VBM and DTI are prominent MRI techniques utilized to assess brain structural changes in neurodegenerative disorders, including PD and its cognitive manifestations such as PDD and mild cognitive impairment (MCI) (Pan et al., 2013; Wu et al., 2018). VBM quantifies gray matter volume and morphology, while DTI provides insights into white matter microstructural integrity, making them essential tools for exploring the neuropathological underpinnings of PD-CI.

VBM enables the quantitative analysis of gray matter volume, providing detailed insights into localized cortical atrophy in PD-CI (Wu et al., 2018). Its strength lies in its capacity to precisely identify structural abnormalities related to cognitive decline. However, VBM results may be influenced by the inherent variability in image preprocessing and normalization protocols, which can affect cross-study comparisons (Wu et al., 2018). Additionally, VBM is limited in its ability to investigate white matter changes, thus providing an incomplete picture of the neural degeneration in PD-CI (Wu et al., 2018).

DTI offers valuable insights into white matter microstructure, which is often compromised in PD-CI (Zhang et al., 2015). Its ability to detect subtle microstructural changes, even in the absence of significant gray matter atrophy, makes DTI particularly useful for identifying early markers of cognitive decline (Agosta et al., 2014). However, the technique’s sensitivity to motion artifacts and its limitations in resolving complex fiber architecture, such as in areas of crossing fibers, can complicate data interpretation. Despite these limitations, DTI serves as a critical complement to VBM, providing a fuller understanding of the structural underpinnings of PD-CI.

Zhang et al. (2015) applied VBM to compare gray matter volumes between PD patients with and without MCI and healthy controls. The study identified extensive cortical atrophy in PD patients relative to controls, particularly in the temporal, occipital, parietal, and frontal lobes, as well as the right insular cortex, cerebellar posterior lobe, and left caudate nucleus. In PD-MCI patients, significant gray matter reduction was observed in regions including the bilateral cingulate gyrus, lingual gyrus, left anterior cingulate cortex, insula, right superior temporal gyrus, orbitofrontal cortex, central gyrus, and precuneus, highlighting these areas as potential contributors to cognitive decline in PD.

Among hospitalized PD patients, those with MCI exhibited further gray matter reduction in the anterior cingulate gyrus, middle temporal gyrus, right precuneus, orbitofrontal cortex, and left cingulate gyrus compared to PD patients without MCI. These findings suggest a pattern of cortical atrophy that may underlie cognitive impairment in PD, particularly implicating the temporoparietal association cortex, right orbitofrontal cortex, and middle temporal gyrus as critical regions involved in the cognitive deficits observed in PD-MCI.

Stewart et al. (2023) investigated the association between cognitive impairment, olfactory dysfunction, and white matter integrity in early PD patients using DTI. The study focused on the anterior olfactory structures (AOS), finding that PD-MCI patients had significantly poorer olfactory function and exhibited abnormalities across all DTI metrics in the AOS compared to PD patients without cognitive impairment. These results imply that olfactory deficits and disrupted white matter integrity in olfactory regions could serve as early biomarkers for cognitive decline in PD-MCI (Figure 1).

The combined findings from these studies underscore the role of widespread gray matter atrophy and white matter microstructural changes as key features of cognitive impairment in PD. The identification of specific brain regions associated with these changes enhances our understanding of the neural substrates of PD-CI and suggests potential targets for future research and therapeutic intervention. Further exploration of these neuroimaging biomarkers could contribute to earlier diagnosis and more personalized management strategies for cognitive impairment in PD.

FMRI is a crucial tool for investigating the brain functional alterations associated with PD-CI (Baggio et al., 2014). Two primary fMRI modalities are commonly employed in research: resting-state fMRI (rs-fMRI) and task-based fMRI (ts-fMRI). Each modality offers distinct insights into the neural mechanisms underlying cognitive decline in PD, making them complementary methods in understanding PD-CI.

VBM is an imaging analysis technique that quantifies differences in brain anatomy by measuring gray matter volume across individuals (Sivaranjini and Sujatha, 2024). It is particularly effective in identifying structural changes, such as atrophy in brain regions, in patients with neurodegenerative diseases, including PD-CI (Sivaranjini and Sujatha, 2024).

DTI is another MRI technique that assesses the diffusion of water molecules in brain tissue, allowing for the evaluation of white matter integrity (Wang et al., 2020). This method is valuable for detecting subtle changes in white matter, which are often associated with cognitive decline in PD (Wang et al., 2020).

In the context of functional brain activity, amplitude of low-frequency fluctuations (ALFF) measures the amplitude of spontaneous low-frequency fluctuations during rs-fMRI (Wang et al., 2018). This analysis detects regional brain activity changes, which are particularly relevant in assessing cognitive impairments in PD-CI (Wang et al., 2018).

Regional homogeneity (ReHo) evaluates the synchronization of brain activity across neighboring voxels, providing insights into local brain connectivity. Alterations in ReHo can signal disruptions in functional networks, which are commonly observed in PD-CI (Li et al., 2020).

Functional connectivity (FC) assesses the temporal correlation between spatially distant brain regions, highlighting how different parts of the brain communicate. FC analysis is particularly useful in identifying network disruptions, a hallmark of PD-CI (Amboni et al., 2015).

Rs-fMRI captures spontaneous brain activity while the subject is at rest, relying on blood oxygen level-dependent (BOLD) signal fluctuations to assess intrinsic brain networks like the default mode network (DMN) (Day et al., 2019). Rs-fMRI is highly effective for detecting disruptions in FC within these networks, which are often implicated in PD-CI. Additionally, rs-fMRI can identify compensatory mechanisms, where increased activity in specific regions reflects the brain’s attempts to counterbalance cognitive deficits (Kaut et al., 2020). Since rs-fMRI does not rely on task performance, it is particularly sensitive to early functional changes that might go undetected during cognitive tasks.

On the other hand, ts-fMRI examines brain activation patterns in response to specific cognitive tasks or stimuli. It provides a more targeted approach, allowing researchers to observe how task-related brain regions function and how these processes deteriorate in PD-CI (Baggio and Junqué, 2019). Ts-fMRI is particularly useful for isolating the neural dynamics related to specific cognitive functions, such as memory or attention. While both rs-fMRI and ts-fMRI can detect compensatory mechanisms, rs-fMRI tends to be more sensitive to early intrinsic changes since it captures spontaneous brain activity without the need for external stimuli (Baggio and Junqué, 2019).

Key analysis methods used in rs-fMRI include ALFF, ReHo, and FC. These metrics help provide a comprehensive view of how brain regions synchronize and communicate, thus revealing detailed patterns of how functional networks are disrupted in PD-CI (Sun et al., 2021; Ten Kate et al., 2018; Hao et al., 2022; Li et al., 2021) (Figure 1). In contrast, ts-fMRI focuses on task-induced activations, which helps pinpoint which specific brain functions are compromised by PD (Kang et al., 2021) (Figure 1).

Despite the valuable insights both modalities offer, they have limitations. Rs-fMRI excels in detecting intrinsic network dysfunction without external stimuli, making it particularly useful for identifying early cognitive impairments (Lemée et al., 2019). Ts-fMRI, however, provides a focused analysis of brain activity during specific tasks. Both modalities face challenges, such as fMRI’s relatively low spatial resolution, which may obscure subtle neural activity (Smith et al., 2011). Moreover, they are susceptible to artifacts from head motion and physiological noise, potentially reducing the reliability of connectivity measurements (Smith et al., 2011). Additionally, interpreting fMRI data is complex due to the intricate interactions between various brain regions and networks (Smith et al., 2011).

Nevertheless, fMRI remains an indispensable tool for understanding the dynamic functional changes that accompany cognitive decline in PD. It provides valuable insights into the underlying neural mechanisms of PD-CI, supporting ongoing advancements in research and clinical practice.

A significant limitation is the lack of longitudinal studies that compare different stages of PD-CI. Such studies are crucial for understanding the progression of cognitive impairment over time and how it may be influenced by compensatory mechanisms or the effects of treatment. Future research should aim to include longitudinal designs to better capture the dynamic changes in brain structure and function associated with the progression of PD-CI. Specifically, studies could focus on tracking individual patients over several years, monitoring how their brain connectivity patterns change, and identifying early biomarkers for cognitive decline. Additionally, longitudinal imaging studies combined with cognitive assessments may help differentiate the trajectories of MCI from those leading to dementia, thus informing personalized therapeutic strategies.

There is a notable scarcity of ts-fMRI studies focused on PD-CI conducted by both domestic and international researchers. Ts-fMRI is valuable for assessing brain activation patterns during specific cognitive tasks, yet its application in PD-CI remains underexplored. Future studies should prioritize ts-fMRI experiments, particularly those examining the effects of cognitive tasks, attention, and motor training, to provide a more comprehensive understanding of functional alterations in PD-CI. A potential direction would be to design ts-fMRI protocols that target specific cognitive domains, such as working memory, executive function, and visuospatial abilities, which are frequently impaired in PD-CI. Comparing brain activation patterns across different cognitive loads could reveal the compensatory mechanisms engaged by PD-CI patients and inform tailored interventions that strengthen cognitive resilience.

Different MRI techniques, such as VBM, DTI, rs-fMRI, and ts-fMRI, each offer unique insights into brain structure and function in PD-CI. However, these techniques emphasize different aspects of brain alterations, and there is currently no consensus on which single technique or combination of techniques provides the most accurate predictive value for cognitive impairment in PD. Future research should focus on systematically comparing these techniques or exploring integrated approaches to identify the most effective diagnostic and prognostic tools for PD-CI. Multimodal imaging studies that combine structural and functional MRI data could provide more comprehensive biomarkers for early cognitive decline in PD. Moreover, machine learning algorithms could be employed to integrate data from multiple modalities and enhance diagnostic accuracy.

As MRI technology continues to evolve, there is significant potential for further advancements in the diagnosis and treatment of PD-CI. Enhanced imaging resolution, more sophisticated analysis methods, and the development of multimodal imaging approaches could provide deeper insights into the neurobiological mechanisms underlying cognitive impairment in PD. Future research should leverage these technological advancements to improve the clinical application of MRI in the management of PD-CI. For example, high-field MRI scanners could offer more detailed structural and functional brain images, allowing for the detection of subtle changes that are currently missed by conventional MRI. Additionally, the integration of MRI with other imaging modalities, such as positron emission tomography, may help identify neurochemical changes alongside structural and functional alterations in the brain.

Addressing these limitations will be critical for advancing our understanding of PD-CI and improving diagnostic and therapeutic strategies through the continued development of MRI technology.