motor area (SMA), temporal gyri, pre-SMA and anterior cingulate cortex (ACC) are active in temporal perception (the perception of the magnitude of time interval) (Figure 4).The SMA is involved in tactile [21,22], visual [23] and auditory [24] forms of temporal discrimination.The pre-supplementary motor area (pre-SMA) is shown to be activated in tactile [21,25] and auditory [26] forms of temporal discrimination.The temporal lobe (including the superior temporal gyrus and temporal gyri) is involved in tactile [21,27] and visual [23] forms of temporal discrimination.The PPC is involved in visual and auditory [28] forms of temporal discrimination.The PPC is also involved in tactile forms of temporal discrimination [22,29] although neither of these studies controlled for brain-area activation due to the tactile aspect of the response.However, in light of the Bueti et al. [28] finding that found PPC involvement during temporal discrimination assessed by two different sensory modalities (visual and auditory), these results are likely reflective of temporal processing.All of these areas demonstrate bilateral involvement in temporal processing [21,[23][24][25]29], although some studies have shown that only the ipsilateral or contralateral side was active [21,[26][27][28].
The putamen is also involved in temporal discrimination [21,22,30] and duration discrimination [23].Additionally, the putamen is involved in temporal processing in conjunction with auditory [30], visual [23] and tactile [21] stimulation.Nenadic [30], Ferandez [23] and Huang [21] compared temporal with other types of discrimination in one sensory modality, and found that the putamen was specifically involved in temporal processing.

What Brain Areas Are Found to Function Abnormally in Temporal Processing in
People with Parkinson's Disease?Koch [31] found that repetitive TMS applied to the dorsolateral prefrontal cortex (DLPFC) could improve time perception in PD but not HCs (Figure 4).Repetitive TMS to the SMA had no significant effect for PwPD or controls.Elsinger [24] found decreased SMA activity during a pace-continuation condition in PwPD off medication, but not on medication, when compared to HCs.The pace-continuation condition required patients to continue tapping their finger at a pace previously demonstrated with auditory tones after cessation of the auditory guide.Dusek [32] found that PwPD on medication, compared to PwPD off medication, had increased bilateral precuneus activation during retrieval of a remembered interval duration (temporal estimation), which correlated with a trend to improved performance in PwPD when on medication, suggesting the importance of the precuneus in temporal estimation in PD.Elsinger [24] also found decreased putamen activity in PwPD off medication, compared to HCs, during a pace-continuation condition that may (as mentioned above) either reflect pure temporal processing or memory retrieval.
Studies exploring brain connectivity in non-PD cohorts have provided evidence of a network of temporal perception, which may help identify a neuropathological substrate underlying abnormal temporal perception in PD cohorts.Lacruz [22] noted functional impacts of lesion locations in the brain and through this noted the main cortical output from the basal ganglia is to the SMA.The SMA then projects heavily to the superior parietal lobe, with the SMA [21][22][23][24] and PPC [22,28,29] noted to be involved in temporal processing.In a healthy cohort study, Pastor [25] noted that the pre-SMA receives input from the basal ganglia and cerebellum and sends output to the striatum and STN.Also in a healthy cohort, Huang [21] noted that ability in the somatosensory temporal-discrimination task was predicted not only by activity in the pre-SMA, ACC and dorsal putamen, but by the connection between these areas.The reduced temporal-processing ability in PD [24,31,33,34], the diseased state of the putamen [35] in PD and the abnormal SMA and putaminal activity [24] noted during temporal-perception testing in a PD cohort, provide further evidence for this putamen, SMA, pre-SMA and ACC network of temporal perception.

What Brain Areas Are Involved in Auditory Processing in Healthy Cohorts?
Multiple studies have found that the bilateral superior temporal gyri [24,30] and ACC (left or right) [26,30] are important in auditory processing (Figure 5).Pastor [26] found the right head of caudate and putamen are active in both auditory spatial and auditory temporal discrimination.This finding is supported by the animal study by Nagy [20] that found 6% of neurons in the caudate tested in cats responded to auditory stimuli on MER.Nagy et al. also found that SN neurons also respond to auditory stimuli.Rothblat [36] found, via microelectric recording in cats, that 22% of globus pallidus externa (GPe) and 8% of GPi neurons responded to auditory stimulation.Pekkonen [37] and Philipova [38] found a reduced amplitude N1 and frontal mismatch negativity (MMN) or P3 response on an electroencephalogram (EEG) during auditory frequency discrimination in PwPD (Figure 5).The N1 response is thought to originate in the superior temporal auditory cortex during auditory stimulus detection.The frontal MMN and P3 responses are thought to reflect auditory stimulus evaluation and discrimination.Rossi [39] found reduced amplitude and prolonged latencies of auditory evoked potentials in PwPD, when compared to HCs, but Weise [40] found no difference.With regards to basal ganglia changes in PwPD in auditory processing, Rothblat [36] found on MER that GPe response to auditory stimulation reduced from 22% to 5%, and GPi response reduced from 8% to 4%.Pekkonen [37] and Philipova [38] found a reduced amplitude N1 and frontal mismatch negativity (MMN) or P3 response on an electroencephalogram (EEG) during auditory frequency discrimination in PwPD (Figure 5).The N1 response is thought to originate in the superior temporal auditory cortex during auditory stimulus detection.The frontal MMN and P3 responses are thought to reflect auditory stimulus evaluation and discrimination.Rossi [39] found reduced amplitude and prolonged latencies of auditory evoked potentials Brain Sci.2023, 13, 1552 8 of 19 in PwPD, when compared to HCs, but Weise [40] found no difference.With regards to basal ganglia changes in PwPD in auditory processing, Rothblat [36] found on MER that GPe response to auditory stimulation reduced from 22% to 5%, and GPi response reduced from 8% to 4%.Several studies found occipital activity during visual discrimination tasks [23,41] (Figure 6).Schmiedt et al. additionally found when assessing the event-related potential response that occipital excitability was early, and likely related to low-level visual processing.Ferrandez also found evidence of caudate, putamen and SN/red nucleus involvement on fMRI during visual discrimination tasks.Ferrandez hypothesized that the putamen and caudate were active due to the demand of making sequential decisions, and the SN/red nucleus was responsible for linking other important structures together.However, animal studies suggest that the basal ganglia activation may be directly induced by visual stimuli.Nagy [20] found in MER in cats that 17% of caudate neurons tested responded to visual (light) stimuli.There was a similar pattern in the SN.Rothblat [36] found 16% of GPe equivalent neurons to respond to visual light stimulation, also via MER in cats, and 11% of GPi equivalent neurons to do the same.
originate in the superior temporal auditory cortex during auditory stimulus d The frontal MMN and P3 responses are thought to reflect auditory stimulus ev and discrimination.Rossi [39] found reduced amplitude and prolonged late auditory evoked potentials in PwPD, when compared to HCs, but Weise [40] f difference.With regards to basal ganglia changes in PwPD in auditory pro Rothblat [36] found on MER that GPe response to auditory stimulation reduced fr to 5%, and GPi response reduced from 8% to 4%.

What Brain Areas Are Involved in Visual Processing in Healthy Cohorts?
Several studies found occipital activity during visual discrimination task (Figure 6).Schmiedt et al. additionally found when assessing the event-related p response that occipital excitability was early, and likely related to low-leve processing.Ferrandez also found evidence of caudate, putamen and SN/red involvement on fMRI during visual discrimination tasks.Ferrandez hypothesized putamen and caudate were active due to the demand of making sequential decisi the SN/red nucleus was responsible for linking other important structures t However, animal studies suggest that the basal ganglia activation may be directly by visual stimuli.Nagy [20] found in MER in cats that 17% of caudate neuron responded to visual (light) stimuli.There was a similar pattern in the SN.Roth found 16% of GPe equivalent neurons to respond to visual light stimulation, also in cats, and 11% of GPi equivalent neurons to do the same.During visual stimulation tasks, there is reduced visual cortex activity in PD compared to HCs [42][43][44][45] (Figure 6).There is additionally pattern electroretinogram evidence of retinal dysfunction during visual stimulus in PD [42,44,46,47].Rothblat [36] additionally demonstrated reduced GPi and GPe microelectrode responses to visual stimuli in an 1-methyl-4-phenyl-propionoxy-piperidine (MPTP) model of PD in cats.

What Brain Areas Are Involved in Tactile Processing in Healthy Cohorts?
Unsurprisingly, a number of studies have demonstrated the importance of the contralateral primary sensory cortex (S1) in tactile sensation processing [25,27,[48][49][50] (Figure 7).Mowery [51] also demonstrated the importance of the contralateral S1 in a non-PD animal model with MER during whisker deflection in rats.Pastor additionally found S1 and the inferior parietal lobule (IPL) activation common to both tactile spatial and tactile temporal  Pastor [25] found bilateral basal ganglia nuclei activity in both tactile spatial and tactile temporal discrimination, including the caudate and SN.This is supported b animal models looking at MER response to tactile stimulation in the basal ganglia Mowery [51] and Nagy [20] found caudate responsiveness to tactile stimuli.Nagy als noted a similar response in the SN.The DeLong [53] and Rothblat [36] animal studies als noted neuronal responses in the GPi and GPe to tactile stimulation.Pastor [25] found bilateral basal ganglia nuclei activity in both tactile spatial and tactile temporal discrimination, including the caudate and SN.This is supported by animal models looking at MER response to tactile stimulation in the basal ganglia.Mowery [51] and Nagy [20] found caudate responsiveness to tactile stimuli.Nagy also noted a similar response in the SN.The DeLong [53] and Rothblat [36] animal studies also noted neuronal responses in the GPi and GPe to tactile stimulation.Weder [54] and Zhao [52] revealed reduced contralateral S1, bilateral parietal lobe and contralateral or bilateral premotor area activity during tactile spatial discrimination tasks (Figure 7).These findings were supported by Palomar [48] who was able to improve tactile amplitude discrimination with paired-pulse transcranial magnetic stimulation (ppTMS) to the contralateral S1 in HCs, but the same benefit was not found in PwPD on medication.The finding was not significant in patients off medication.In animal study MER, Rothblat [36] demonstrated decreased globus pallidus neuronal function to tactile stimuli, as GPe-equivalent neurons reduced from 31.4 to 12.2% response, and GPi-equivalent neurons reduced from 29% to 13% response.

What Brain Areas Are Involved in Proprioceptive Processing in Healthy Cohorts?
Kalmar [35] and Boecker [55] demonstrated contralateral S1 activity during proprioceptive stimulation, with Boecker showing this extended to the contralateral S2, and Kalmar also revealing activity in the contralateral primary motor cortex (M1) and SMA (Figure 8).In an HC population, Kalmar found the contralateral putamen to activate on fMRI in response to proprioceptive stimuli, while the Boecker study revealed contralateral globus pallidus activation.Non-PD animal studies [53,56] have revealed neuronal activation on MER in the GPe, GPi, and STN in response to proprioceptive stimuli of the face, arm and leg.
Brain Sci.2023, 13,1552 proprioceptive stimulation, with Boecker showing this extended to the contralateral and Kalmar also revealing activity in the contralateral primary motor cortex (M1) SMA (Figure 8).In an HC population, Kalmar found the contralateral putamen to acti on fMRI in response to proprioceptive stimuli, while the Boecker study revea contralateral globus pallidus activation.Non-PD animal studies [53,56] have revea neuronal activation on MER in the GPe, GPi, and STN in response to propriocep stimuli of the face, arm and leg.Seiss [57] found an early evoked EEG response to proprioception, attributed focus in the primary motor cortex, was equivalent between PwPD and HCs, but a l response, attributed to a source in the S1 cortex, was of opposite polarity in PwPD.Th supported by Boecker's [55] findings that the contralateral S1 had decreased activity in patients compared with controls (Figure 8), although Boekcer also found that contralateral M1, lateral premotor, S2 and posterior cingulate had decreased activit PwPD.
A number of studies have demonstrated that the GP [58,59], STN [60][61][62][63][64] pedunculopontine nucleus [65] are responsive to proprioception in PD cohorts, altho without comparison to HCs to assess for effects of the disease state.Stefani demonstrated in PwPD that there is an increase in the contralateral STN neuronal fi to passive movement, with this firing rate reduced by apomorphine.Boecker [55] fo Seiss [57] found an early evoked EEG response to proprioception, attributed to a focus in the primary motor cortex, was equivalent between PwPD and HCs, but a later response, attributed to a source in the S1 cortex, was of opposite polarity in PwPD.This is supported by Boecker's [55] findings that the contralateral S1 had decreased activity in PD patients compared with controls (Figure 8), although Boekcer also found that the contralateral M1, lateral premotor, S2 and posterior cingulate had decreased activity in PwPD.
A number of studies have demonstrated that the GP [58,59], STN [60][61][62][63][64] and pedunculopontine nucleus [65] are responsive to proprioception in PD cohorts, although without comparison to HCs to assess for effects of the disease state.Stefani [66] demonstrated in PwPD that there is an increase in the contralateral STN neuronal firing to passive movement, with this firing rate reduced by apomorphine.Boecker [55] found the contralateral GP to have decreased activation in a PD cohort during proprioceptive stimulation, with the contralateral putamen demonstrating a trend towards the same effect.Kalmar [35] demonstrated an increase in ipsilateral putamen activity only in right-handed PwPD, which may suggest a diseased state of the contralateral putamen.

The Breadth of Sensory Response and Multisensory Response of the Basal Ganglia
DeLong [53] looked at MER in HC monkeys in the GPe, GPi and STN after passive joint movement (proprioception > tactile), joint palpation (tactile and possibly proprioception), muscle palpation (tactile and possibly proprioception), tendon taps (tactile and proprioception), body hair stimulation (thought to be purely tactile stimulation) and light touch (tactile stimulation alone).All of the tested basal ganglia nuclei had by far the greatest activity during passive joint movement, which may be thought as the closest representation of proprioception.In the same study, MER revealed no response to a light touch, supporting an argument that the basal ganglia specifically detect proprioception, and indirectly support motor function through this mechanism.Low-magnitude responses in these nuclei were recorded during joint palpation, muscle palpation and tendon taps.This response may reflect low-magnitude proprioceptive responses with these forms of stimulation.Similarly in HC humans, Kamar [35] demonstrated putaminal fMRI activation during passive joint movement and Boecker [55] demonstrated globus pallidus PET activation to high-frequency vibrational stimulus to a joint.There are additionally numerous studies investigating proprioceptive responses in the basal ganglia with MER in PD humans, which were conducted in the setting of DBS surgeries.Stefani [66] found contralateral STN activation on MER in PD humans during proprioceptive stimuli.Additionally, there are many studies investigating the somatotopic organization of the STN in PD humans to proprioceptive stimuli [58][59][60][61][62][63][64][67][68][69].The low-magnitude activation during hair stimulation reported by DeLong and colleagues [53], however, is difficult to rationalize as the result of activation of proprioceptive afferents, and therefore this finding may argue towards the ability of the basal ganglia to also detect non-proprioceptive sensations.
Pastor [25] demonstrated with fMRI a tactile response at the head of caudate, STN and SN, and an auditory response at the head of caudate and putamen in human HCs.Additionally, studies have demonstrated basal ganglia responses to various sensory stimuli via MER in HC animals [20,36,51,53], supportive of the notion that the basal ganglia are involved in non-proprioceptive sensory perception (hair stimulation, visual, tactile, auditory stimuli).Pesenti [70], in a human study, however, failed to detect any component of the P100 visual-evoked potential response via MER at the STN in a PD cohort, although did not include a comparison to HCs leaving it unclear if this was due to the impact of PD or merely a lack of STN activity with visual stimulus.Only one study demonstrated the multisensory response of basal ganglia neurons.Nagy [20] examined neuronal responses in the caudate and SN to visual, auditory and tactile stimulation in HC cats.Fascinatingly, Nagy found that in the caudate, more neurons responded to more than one sensory modality than to a single modality.This single neuronal response to multisensory modalities was greater than the additive response of the different modalities alone, suggesting a preference of these neurons for a multisensory response.In addition, latencies to multimodal sensory responses were actually shorter than to unimodal responses.The SN also has a multi-sensory function, although with a higher proportion of neurons responding to one modality only.Rothblat et al. [36] also provided some insight into changes that may occur to the multi-sensory response in the basal ganglia in PD.They found, in an MPTP model of PD in cats, there was increased multi-sensory response in the GPe and GPi, in that more neurons responded to multiple sensory types than before the induction of Parkinsonism.

Discussion
This review synthesizes findings from 101 studies conducted between 1982 and 2022, involving 2853 humans and 137 animals, addressing key questions regarding sensory processing abnormalities in PD.In observing the active brain areas in the visual, auditory, tactile, proprioceptive and temporal senses, and the abnormalities in PD, a hypothesis emerges that PwPD demonstrate sensory-processing abnormalities due to abnormal functioning of the basal ganglia.Amongst tactile, auditory, visual and temporal perception and proprioception, there is little overlap in active and abnormal brain regions outside of the basal ganglia (Figures 3-8).An exception to this is found in the ACC and temporal lobe with regards to auditory and temporal perception, which perhaps relates to the relationship between internal timekeeping and internal "counting".These findings highlight the importance of the basal ganglia in sensory perception, which has not been well recognised historically.
The repeated demonstration of a multisensory response within the basal ganglia, along with their activation in response to tactile, visual and auditory stimuli, has clarified that while the basal ganglia may respond most strongly to proprioception, there is a clear response to the tactile, visual and auditory sensory modalities as well.Additionally, the larger magnitude of the neuronal response within the basal ganglia to multiple simultaneous modalities of sensory stimulation compared to a single modality [20] suggests that the role of the basal ganglia in sensory perception may not be only as a node in a relay network, but as a processing facility for integrating various simultaneous sensory inputs.

What Could Be the Reason for Reduced Sensory Function in Parkinson's Disease?
Increased noise or decreased specificity in the PD sensory system [36,[71][72][73][74][75][76][77][78]] may be attributed to various factors including an inherent reduction in the specificity of basal ganglia signalling in response to sensory stimuli [36,71,75], an increase in receptive field size considering the somatopic organization of the basal ganglia sensory system [53,56,[58][59][60][61][62][63][67][68][69]79] and potentially continuous proprioceptive signalling, due to impaired muscular relaxation in PD, leading to overburdening of neuronal firing or noise [80].Indeed, this loss of specificity may extend beyond the basal ganglia as Escola [72] found evidence of decreased cortical specificity to sensory responses via MER in the pre-SMA and SMA in primate models of PD.Somatosensory-evoked potential studies in PD cohorts have also revealed a number of features such as enlarged high-frequency oscillations in the S1 region [81,82], which are possibly caused by the high burden of proprioceptive signalling.The hypothesis that reduced sensory response in PD may be related to increased proprioceptive noise caused by high resting muscular tone is supported by the Pierantozzi [83] study which revealed that a reduced somatosensory-evoked potential N30 response in PD had a remarkable increase in amplitude after a peripheral nerve block that reduced the tone.The Onofrj [84] study further supported this idea by showing that the equivalent somatosensory-evoked potential response in animals improves after a GABA (γ-Aminobutyric acid) activating anaesthetic agent, as opposed to dopaminergic medication, possibly due to reduction in tone by the former.'Sensorimotor gating' observed in non-PD populations shows that motor neuronal activity can inhibit sensory neuronal activity [14,50,80,85,86] and sensory neuronal activity can inhibit motor neuronal activity [87,88].This phenomenon is reduced in PD [14,15,80,[87][88][89], potentially impacting the precision of motor commands [88] and possibly contributing to excess motor output.Conversely, a high burden of proprioceptive noise is likely to lead to reduced motor output in PD due to sensorimotor gating [16].These irregularities in the normal homeostatic mechanisms of motor control in PD may underlie the paradox of excessive motor output (increased tone) and reduced motor output (bradykinesia).

Methodological Limitations within the Reviewed Studies
The Recurrent Problem of Determining Brain Response Related to a Pure Sensory Stimulus in the Setting of Confounds Several studies emphasize the importance of the contralateral S1 in the somatosensory temporal-discrimination task [22,49,50,85,86,[90][91][92][93][94][95] but there was no clear distinction between temporal and tactile processing.This ambiguity extends to other brain areas as well, with studies such as Pastor [26] and Lacruz [22] reporting involvement of the caudate in temporal processing but not controlling for the confound of brain-area activation to other forms of tactile stimuli.Additionally, Ferrandez [23] and Schmiedt [41] have highlighted examples of frontal-brain activity in sensory perception, suggesting higher-order "topdown" sensory processing occurring concurrently or shortly after pure sensory perception.The hypothesis is that these higher-order areas are distinct from brain regions involved in pure sensory perception and reflect processes of categorization, classification, problem solving, and working memory.These outcomes imply that the most robust evidence for brain-area involvement in pure sensory perception will be derived from studies that isolate brain areas activated by utilizing varying experimental paradigms to interrogate a single sensation perception.

Limitations within the Present Review
Despite extensive efforts to encompass a vast array of studies probing brain networks involved in sensory function and abnormalities in PD, some relevant studies, especially non-English ones, might have been overlooked.However, multiple reviewers screened the papers and there was no date restriction for publication, with the search strategy including all databases.The reviewing team included clinicians and a neuroscientist for a broader interpretation.
There are established connections between ascending sensory pathways and the circuits involved in novelty-reward processing (including the nucleus accumbens).However, studies with experimental paradigms designed to assess the novelty-reward dopaminergic system in combination with sensory testing were not captured within our inclusion/exclusion criteria (see the Supplementary Materials).The question of involvement of the novelty-reward dopaminergic system therefore could not be addressed within this review.However, it is conceivable that if this system was activated by the sensory-testing experimental paradigm, PwPD would have a reduced response due to the reduction in dopaminergic function in this area.This is an important consideration that should be addressed in a future literature synthesis.Notably, the sensory-testing methodologies that were included in our review (see Section 3.2.2) included no rewards for participation in activities, nor for achieving a goal performed through sensory system activation, for animals or humans.Animal studies in particular usually occurred in a restrained animal which was exposed to a pure sensory stimulus.It is conceivable that the motivation-reward circuit may have activated in humans who wanted to do well in a task; however, there was generally no feedback for participants as to whether responses were good/bad or correct/incorrect.As such, any activation of reward circuitry would likely be intrinsic to the tasks (i.e., stimulus novelty) rather than explicit to the experimental designs.

Difficulty Interrogating Deep Brain Nuclei for Sensory Function in Human Models
Inconsistencies emerged in studies focusing on individual sensory systems, particularly those interrogating the basal ganglia, due to challenges in controlling for costimulation or cross-talk, and the limitations of existing technologies like fMRI, PET, and surface EEG in studying small deep nuclei.Additionally, human MER is not possible in the research setting unless in the context of a therapeutic intervention, which therefore limits the studied population to diseased cohorts.A potential technology that may be able to provide further insights is brain-focused ultrasound, which has the ability to reversibly modulate the function of deep brain nuclei in a non-invasive procedure, with a neuromodulation focus as small as 2 mm [96].Another potential method to interrogate sensory function of deep nuclei would be to test various sensory functions before and after DBS implantation [8].

Conclusions
This review described brain-area activation in HC and abnormalities in PD associated with tactile, proprioceptive, auditory, visual, and temporal sensory modalities.In PD, the basal ganglia are consistently implicated in sensory dysfunction across all sensory modalities.The sensory dysfunction may be attributed to concepts like "increased noise" and "decreased neuronal specificity," possibly due to factors like high resting muscular tone, increased receptive field size of sensory neurons, and inherent decreased specificity at the basal ganglia.Sensory-system dysfunction may have significant clinical importance as it may, in part, lead to motor-system dysfunction due to sensorimotor gating.Irregularities in the normal homeostatic sensorimotor mechanisms of motor control in PD may underlie the paradox of excessive motor output (increased tone) and reduced motor output (bradykinesia).Irrespective of their impact on motor function, sensory abnormalities are also associated with reduced function and quality of life in their own right (Chaudhuri and Schapira 2009).The analysis of sensory-system dysfunction therefore may be paramount in understanding functional disability in PD.
Exploring brain networks involved in sensory modalities in normal populations and assessing the associated dysfunctions in PD offers valuable insights into the pathological processes in PD and their functional implications.Despite the historical focus on the motor

Figure 2 .
Figure 2. Breakdown of the numbers of studies extracted between disease state, animal or human cohort and sensory modality.

Figure 2 .
Figure 2. Breakdown of the numbers of studies extracted between disease state, animal or human cohort and sensory modality.

20 Figure 5 .
Figure 5. Brain areas involved in auditory processing; the green shading denotes areas found to be involved in healthy controls (ACC = 2 studies, superior temporal gyrus = 2 studies, putamen = 1 study, globus pallidus = 1 study, SN = 1 study, caudate = 1 study) and the red outline denotes areas found to function abnormally in patients with Parkinson's disease (superior temporal gyrus = 3 studies, globus pallidus = 1 study).This summarises findings from 7 human studies and 2 animal studies.ACC = anterior cingulate cortex.GPi = globus pallidus internal, GPe = globus pallidus externa.3.3.4.What Brain Areas Are Found to Function Abnormally during Auditory Processing in Patients with Parkinson's Disease?

Figure 5 .
Figure 5. Brain areas involved in auditory processing; the green shading denotes areas found to be involved in healthy controls (ACC = 2 studies, superior temporal gyrus = 2 studies, putamen = 1 study, globus pallidus = 1 study, SN = 1 study, caudate = 1 study) and the red outline denotes areas found to function abnormally in patients with Parkinson's disease (superior temporal gyrus = 3 studies, globus pallidus = 1 study).This summarises findings from 7 human studies and 2 animal studies.ACC = anterior cingulate cortex.GPi = globus pallidus internal, GPe = globus pallidus externa.

3. 3 . 4 .
What Brain Areas Are Found to Function Abnormally during Auditory Processing in Patients with Parkinson's Disease?

3. 3
.5.What Brain Areas Are Involved in Visual Processing in Healthy Cohorts?

Figure 6 .
Figure 6.Brain areas involved in visual processing; the green shading denotes areas found to be involved in healthy controls (primary visual cortex = 2 studies, caudate = 2 studies, putamen = 1 study, substantia nigra = 2 studies, globus pallidus = 1 study) and the red outline denotes areas found to function abnormally in patients with Parkinson's disease (primary visual cortex = 4 studies, retina = 4 studies, globus pallidus = 1 study).This summarises findings from 8 human studies and 2 animal studies.GPi = globus pallidus interna, GPe = globus pallidus externa.teriorcingulate cortex.GPi = globus pallidus internal, GPe = globus pallidus externa.3.3.6.What Brain Areas Are Found to Function Abnormally during Visual Processing in People with Parkinson's Disease?
work ofZhao [52]  in HCs further supports the importance of bilateral S1 and contralateral IPL (Brodmann area 40) in tactile sensory perception.3.3.7.What Brain Areas Are Involved in Tactile Processing in Healthy Cohorts?Unsurprisingly, a number of studies have demonstrated the importance of th contralateral primary sensory cortex (S1) in tactile sensation processing[25,27,48-50 (Figure7).Mowery[51] also demonstrated the importance of the contralateral S1 in a non PD animal model with MER during whisker deflection in rats.Pastor additionally found S1 and the inferior parietal lobule (IPL) activation common to both tactile spatial and tactile temporal discrimination.The work ofZhao [52]  in HCs further supports th importance of bilateral S1 and contralateral IPL (Brodmann area 40) in tactile sensor perception.

3. 3 . 8 .
What Brain Areas Are Found to Function Abnormally during Tactile Processing in Patients with Parkinson's Disease?

3. 3 . 8 .
What Brain Areas Are Found to Function Abnormally during Tactile Processing in Patients with Parkinson's Disease?

4. 1 .
What Brain Areas Are Involved in Sensory Processing, What Are the Abnormalities in Parkinson's Disease, and What Is the Importance of the Basal Ganglia?

4. 3 .
How Does the Brain Integrate Sensory and Motor Function, and What Is the Significance of This with Regard to Parkinson's Disease?