The use of neuroimaging has provided many new understandings of the pathological basis of mental disorders. This highlights neuroimaging's great capacity to act as diagnostic tools, predictors of how diseases will develop, and gauges of whether treatments are working (Gerlach et al., 2022; Pilmeyer et al., 2022; Tu, 2021). Various mental disorders demonstrate distinct neurological differences visible through imaging methods including task-fMRI, resting state-fMRI, and structural scans as compared to healthy individuals. For example, major depressive disorder (MDD) and schizophrenia (SZ) show variations in cortical thickness, amygdala and hippocampus shape/size, and white matter condition (Deming et al., 2022; Dunlop & Mayberg, 2017; Sobral et al., 2022). A recent review of research on mental disorders indicates the presence of structural abnormalities in brain areas such as the inferior temporal gyrus, cingulate gyrus, inferior and middle frontal gyri, auditory cortex, insular cortex, and hippocampus (Cattarinussi et al., 2022; Deming et al., 2022). These imaging findings offer a more precise and diverse approach to diagnose and predict future mental disorders. Furthermore, MRI scans that incorporate multiple sequences provide a more detailed mapping of the underlying pathophysiology of mental disorders.

Despite significant advancements in MRI studies, the focus has primarily been on examining functional and structural changes in cortical and some subcortical regions of the brain. However, limited knowledge exists regarding the primary sources of neurotransmitters, such as the substantia nigra (SN) and locus coeruleus (LC). These regions form the cortex-subcortical ganglion loop with the thalamus and cortex and release neurotransmitters such as dopamine (DA) or Norepinephrine (NE) that project widely within the brain (Greene et al., 2020; Gunaydin & Kreitzer, 2016; Lacerda et al., 2003), as depicted in Figure 1. Dysfunction of these neural nuclei can disrupt neurotransmitter metabolism, transport, and interactions. This may consequently precipitate abnormalities throughout neural circuits and higher-order cortical functions (Ramos & Arnsten, 2007; Robbins & Arnsten, 2009), potentially serving as the initial starting point for the pathogenesis of mental disorders, such as depression (Graybiel, 1990; Jiang et al., 2022), anxiety (Goddard et al., 2010; Gong et al., 2021; McCall et al., 2015), schizophrenia (Malik et al., 2023), and addiction (Koob & Volkow, 2010; Volkow et al., 2009). Currently, assessing the monoamine systems non-invasively remains challenging, with dopamine function primarily evaluated through animal models and positron emission tomography (PET) (Cervenka et al., 2022; Gu et al., 2023; Urban et al., 2012; Wong et al., 1986). Nonetheless, the invasive and specialized features of PET hinder its practicality for clinical use and research, especially when considering pediatric cohorts and conducting repeated assessments to track disease progression and treatment response (Pizzagalli et al., 2019). Regarding the LC, its widespread projection patterns and small size present challenges for its precise localization and comprehensive study, given that its nucleus is located in the brainstem.

Previously developed NM-MRI sequences are capable of generating high-resolution T1-weighted neuromelanin-sensitive MRI (NM-MRI), offering a promising avenue for exploring alterations in the DA and NE systems in mental disorders (Sasaki et al., 2006; Tosk et al., 1992). Early studies have indicated that neuromelanin (NM) is a byproduct of DA and NE metabolism, primarily located in the dopamine neurons of the SN and the norepinephrine neurons of the LC in the human brain.(Sulzer et al., 2000; Wakamatsu et al., 2003, 2012; Zucca et al., 2017). NM-MRI contrast primarily stems from T1 effects and MT effects, iron within NM and its formation of complexes with ferritin may cause periodic changes in T1 signals and MT effects, both of which significantly influence imaging interpretations (Sasaki et al., 2006; Zucca et al., 2017). Recent Experiments demonstrate that NM-Fe complexes significantly influence MRI relaxation times and exchange processes. The physical structure of NM is found to affect the signal. Findings suggest that alterations in T1 and T2 induce opposite effects on DS linewidth, while MT-related effects affect the longitudinal magnetization component at higher frequencies(Priovoulos et al., 2020; T. Watanabe, 2023). In areas equivalent to NM-rich SN, the overall physical principles underlying NM-MRI's high signal involve the complex impact of NM's distinctive structure on relaxation and MT effects, culminating in the prominent high signal presentation of NM-rich regions (Sulzer et al., 2018; Trujillo et al., 2017). Tissues containing NM emit strong signals in NM-MRI sequences, as depicted in Figure 2, and can serve as a proxy measurement for the SN and LC, as they exhibit contrast with the surrounding tissues (Keren et al., 2015; Trujillo et al., 2017; Zecca et al., 1996).

Previous studies have demonstrated that the NM-MRI modality can effectively uncover notable differences in neuromelanin (NM) content between patients diagnosed with neurodegenerative diseases and healthy control groups (Blazejewska et al., 2013; Schwarz et al., 2017). Autopsy reports have additionally confirmed that the signal intensity of NM can serve as a relatively consistent biomarker for neurodegenerative diseases (Cho et al., 2021; Gaurav et al., 2021; Hou et al., 2022; Sulzer et al., 2018). Recent experimental evidence highlights significant differences in NM-MRI signals between patients with mental disorders and HC in the LC and SN. Additionally, NM-MRI displays variations among healthy individuals, with contrast-to-noise ratio (CNR) associated with stress (Bachman et al., 2022), arousal state (Mather et al., 2020), memory (Clewett et al., 2018; D et al., 2018; Dahl et al., 2019),cognitive reserve (Clewett et al., 2016), Inhibitory Control (Tomassini et al., 2022) and several cognitive features related to DA or LC function (Elman et al., 2021; Hussain et al., 2023). These findings emphasize the prospective utility of NM-MRI in detecting deviations in the DA and NE systems in mental disorders. However, NM-MRI's application in mental disorders is still exploratory, lacking a comprehensive systematic review encompassing findings across all mental disorders. Moreover, there is no consensus on the sequence parameter characteristics and analysis methods for the best proxy measures of DA and NE function using NM-MRI. Thus, reliable and valid recommendations are imperative to address the shortcomings of extant studies and assess the use of NM-MRI as a biomarker for mental disorders.

In this systematic review, we first comprehensively summarize the new findings of current research related to NM-MRI in mental disorders. Then we compare different NM-MRI sequence parameters and analytical methods. Finally, we critically evaluate the potential effects of different sequence parameters on outcomes, and provide recommendations to improve the reliability and validity of future NM-MRI applications in mental disorders.