Neuroimaging Biomarkers in Parkinson's Disease: Current Applications and Future Directions

1. Introduction

Parkinson's disease (PD) is the second most common neurodegenerative disorder worldwide and affects more than 10 million individuals globally [1]. It is characterized by progressive degeneration of dopaminergic neurons within the SNpc and the accumulation of misfolded αSyn pathology throughout the central and peripheral nervous systems [2,3,4]. Although PD is traditionally defined by its cardinal motor manifestations, including bradykinesia, rigidity, resting tremor, and postural instability, pathological and biomarker studies indicate that neurodegeneration begins many years before clinical diagnosis, with substantial nigrostriatal neuronal loss already present at symptom onset [5,6]. Neuropathological studies suggest that αSyn pathology progresses in a stereotyped pattern, initially involving the olfactory bulb and lower brainstem before ascending to midbrain and cortical regions [7]. In addition, αSyn pathology has been identified within sympathetic and parasympathetic neurons, including preganglionic and postganglionic autonomic structures, supporting the concept that autonomic involvement occurs early in PD and may precede the onset of classical motor manifestations [8].

The recognition of prodromal and preclinical phases of PD has transformed the field's approach to disease detection and therapeutic development [9]. Increasing attention has focused on biomarkers capable of identifying underlying pathology before the appearance of overt motor symptoms, improving diagnostic accuracy, monitoring disease progression, facilitating biological staging, and enabling the evaluation of disease-modifying therapies [10,11,12]. This paradigm shift has been further reinforced by emerging biologically based disease frameworks that emphasize molecular and pathological evidence of disease rather than clinical manifestations alone [13,14].

Among available biomarker modalities, neuroimaging offers a unique opportunity to visualize disease-related alterations in vivo. Advanced imaging techniques can assess nigrostriatal dopaminergic function, cerebral metabolic activity, autonomic nervous system involvement, synaptic integrity, neuroinflammation, and structural changes associated with neurodegeneration. Consequently, neuroimaging has become a central component of contemporary PD research and increasingly contributes to diagnosis, prognostication, patient stratification, and clinical trial design [15,16].

An ideal biomarker should demonstrate biological relevance, high sensitivity and specificity, reproducibility across centers, responsiveness to longitudinal change, and meaningful clinical utility (Table 1) [17]. Despite substantial progress, no currently available neuroimaging biomarker fulfills all of these characteristics. Nevertheless, several imaging modalities have demonstrated important applications across the disease continuum, ranging from prodromal detection and biological staging to monitoring disease progression and assessing therapeutic target engagement.

In this review, we summarize current and emerging neuroimaging biomarkers in PD, including molecular imaging and advanced MRI techniques. We discuss their biological basis, clinical applications, role in prodromal disease and disease progression, utility in clinical trials, and future integration within biomarker-driven frameworks aimed at advancing precision medicine in PD (Table 2).

2. Methodology

This narrative review was developed through a comprehensive literature search of the PubMed, Embase, and Google Scholar databases to identify studies evaluating neuroimaging biomarkers in PD. The search focused on publications available through July 2026 and used combinations of terms including PD, neuroimaging biomarkers, DAT, DAT-SPECT, PET, fluorodopa, VMAT2, FDG-PET, neuromelanin MRI, QSM, nigrosome-1, FW imaging, diffusion MRI, fMRI, αSyn PET, prodromal PD, and biomarker harmonization. Original research articles, systematic reviews, meta-analyses, consensus statements, and clinical practice guidelines published in English were considered. Particular emphasis was placed on studies examining molecular imaging modalities, including DAT-SPECT, fluorodopa and VMAT2 PET, metabolic imaging, cholinergic imaging, neuroinflammation imaging, synaptic imaging, and emerging α-synuclein PET tracers, as well as MRI-based biomarkers such as neuromelanin-sensitive MRI, QSM, nigrosome-1 imaging, diffusion MRI, FW imaging, and fMRI. Priority was given to landmark investigations, longitudinal cohort studies, multicenter biomarker initiatives such as the PPMI, consensus recommendations, and recent publications addressing biological disease staging, prodromal PD, biomarker standardization, and disease-modifying clinical trials. Additional relevant references were identified through manual review of reference lists from selected articles and expert consensus reports. Owing to the narrative nature of this review, formal systematic-review methodology and quantitative pooled analyses were not performed. Instead, the available literature was critically synthesized to provide a contemporary overview of the biological basis, diagnostic utility, prognostic significance, staging applications, and therapeutic relevance of current and emerging neuroimaging biomarkers across the PD continuum.

3. Molecular Imaging Biomarkers

3.1. Dopaminergic Imaging Biomarkers

Degeneration of nigrostriatal dopaminergic neurons is a pathological hallmark of PD and provides the biological basis for several of the most widely utilized molecular imaging biomarkers. DAT imaging remains the most established molecular imaging modality in clinical practice [18]. DAT-SPECT assesses presynaptic dopaminergic terminal integrity by measuring striatal DAT availability [19]. Marek et al. demonstrated an annual decline of approximately 11% in striatal DAT binding in patients with PD compared with less than 1% in healthy controls, establishing DAT imaging as an objective measure of ongoing nigrostriatal degeneration [20]. The clinical importance of DAT imaging was further demonstrated in the PRECEPT study, where individuals with SWEDD showed minimal clinical and imaging progression over 22 months compared with DAT-deficit subjects, suggesting that most SWEDD cases are unlikely to represent idiopathic PD and highlighting the value of DAT imaging for diagnostic enrichment in clinical trials [21].

Reduced striatal DAT binding reflects loss of nigrostriatal projections and reliably distinguishes neurodegenerative parkinsonism from conditions that lack presynaptic dopaminergic degeneration, such as essential tremor and drug-induced parkinsonism [22,23]. However, dopaminergic denervation is not specific to PD and may also be observed in MSA, PSP, and corticobasal degeneration, limiting the ability of DAT imaging to differentiate among neurodegenerative parkinsonian disorders [24,25]. Additional PET-based dopaminergic biomarkers provide complementary information regarding nigrostriatal function. 18F-FDOPA PET evaluates dopamine synthesis capacity, whereas VMAT2 imaging reflects vesicular dopamine storage [26]. These techniques have significantly improved understanding of the temporal evolution of nigrostriatal degeneration and have demonstrated measurable abnormalities during prodromal stages of disease [6,16]. Early FDOPA PET studies also demonstrated a characteristic spatial pattern of nigrostriatal degeneration, with the dorsal putamen showing the greatest reduction in dopaminergic function early in the disease course, followed by progressive involvement of ventral putaminal regions as clinical severity increases [27]. Direct comparisons between presynaptic dopaminergic imaging modalities have demonstrated strong correlations between striatal DAT binding measured with FP-CIT SPECT and presynaptic dopaminergic function assessed with F-DOPA PET. Both techniques showed comparable associations with disease severity and duration and effectively distinguished early from advanced PD, supporting the clinical utility of DAT-SPECT as a more widely accessible alternative to PET-based assessments [28]. Longitudinal F-DOPA PET studies have demonstrated progressive declines in striatal dopaminergic function over time, with the earliest and most pronounced deficits occurring in the posterior putamen. Annual reductions in F-DOPA uptake of approximately 8% to 10% have been reported [29].

The utility of DAT imaging in prodromal synucleinopathies was demonstrated by Iranzo et al., who showed that patients with iRBD exhibiting reduced striatal DAT uptake and/or SN hyperechogenicity were at substantially increased risk of developing PD, DLB, or MSA during follow-up, whereas individuals with normal imaging remained disease-free [30].

An important challenge in dopaminergic imaging is harmonization of quantitative measurements across different radiotracers and imaging platforms. Similar to the Centiloid framework developed for amyloid PET imaging [31], recent efforts have focused on standardization of DAT and VMAT2 imaging analyses to improve comparability across studies and facilitate incorporation of these biomarkers into multicenter clinical trials [32,33]. More recently, the Centamine framework has been proposed as a tracer-independent quantitative scale for dopaminergic imaging. Using healthy control data obtained with 123I-ioflupane SPECT as a reference standard, this framework allows harmonization of measurements obtained with different dopaminergic tracers, including 18F-AV133 PET [33]. Preliminary studies have demonstrated strong correlations between standardized uptake measures and Centamine-derived values across striatal regions, supporting the feasibility of cross-tracer calibration and quantitative comparison between studies and imaging platforms [33].

3.2. Metabolic Network Imaging

FDG-PET provides an indirect measure of neuronal and synaptic activity through assessment of regional glucose metabolism. Unlike dopaminergic imaging, FDG-PET characterizes disease-related network dysfunction rather than specific neurotransmitter deficits. This approach has enabled identification of a characteristic PD-related pattern (PDRP), which consists of relative hypermetabolism within pallidothalamic, pontine, and cerebellar structures accompanied by relative hypometabolism in cortical association regions [34,35]. The original metabolic topography studies by Eidelberg et al. demonstrated a characteristic covariance pattern characterized by relatively increased metabolism in the lentiform nucleus and thalamus together with reduced activity in frontal and parietal association cortices [36]. Expression of the PDRP has shown potential utility for diagnosis, monitoring disease progression, and evaluating treatment response, highlighting the importance of network-level dysfunction in PD pathophysiology [37,38]. Feigin et al. showed that levodopa administration reduced PDRP expression and that the magnitude of network suppression correlated with improvement in UPDRS motor scores [39]. Tang et al. demonstrated that automated FDG-PET pattern analysis accurately differentiated PD, multiple system atrophy, and progressive supranuclear palsy, achieving specificities of 97%, 96%, and 94%, respectively, highlighting the utility of metabolic imaging for the differential diagnosis of parkinsonism [40].

3.3. Cholinergic Imaging

Although dopaminergic dysfunction remains the defining neurochemical abnormality of PD, degeneration of non-dopaminergic systems contributes substantially to clinical heterogeneity and disease progression. Among these systems, cholinergic dysfunction has emerged as an important contributor to cognitive impairment (especially, attention and executive performance) [41], gait dysfunction, postural instability, olfactory deficits, and RBD [42,43]. PET studies employing acetylcholinesterase tracers have demonstrated reduced cortical and thalamic cholinergic innervation in patients with PD, particularly among those with cognitive impairment and gait disturbances [43,44]. Cholinergic dysfunction has also been associated with olfactory deficits and RBD, suggesting that cholinergic involvement may occur relatively early in the disease course [45,46]. More recently, abnormalities in cholinergic networks have been identified in genetic forms of PD, further supporting the contribution of non-dopaminergic pathways to disease heterogeneity [47]. Cholinergic imaging has also provided important insights into axial symptoms of PD, Bohnen et al. demonstrated that PD patients with a history of falls exhibited reduced cortical and thalamic cholinergic activity compared with non-fallers, whereas nigrostriatal dopaminergic denervation did not differ between groups [48]. Expanding on these findings, Müller et al. demonstrated that reduced thalamic cholinergic innervation was independently associated with impaired postural sensory integration and increased postural sway in patients with PD, whereas cortical cholinergic and striatal dopaminergic deficits showed no significant relationship [49].

3.4. Neuroinflammation Imaging

Neuroinflammation has increasingly been recognized as a potential contributor to PD pathogenesis and progression. PET tracers targeting the TSPO, a marker of activated microglia, have demonstrated increased uptake in the SN, striatum, and cortical regions of patients with PD [50,51,52]. Although findings have been somewhat inconsistent because of methodological variability and limitations of first-generation TSPO tracers, more recent studies continue to support the presence of microglial activation throughout the disease course [53,54]. Emerging PET tracers targeting the CSF1R, another marker of microglial activation, have demonstrated increased binding in patients with mild and moderate PD, with tracer uptake correlating with motor severity. These findings suggest potential utility for neuroinflammation imaging as a biomarker of disease activity and progression [55].

3.5. Synaptic Imaging

Synaptic dysfunction is increasingly recognized as a fundamental component of PD pathology and may occur earlier than overt neuronal loss. PET tracers targeting SV2A, a marker of presynaptic density, have enabled in vivo assessment of synaptic integrity. Using SV2A PET imaging, Delva et al. demonstrated reduced synaptic density within the SN and BG of patients with PD, supporting the presence of widespread synaptic loss [56,57]. More recently, Holmes et al. reported significant reductions in SV2A density within the SN and red nucleus, with lower tracer uptake correlating with greater motor impairment. These observations suggest that synaptic imaging may provide a sensitive marker of disease burden and neuronal dysfunction beyond conventional dopaminergic imaging [58].

3.6. αSyn Imaging

Direct visualization of αSyn pathology represents one of the most important unmet needs in PD biomarker research (Table 3). Despite major advances in molecular imaging, αSyn aggregation remains inaccessible to routine in vivo quantification. This limitation has important implications because αSyn pathology constitutes the central pathological process underlying PD, DLB, and MSA [59,60]. Development of αSyn PET tracers has proven substantially more challenging than development of Aβ or tau imaging agents. Major obstacles include the relatively low concentration of αSyn aggregates, their predominantly intracellular distribution, significant conformational heterogeneity among synucleinopathies, and the need to avoid off-target binding to coexisting Aβ and tau pathology [61,62]. Nevertheless, several candidate tracers have recently entered clinical evaluation. The tracer 18F-ACI-12589 has demonstrated promising results in patients with MSA, particularly the cerebellar subtype, where retention is observed predominantly within cerebellar white matter and middle cerebellar peduncles [63]. Although discrimination between PD and healthy controls remains limited, these findings suggest utility for imaging MSA-related αSyn pathology [64].

Additional tracers under investigation include 18F-FD4 and 11C-CYS08 (SYS08). Preliminary studies of FD4 have demonstrated uptake within regions affected by synuclein pathology across cohorts with PD, iRBD, and MSA, supporting its potential utility as a biomarker of αSyn-related neurodegeneration and disease staging (https://www.ablitherapeutics.com/news/detail/9633/synusight-biotech-abli-therapeutics-and-xingimaging-announce-strategic-collaboration-to-implement-alpha-synuclein-pet-imaging-into-clinical-trials-evaluating-risvodetinib-as-a-disease-modifying-therapy-for-parkinsons-disease, accessed on Aug 17th, 2026). SYS08, developed at Massachusetts General Hospital and Harvard Medical School, has demonstrated high affinity for αSyn fibrils, favorable brain penetration, limited off-target binding, and substantially greater selectivity for αSyn than for Aβ or tau pathology. Early first-in-human studies have demonstrated regional uptake differences among healthy controls, PD, and DLB cohorts, suggesting that in vivo visualization of αSyn pathology may be achievable (https://www.alzforum.org/news/conference-coverage/mjff-gathers-scientists-advance-synuclein-tracer-development, accessed on Aug 17th, 2026).

Although no αSyn PET tracer has yet achieved widespread clinical implementation, successful development of these agents could fundamentally transform diagnosis, biological staging, and therapeutic development. Similar to the impact of amyloid and tau PET imaging in Alzheimer's disease, αSyn imaging could enable biomarker-defined enrollment strategies, facilitate prevention trials, and provide objective measures of therapeutic target engagement and disease progression.

3.7. Imaging of Other Neurotransmitter Systems

3.7.1. Noradrenergic Imaging

Degeneration of the LC is among the earliest pathological changes identified in PD and may precede substantial nigral degeneration in some individuals [65]. The noradrenergic system contributes to attention, arousal, autonomic regulation, gait control, mood, fatigue, and cognitive resilience. PET tracers targeting the NET, including 11C-methylreboxetine (11C-MRB) and 11C-MeNER, allow in vivo assessment of noradrenergic terminal integrity [66,67]. Studies have demonstrated reduced tracer binding within the thalamus, LC projection regions, and widespread cortical areas, supporting extensive noradrenergic involvement in PD [68]. Recent multimodal PET-MRI investigations combining 11C-yohimbine PET with neuromelanin-sensitive MRI demonstrated both reduced LC neuromelanin signal and decreased α2-adrenergic receptor availability in PD [68,69]. Importantly, these abnormalities correlated with bradykinesia, tremor, fatigue, apathy, constipation, and anxiety, highlighting the clinical relevance of noradrenergic dysfunction [68]. The potential utility of noradrenergic imaging extends beyond symptom characterization. Because LC pathology may occur early in the disease process, noradrenergic biomarkers could contribute to prodromal disease identification, biological staging, and patient stratification within emerging brain-first and body-first disease models [4,70].

3.7.2. Serotonergic Imaging

Pathological studies indicate that serotonergic nuclei within the raphe complex are affected early in PD, potentially preceding significant nigrostriatal degeneration. PET imaging using ligands targeting the SERT, including 11C-DASB, has demonstrated widespread serotonergic deficits involving the raphe nuclei, striatum, limbic structures, and neocortex [71]. Serotonergic dysfunction has been associated with depression, anxiety, apathy, fatigue, RBD, and cognitive impairment [72]. Furthermore, one of the best-established applications of serotonergic imaging relates to LID. PET studies have demonstrated an abnormally elevated serotonergic-to-dopaminergic terminal ratio within the putamen of patients with LID [73]. Because serotonergic terminals convert exogenous levodopa into dopamine without appropriate autoregulatory control, excessive extracellular dopamine fluctuations may contribute directly to dyskinesia development. These findings have important therapeutic implications, as modulation of 5-HT_1A and 5-HT_1B receptors has been shown to reduce abnormal dopamine release and mitigate dyskinesia severity [74]. Politis et al. demonstrated excessive serotonergic innervation within the grafted striatum of transplanted patients who developed severe graft-induced dyskinesias, with symptom improvement following administration of a 5-HT1A receptor agonist [75].

3.7.3. Glutamatergic Imaging

Glutamatergic neurotransmission plays a central role in BG circuitry and is increasingly recognized as an important mediator of motor complications and network dysfunction in PD. Hyperactivity of glutamatergic pathways is thought to contribute to LID, motor fluctuations, and abnormal cortical excitability [76]. Emerging PET tracers targeting mGluR5 including 11C-ABP688, have enabled in vivo assessment of glutamatergic signaling [77]. Although clinical experience remains limited, studies suggest that altered mGluR5 expression may be associated with LID susceptibility and abnormal cortico-striatal plasticity [78]. As disease-modifying and circuit-based therapeutic approaches evolve, glutamatergic imaging may provide valuable information regarding network-level dysfunction and treatment target engagement [79].

3.7.4. Adenosine A2A Receptor Imaging

Adenosine A2A receptors are highly expressed within the indirect pathway of the BG and modulate dopaminergic neurotransmission through interactions with D_2 receptors [80]. This pathway has particular clinical relevance because A2A receptor antagonists, including istradefylline, are approved therapies for motor fluctuations in PD [81]. PET imaging using radioligands such as 11C-TMSX and other A2A receptor tracers has demonstrated altered receptor availability in striatal regions of patients with PD [82,83]. These abnormalities may reflect compensatory responses to dopamine depletion and may influence response to dopaminergic therapy [84].

3.7.5. Opioid Receptor Imaging

The endogenous opioid system participates in BG regulation and motor control and may contribute to pain processing, affective symptoms, and LID [85]. PET studies employing μ-opioid receptor tracers such as 11C-diprenorphine have demonstrated altered opioid receptor binding in striatal, thalamic, and cortical regions in PD [86]. Alterations in opioid signaling appear particularly relevant in patients with motor complications and chronic pain syndromes [87].

3.7.6. Autonomic Imaging

Cardiac sympathetic denervation is a well-established feature of PD and reflects involvement of postganglionic autonomic neurons by αSyn pathology [88]. Myocardial scintigraphy using 123I-MIBG, a norepinephrine analogue, permits in vivo assessment of cardiac sympathetic innervation and has emerged as one of the most extensively studied autonomic biomarkers in PD [89]. Reduced cardiac MIBG uptake is observed in many patients with PD and DLB, whereas uptake is generally more preserved in MSA and PSP, although overlap and atypical findings may occur, reflecting differences in the distribution of autonomic pathology [90,91,92]. Supporting the biological basis of these findings, Goldstein et al. demonstrated evidence of cardiac sympathetic denervation in many patients with PD, including some without overt autonomic failure, indicating that peripheral catecholaminergic degeneration is a common feature of the disease and may occur independently of motor severity or levodopa exposure [93]. Consequently, cardiac MIBG scintigraphy may assist in differentiating Lewy body disorders from other causes of parkinsonism and has been incorporated into diagnostic criteria for DLB [94]. Beyond its diagnostic utility, cardiac sympathetic imaging has provided important insights into disease pathophysiology and prodromal neurodegeneration. Abnormal cardiac MIBG uptake has been demonstrated in individuals with iRBD, PAF, and incidental LBD, suggesting that cardiac sympathetic denervation may precede the onset of classical motor symptoms by many years [95,96]. These findings support the concept that autonomic dysfunction represents an early manifestation of αSyn pathology in at least a subset of patients [97]. Neuropathological studies support these observations by demonstrating that αSyn aggregates accumulate within distal cardiac sympathetic axons before substantial neuronal loss occurs in sympathetic ganglia, suggesting that cardiac sympathetic degeneration may represent an early and relatively selective manifestation of Lewy body pathology [98]. More recently, cardiac MIBG imaging has assumed a prominent role in studies evaluating biological heterogeneity in PD. According to the brain-first versus body-first model, some patients may develop αSyn pathology initially within the peripheral autonomic nervous system before subsequent propagation to the brain. Reduced cardiac MIBG uptake has been linked to the body-first phenotype and is frequently observed in patients with iRBD and PD accompanied by RBD [99], whereas patients with brain-first disease may demonstrate relatively preserved cardiac sympathetic innervation during earlier disease stages [100]. Consequently, autonomic imaging may contribute not only to diagnosis but also to biological subtyping, disease staging, and future precision medicine approaches.

4. MRI Biomarkers in PD

4.1. Conventional Structural MRI

Conventional MRI is frequently normal in early PD but remains essential for excluding alternative causes of parkinsonism (Table 4). Although routine structural imaging has limited sensitivity for detecting early PD, characteristic abnormalities may assist in the diagnosis of atypical parkinsonian disorders, including putaminal changes in MSA and midbrain atrophy in PSP [101,102]. Consequently, conventional MRI remains an important component of the diagnostic evaluation despite its limited ability to directly visualize nigrostriatal neurodegeneration.

4.2. Neuromelanin-Sensitive MRI

Neuromelanin-sensitive MRI enables visualization of neuromelanin-containing neurons within the SN and LC [103]. This technique is increasingly regarded as a direct imaging marker of SNpc neurodegeneration [104]. Reductions in neuromelanin signal correlate with dopaminergic neuronal loss, disease severity, and clinical progression, supporting its potential role as a biomarker for both diagnosis and longitudinal monitoring [105,106]. Several studies have reported strong diagnostic performance, with area-under-the-curve values ranging from approximately 0.85 to 0.96 for differentiating PD from healthy controls. Visual assessment by experienced readers has also demonstrated diagnostic accuracies approaching 85%, supporting the potential clinical applicability of this technique [107,108]. Early work using automated neuromelanin-MRI segmentation demonstrated excellent diagnostic performance, with substantia nigra pars compacta volumetry achieving AUC values of 0.93 to 0.94 and sensitivities and specificities approaching 90% for distinguishing PD from healthy controls [109].

Beyond the SN, neuromelanin-sensitive MRI can visualize the LC, a major noradrenergic nucleus affected early in PD. Reduced neuromelanin signal in both structures supports the concept that PD involves degeneration of multiple catecholaminergic systems rather than isolated nigrostriatal dysfunction [110,111]. Emerging evidence suggests that neuromelanin MRI abnormalities may be detectable during prodromal stages of disease. Reduced SN neuromelanin volume and signal intensity have been observed in individuals with iRBD and asymptomatic LRRK2 mutation carriers, suggesting that neuromelanin imaging may identify neurodegenerative changes before the onset of classical motor manifestations [107,112].

The biological basis of the neuromelanin MRI signal remains an active area of investigation. Experimental studies suggest that signal hyperintensity reflects the paramagnetic properties of neuromelanin-iron complexes as well as local tissue water content and magnetization-transfer effects [113]. Consequently, reductions in neuromelanin signal likely reflect a combination of neuronal loss and alterations in tissue composition accompanying neurodegeneration [114,115,116]. Importantly, neuromelanin MRI findings correlate with established molecular imaging markers of dopaminergic degeneration. Significant associations have been reported between SN neuromelanin loss and reduced striatal DAT binding measured by DAT-SPECT and DAT-PET, supporting the biological validity of neuromelanin MRI as a surrogate marker of nigrostriatal degeneration [117,118,119]. Longitudinal modeling studies suggest that neuromelanin-sensitive MRI abnormalities emerge later than reductions in striatal DAT binding, becoming detectable approximately 5 to 6 years before clinical diagnosis [120,121]. These observations suggest that neuromelanin MRI and dopaminergic molecular imaging capture distinct phases of disease evolution. Neuromelanin MRI changes have also been associated with cognitive performance, attentional dysfunction, and neuropsychiatric symptoms such as apathy, suggesting that this biomarker may reflect broader disease-related network dysfunction extending beyond the motor system [122].

4.3. Swallow-Tail Sign (Nigrosome-1 Imaging)

The swallow-tail sign refers to the normal appearance of nigrosome-1 within the dorsolateral SN on high-resolution SWI. In healthy individuals, nigrosome-1 appears as a characteristic hyperintense region resembling a swallow's tail [123]. Loss of this appearance has been associated with degeneration of nigrosome-1 dopaminergic neurons and local iron accumulation within the SN, providing a structural imaging marker of underlying neurodegeneration in PD [124]. Several studies have demonstrated high diagnostic accuracy for differentiating PD from healthy controls, with visual assessment of high-resolution SWI achieving accuracies exceeding 90% in some cohorts [125,126,127]. Also, Reiter et al. reported an overall diagnostic accuracy of approximately 95% for distinguishing neurodegenerative parkinsonism from healthy controls based on the absence of dorsolateral nigral hyperintensity on 3T SWI [128]. However, variability in acquisition protocols, magnetic field strength, and image interpretation has limited widespread clinical implementation. Histopathological and ultra-high-field MRI studies suggest that loss of the dorsal nigral hyperintensity reflects increased iron accumulation within nigrosome-1 [124]. Similar abnormalities have been reported in prodromal populations, including individuals with iRBD, supporting its potential role as an early marker of neurodegeneration [129,130].

4.4. Iron-Sensitive MRI (SWI and QSM)

Iron accumulation within the SN is a well-established pathological feature of PD. SWI and QSM allow in vivo assessment of iron deposition and consistently demonstrate increased nigral iron content in PD compared with healthy controls [131,132]. Histopathological correlation studies have demonstrated strong relationships between MRI-derived susceptibility measures and postmortem brain iron concentrations, supporting the biological validity of iron-sensitive MRI as an in vivo marker of iron deposition [133]. Both R2* relaxometry and QSM have shown robust correlations with regional brain iron content [131,134,135]. Iron deposition appears to be most prominent within the posteroventral SN, a region particularly vulnerable to dopaminergic neuronal degeneration [136]. Longitudinal QSM studies have demonstrated progressive increases in magnetic susceptibility over time, suggesting that iron accumulation continues throughout disease progression [137]. Among currently available techniques, QSM appears more sensitive than conventional R2* relaxometry for detecting nigral iron accumulation. Several studies have reported diagnostic accuracies ranging from approximately 0.80 to greater than 0.90 for differentiating PD from healthy controls, supporting its growing role as a quantitative biomarker of pathology [132].

Longitudinal studies further suggest that abnormalities in iron-sensitive MRI may emerge during very early stages of disease and could precede some structural changes observed with neuromelanin-sensitive MRI. Progressive increases in susceptibility measurements have been observed across disease stages, supporting the potential role of iron imaging as a biomarker of biological progression [120,138]. Additional support for early iron dysregulation comes from studies of iRBD, in which increased SN iron accumulation has been observed relative to healthy controls, suggesting that iron deposition may precede clinically manifest PD [139]. MRI measures of nigral iron accumulation have also demonstrated associations with motor impairment, particularly bradykinesia and rigidity, as well as emerging relationships with cognitive performance [140,141]. Moreover, experimental evidence further suggests that a substantial component of the iron-related MRI signal within nigrosome-1 reflects iron bound to neuromelanin-containing dopaminergic neurons, linking neuromelanin-sensitive and iron-sensitive MRI biomarkers to a common biological substrate [142,143].

4.5. Diffusion MRI and FW Imaging

Diffusion MRI techniques evaluate microstructural tissue integrity and provide information regarding tissue architecture beyond conventional structural imaging. Among these approaches, FW imaging has emerged as a particularly promising biomarker [144]. Increased FW content within the posterior SN has been associated with disease severity and longitudinal progression in PD [145,146]. Biologically, FW imaging is thought to reflect expansion of the extracellular compartment and may be influenced by neuroinflammation, edema, and neurodegeneration-related tissue remodeling [147]. Experimental studies have demonstrated increases in FW measurements in association with inflammatory processes, while human investigations have reported correlations with biomarkers of both neurodegeneration and neuroinflammation [148,149]. Several studies have demonstrated that increased FW content within the posterior SN correlates with reductions in presynaptic dopaminergic function measured by DAT-SPECT and FDOPA PET. These findings suggest that FW imaging reflects biologically meaningful nigrostriatal degeneration and may provide a noninvasive marker of disease severity and progression [150,151]. Importantly, FW abnormalities have also been identified in individuals with iRBD and appear to increase longitudinally over time, supporting the potential role of FW imaging as both an early and progression-sensitive biomarker of αSyn-related neurodegeneration [152,153].

4.6. fMRI

fMRI provides insight into large-scale neural network dysfunction in PD. Resting-state connectivity studies have demonstrated abnormalities involving motor, cognitive, and limbic networks, indicating that disease-related dysfunction extends beyond the BG [154,155]. Early rs-fMRI investigations demonstrated abnormal connectivity within motor networks, including reduced connectivity between the pre-supplementary motor area and the putamen, premotor cortex, and parietal regions [156]. Tessitore et al. demonstrated reduced default mode network connectivity in cognitively unimpaired patients with PD, particularly involving the medial temporal and inferior parietal regions, suggesting that functional network abnormalities may precede overt cognitive impairment [157]. These findings support contemporary views of PD as a multisystem disorder involving widespread disruption of distributed neural networks. Although fMRI remains primarily a research tool, it has contributed substantially to understanding the neural substrates underlying both motor and non-motor symptoms.

4.7. Multimodal MRI Biomarkers

Increasing evidence suggests that multimodal MRI approaches may outperform individual imaging techniques. Studies combining neuromelanin-sensitive MRI, QSM, FW imaging, and nigrosome-1 assessment have demonstrated higher diagnostic accuracy than any single MRI biomarker alone [158,159,160]. Longitudinal studies have also shown that neuromelanin MRI, iron-sensitive imaging, and FW imaging may track disease progression over time and represent potential biomarkers for disease-modifying clinical trials. However, further multicenter validation, harmonization, and technical standardization are required before routine clinical implementation [161,162].

5. Discussion

5.1. Neuroimaging Biomarkers Across the Disease Continuum

Neuroimaging has transformed the study of PD by providing objective biomarkers of neurodegeneration throughout the disease continuum (Table 5). Beyond its diagnostic role, neuroimaging increasingly contributes to biological staging, prognostication, longitudinal monitoring, and therapeutic development. The expanding application of imaging biomarkers reflects the ongoing transition from symptom-based diagnosis toward biologically defined disease characterization [13,14,15].

5.1.1. Neuroimaging Biomarkers in Prodromal PD

One of the most promising applications of neuroimaging is the identification of individuals at risk for developing PD before motor symptom onset. Hyposmia represents one of the most common prodromal manifestations of PD. In the Parkinson Associated Risk Syndrome (PARS) study, hyposmic individuals with DAT imaging abnormalities exhibited a substantially increased risk of future conversion to clinically manifest PD compared with hyposmic individuals without imaging deficits [163,164]. Similarly, iRBD has emerged as one of the strongest predictors of future αSyn pathology. In a large multicenter cohort, Postuma et al. demonstrated that most patients with iRBD eventually developed PD, DLB, or related synucleinopathies and identified clinical and biomarker features associated with phenoconversion [165]. Longitudinal studies have demonstrated that reduced striatal DAT uptake predicts conversion to PD, DLB, or MSA, supporting its use as a biomarker of ongoing neurodegeneration during the prodromal phase [166,167]. Neuroimaging studies have also provided valuable insights into genetically defined populations. Some studies demonstrated dopaminergic abnormalities in asymptomatic LRRK2 mutation carriers, suggesting that measurable neurobiological changes may be present years before clinical manifestations become apparent [168]. An important unresolved question concerns the anatomical origin and temporal evolution of PD pathology. Imaging and pathological studies suggest that disease progression may differ among individuals, supporting both brain-first and body-first models of PD [99]. Longitudinal imaging studies have further demonstrated that markers of nigrostriatal dysfunction may become abnormal years before diagnosis, whereas other imaging biomarkers appear to evolve later in the disease course [120,169]. These observations reinforce the concept of biological heterogeneity during the prodromal phase of PD.

Beyond its diagnostic utility, DAT imaging is increasingly being investigated as a quantitative marker of biological staging. Recent analyses from longitudinal cohorts such as the PPMI suggest that the degree of striatal dopaminergic deficit may provide information regarding disease trajectory and potentially predict the timing of phenoconversion among prodromal individuals. This emerging concept, sometimes referred to as a “DAT clock,” represents a shift from binary interpretation of DAT imaging toward the use of dopaminergic biomarkers for estimating disease stage and progression risk [161,170,171]. Recent longitudinal modeling approaches have further expanded the role of DAT imaging in biological staging. Using latent time variable models, investigators have proposed individualized trajectories of DAT decline that estimate the temporal position of a patient along the neurodegenerative continuum. Rather than categorizing imaging findings as simply normal or abnormal, these approaches attempt to quantify disease stage and estimate progression relative to population-based trajectories, supporting the concept of PD as a biologically staged disorder [14,161]. Importantly, imaging and clinical manifestations may not progress synchronously during the prodromal phase. Saari et al. demonstrated that changes in striatal DAT binding did not consistently correlate with motor progression in patients with iRBD [172]. Similarly, Honkanen et al. reported discordance between DAT-SPECT findings and motor performance, highlighting the heterogeneity of prodromal PD and supporting the use of multimodal biomarkers rather than reliance on a single imaging measure [173].

Data-driven analyses incorporating longitudinal DAT-SPECT datasets have demonstrated substantial interindividual heterogeneity in the rate and spatial distribution of nigrostriatal degeneration, supporting the existence of distinct progression trajectories and biological subtypes of PD [174]. More recently, machine-learning approaches applied to DAT-SPECT have enabled simultaneous inference of disease stage and subtype from regional patterns of nigrostriatal dopaminergic degeneration, suggesting that DAT imaging may contribute directly to future biological staging frameworks rather than functioning solely as a marker of dopaminergic deficit [175]. Barber et al. further demonstrated that individuals with iRBD exhibit widespread motor, cognitive, autonomic, and olfactory abnormalities consistent with prodromal PD, with 74% meeting Movement Disorder Society criteria for probable prodromal PD [176].

5.1.2. Neuroimaging Biomarkers of Disease Progression

An important characteristic of an ideal biomarker is the ability to monitor disease progression over time (Table 6). Longitudinal PET studies have demonstrated progressive declines in DAT, fluorodopa, and VMAT2 signals throughout the course of PD. Importantly, these trajectories appear non-linear. Modeling studies by de la Fuente-Fernández and colleagues suggested that measurable abnormalities may begin approximately 17 years before diagnosis for VMAT2 imaging, 13 years before diagnosis for DAT imaging, and 6 years before diagnosis for fluorodopa PET [6]. Neuropathological studies support these findings. Kordower et al. demonstrated a steep decline in nigral neuronal density during early disease stages followed by relative stabilization as surviving neuronal populations become increasingly depleted [5]. These observations suggest that the rate of biological progression may be greatest during prodromal and early symptomatic phases. The recognition of non-linear disease progression has important implications for clinical trial design because interventions targeting disease modification may need to be initiated before extensive neuronal loss has occurred. Advanced longitudinal modeling studies suggest that dopaminergic decline follows a sigmoidal rather than linear trajectory, with relatively slow changes during preclinical stages, accelerated decline around symptom onset, and eventual plateauing in advanced disease. Latent time variable models have been proposed to estimate an individual's position along this trajectory by incorporating repeated DAT measurements over time. Such approaches may improve patient stratification and facilitate the detection of disease-modifying effects in clinical trials [6,161].

Additional support for DAT imaging as a progression biomarker has recently emerged from longitudinal analyses of the PPMI. In a cohort comprising 719 individuals with PD and nearly 2,000 longitudinal observations, decline in putaminal DAT availability was significantly associated with worsening contralateral motor impairment over time [177]. Notably, these associations became stronger when analyses accounted for disease laterality and regional patterns of degeneration, suggesting that previous inconsistencies between DAT imaging and clinical progression may have reflected methodological limitations rather than a true biological dissociation. These findings support the use of serial DAT imaging as an objective biomarker for monitoring disease progression and tracking individualized disease trajectories [177]. In a longitudinal cohort of de novo PD patients followed for more than 5 years, lower baseline striatal DAT binding and greater early declines in DAT uptake were independently associated with worse long-term motor and nonmotor outcomes, including postural instability, cognitive impairment, psychosis, depressive symptoms, and disability, suggesting that DAT imaging may help stratify future disease risk [178]. Supporting the biological heterogeneity of PD, a longitudinal PPMI study demonstrated that clinically defined subtypes exhibit distinct patterns of regional brain atrophy and clinical progression. Patients with the diffuse-malignant subtype showed accelerated cognitive and motor decline accompanied by greater atrophy in the precuneus, temporal cortex, fusiform gyrus, and cerebellum [179].

5.1.3. Neuroimaging Biomarkers in Clinical Trials

Neuroimaging biomarkers have become increasingly important in the evaluation of disease-modifying therapies because they provide objective evidence of target engagement and biological response. The need for robust biomarkers has become a recurring theme in discussions of disease-modifying PD trials [180]. Several restorative therapies, including fetal mesencephalic transplantation and glial cell line-derived neurotrophic factor (GDNF)-based interventions, have demonstrated substantial improvements in dopaminergic imaging measures. PET studies frequently showed increased fluorodopa uptake after treatment, suggesting successful dopaminergic reinnervation and graft survival [181,182]. However, improvements in neuroimaging biomarkers have not always translated into proportional clinical benefit. This discrepancy highlights an important distinction between biomarkers that reflect disease biology and surrogate endpoints capable of predicting meaningful clinical outcomes. Motor disability in PD is influenced by multiple neurotransmitter systems and widespread neural network dysfunction that may not be fully captured by dopaminergic imaging alone [183,184]. Consequently, while neuroimaging remains invaluable for demonstrating treatment-related biological effects, current evidence does not support its use as a standalone surrogate outcome measure in disease-modification trials.

Lessons from Alzheimer's disease illustrate the transformative impact of pathology-specific imaging biomarkers on therapeutic development. Amyloid PET and tau PET imaging have enabled identification of biomarker-positive individuals before dementia onset, enrichment of clinical trial cohorts, reduction of sample-size requirements, and evaluation of biological treatment effects [185,186]. Similar approaches may ultimately be applied to synucleinopathies if reliable αSyn imaging becomes available. Biomarker-defined at-risk populations could facilitate secondary prevention trials by identifying individuals most likely to progress to clinically manifest disease, thereby improving trial efficiency and enhancing the likelihood of detecting disease-modifying effects.

5.2. Strengths and Limitations of Current Neuroimaging Biomarkers

Dopaminergic imaging remains the most validated and widely adopted imaging biomarker and is particularly useful for confirming nigrostriatal degeneration. Nevertheless, reductions in dopaminergic tracer uptake are not specific to PD and can be observed in atypical parkinsonian syndromes. Furthermore, changes in dopaminergic imaging do not always parallel clinical progression or treatment response [15,187]. Although individual MRI biomarkers each provide valuable information, they differ substantially in their biological specificity, clinical applicability, and longitudinal performance (Table 7). Neuromelanin-sensitive MRI appears most closely linked to dopaminergic neuronal loss, QSM provides quantitative assessment of iron accumulation, the swallow-tail sign offers a simple diagnostic marker of nigrosome-1 degeneration, and FW imaging reflects microstructural tissue alterations associated with neurodegeneration and neuroinflammation. These complementary strengths suggest that future MRI biomarker strategies will likely rely on integrated multimodal approaches rather than individual imaging metrics used in isolation [120,159]. No currently available neuroimaging biomarker fulfills all characteristics of the ideal biomarker. Consequently, the future of PD biomarker research will likely depend on combining imaging, fluid, genetic, and clinical biomarkers rather than relying on any single modality.

5.3. Multimodal Imaging Approaches

Increasing evidence suggests that multimodal imaging approaches may outperform individual imaging techniques. Studies combining neuromelanin-sensitive MRI, QSM, FW imaging, and nigrosome-1 assessment have demonstrated higher diagnostic accuracy than any single MRI biomarker alone [159,188]. Similarly, combining molecular imaging and structural MRI may provide complementary information regarding dopaminergic dysfunction, neuronal loss, iron accumulation, and network-level dysfunction. Such multimodal approaches align closely with emerging biological frameworks such as SynNeurGe and may improve patient classification, biological staging, risk prediction, and clinical trial stratification [13,14].

5.4. Improving Existing Biomarkers Through Advanced Quantitative Analysis

Recent efforts have focused on maximizing the performance of established imaging biomarkers through improved quantitative analysis. Re-analysis of DAT imaging data from the PPMI, encompassing thousands of scans from healthy controls, patients with PD, individuals with RBD, hyposmic subjects, and genetic cohorts, has demonstrated that optimized spatial normalization methods and alternative reference regions may substantially improve quantification of striatal binding ratios [161]. These refinements reduce measurement variability and improve sensitivity to longitudinal change, potentially enhancing the utility of DAT imaging in observational studies and disease-modifying clinical trials. Notably, improved analytic pipelines have been associated with higher signal-to-noise ratios and substantially reduced sample-size requirements for clinical trials compared with conventional approaches. These findings suggest that methodological innovation may provide benefits comparable to those achieved through development of entirely new biomarkers. As imaging biomarkers move toward regulatory qualification and broader clinical adoption, standardization and harmonization of image acquisition, processing, and analysis will become increasingly important [14].

Reanalysis of longitudinal DAT-SPECT datasets using advanced pipelines incorporating motion correction, template-space normalization, optimized reference-region selection, and refined quantification algorithms has been associated with markedly improved signal-to-noise characteristics and increased sensitivity to longitudinal change [189]. Importantly, these methodological refinements substantially reduced estimated sample-size requirements for disease-modifying clinical trials, suggesting that improvements in analytical methodology may produce gains comparable to those achieved through development of entirely new biomarkers [189].

6. Future Directions

6.1. Artificial Intelligence and Quantitative Imaging

Several technological advances are expected to shape the next generation of neuroimaging biomarkers. Machine learning, deep-learning-based image analysis, advanced radiomics, automated segmentation algorithms, and improved spatial normalization techniques are increasingly enabling extraction of high-dimensional biological information from multimodal imaging datasets, thereby enhancing sensitivity for the detection of subtle neurodegenerative changes during prodromal and early disease stages [190]. Beyond conventional region-of-interest analyses, these approaches can identify complex spatial and temporal imaging signatures associated with nigrostriatal degeneration, network dysfunction, and disease heterogeneity, potentially improving diagnostic accuracy, biological staging, and patient stratification for disease-modifying interventions. In parallel, quantitative imaging methodologies designed to reduce measurement variability and improve inter-scanner reproducibility may substantially enhance the performance of existing imaging biomarkers, facilitating their implementation in multicenter observational studies and clinical trials and accelerating regulatory qualification efforts [191].

The growing availability of large-scale longitudinal datasets from cohorts such as the PPMI has further catalyzed the development of artificial intelligence-driven biomarker frameworks. Graph-based analytical models and multimodal machine-learning approaches have demonstrated the ability to identify distinct progression subtypes, predict future clinical trajectories, and integrate neuroimaging data with genetic, molecular, and phenotypic information into unified disease models [192]. Such computational approaches are particularly attractive within emerging biologically based classification systems, as they may capture complex interactions among αSyn pathology, neurodegeneration, and genetic susceptibility that are difficult to characterize using individual biomarkers alone [193]. Ultimately, the convergence of artificial intelligence, quantitative imaging, and multimodal biomarker integration may enable individualized biological staging, earlier diagnosis, and more precise assessment of therapeutic target engagement in PD.

6.2. Emerging Imaging Technologies

Advances in imaging hardware may further improve the sensitivity of molecular imaging biomarkers. Digital PET systems offer superior detector performance and improved image quality compared with conventional scanners. More recently, total-body PET systems, such as the NeuroEXPLORER platform, have demonstrated substantially higher sensitivity and spatial resolution than traditional PET technology [194,195]. Early studies suggest that these systems may improve visualization of small brainstem structures and detect subtle molecular abnormalities that would otherwise be difficult to quantify using standard PET scanners. Such improvements could be particularly relevant for imaging nigral pathology and prodromal neurodegenerative processes [196].

6.3. Integration of Imaging and Fluid Biomarkers

Future biomarker strategies will likely combine neuroimaging with fluid, genetic, and digital biomarkers to achieve a more comprehensive characterization of disease biology. Among these approaches, αSyn seed amplification assays (SAA) have emerged as promising tools for detecting pathological αSyn in cerebrospinal fluid, skin, and other tissues [197]. Recent studies suggest that SAA-based measures may not only provide sensitive detection of underlying pathology but may also offer quantitative information regarding disease progression and individual risk trajectories. The combination of αSyn PET imaging with fluid-based biomarkers could provide complementary information, enabling large-scale screening using SAA followed by in vivo regional characterization using molecular imaging. Such multimodal approaches may be particularly useful for early diagnosis, biological staging, participant selection for prevention trials, and assessment of therapeutic target engagement. Experience from amyloid PET imaging suggests that successful translation of αSyn PET into clinical trials will require coordinated development of imaging biomarkers, fluid biomarkers, and standardized analytical frameworks [198]. Quantitative αSyn assays may provide scalable methods for identifying biologically affected individuals, whereas PET imaging offers unique information regarding the regional distribution and burden of pathology. Future studies should prioritize direct comparison of αSyn PET tracers, harmonization of image-processing pipelines, and integration of PET findings with SAA-based biomarkers [199]. Such convergence of fluid and imaging biomarkers may accelerate biological staging, participant selection for prevention trials, and validation of disease-modifying therapies.

6.4. Biomarker Harmonization

One of the major challenges limiting widespread implementation of neuroimaging biomarkers is variability introduced by differences in imaging protocols, scanners, radiotracers, and analytical pipelines (Table 8). Large multicenter initiatives, including the PPMI [200], have highlighted the importance of harmonized image processing methods and standardized quantitative metrics (Table 9). Future efforts may lead to biomarker-specific calibration frameworks analogous to the Centiloid system developed for amyloid PET imaging, enabling direct comparison of results across studies and institutions [31]. The development of biomarker harmonization frameworks such as Centamines represents an important step toward standardization of dopaminergic imaging. By creating tracer-independent scales that permit conversion of DAT-SPECT and VMAT2 PET measurements into common quantitative units, these approaches may facilitate multicenter studies, meta-analyses, and regulatory qualification of imaging biomarkers. Similar harmonization strategies have transformed the amyloid imaging field and may ultimately enable broader clinical implementation of molecular imaging biomarkers in PD [31]. The Centamine framework uses healthy subjects imaged with 123I-ioflupane SPECT as a reference population, allowing measurements obtained from different dopaminergic tracers to be converted into standardized quantitative units. Such harmonization may improve reproducibility, accelerate clinical trial implementation, and support wider adoption of dopaminergic imaging biomarkers in both research and clinical practice [33].

6.5. Precision Medicine and Biological Staging

The development of αSyn imaging may represent the most important unmet need in PD biomarker research. Analogous to the impact of amyloid and tau PET imaging in Alzheimer's disease, successful αSyn PET tracers could enable biomarker-defined enrollment strategies, facilitate secondary prevention trials, improve biological disease staging, and provide direct measures of therapeutic target engagement [13]. More broadly, integration of multimodal imaging with fluid biomarkers, genetics, and clinical phenotyping is expected to support the transition toward precision medicine in PD. Emerging biological classification systems, including the SynNeurGe framework (Table 10), emphasize characterization of disease according to underlying pathology rather than clinical manifestations alone. Neuroimaging biomarkers will likely play a central role in these efforts by providing objective measures of neurodegeneration, network dysfunction, and eventually αSyn burden. Such advances may enable earlier diagnosis, more accurate biological staging, individualized prognostic assessment, and improved selection of patients for targeted therapies [13,14].

7. Conclusions

Neuroimaging biomarkers have become indispensable tools for understanding PD biology and increasingly support diagnosis, biological staging, prognostication, and therapeutic development. Dopaminergic imaging remains the most established and clinically validated biomarker of nigrostriatal degeneration, while advances in MRI have provided complementary measures of neuronal loss, iron accumulation, microstructural alterations, and network dysfunction. Emerging molecular imaging approaches targeting cholinergic dysfunction, neuroinflammation, synaptic integrity, and αSyn pathology are expanding the scope of biomarker research beyond traditional dopaminergic paradigms. At the same time, advances in quantitative imaging, artificial intelligence, biomarker harmonization, and multimodal integration are improving the precision and clinical applicability of existing techniques. Future progress will depend not only on the development of novel biomarkers but also on the standardization, validation, and integration of imaging measures with fluid biomarkers and other biological data.