Linking individual brain variability to clinical phenotypes for early detection of bipolar disorders
Poster No:
604
Submission Type:
Abstract Submission
Authors:
Junneng Shao1, Wei Zhang1, Ting Wang1, Qian Liao1, Cong Pei1, Lien Wang1, Shui Tian2, Zhilu Chen3, Zhijian Yao3, Qing Lu1
Institutions:
1School of Biological Sciences & Medical Engineering, Southeast University, Nanjing, China, 2Nanjing Medical University, Nanjing, China, 3Department of Psychiatry, the Affiliated Nanjing Brain Hospital of Nanjing Medical University, Nanjing, China
First Author:
Junneng Shao
School of Biological Sciences & Medical Engineering, Southeast University
Nanjing, China
School of Biological Sciences & Medical Engineering, Southeast University
Nanjing, China
Co-Author(s):
Zhilu Chen
Department of Psychiatry, the Affiliated Nanjing Brain Hospital of Nanjing Medical University
Nanjing, China
Department of Psychiatry, the Affiliated Nanjing Brain Hospital of Nanjing Medical University
Nanjing, China
Zhijian Yao
Department of Psychiatry, the Affiliated Nanjing Brain Hospital of Nanjing Medical University
Nanjing, China
Department of Psychiatry, the Affiliated Nanjing Brain Hospital of Nanjing Medical University
Nanjing, China
Introduction:
Bipolar disorder (BD) with a depressive episode and unipolar disorder (UD; i.e., major depressive disorder) share similar clinical profile (Phillips & Kupfer, 2013) and consequently, individuals with BD are commonly misdiagnosed as UD (de Almeida & Phillips, 2013). Misdiagnosis can lead to inappropriate treatment, poor prognosis and subsequent complication, such as increased risk of suicide and long-term medical costs (de Almeida & Phillips, 2013; Siegel-Ramsay et al., 2022). Evidence has been shown that clinical characteristics, such as age of onset, anxiety and cognitive, may be clinical precursors of BD (Bolton, Warner, Harriss, Geddes, & Saunders, 2021; Faedda et al., 2019). Therefore, we tried to find objective biomarkers that could distinguish BD from MDD early in the course of the disease by linking clinical phenotypes with brain neuroimaging.
Methods:
A total of 166 healthy controls, 184 UD patients, 148 BD patients, and 72 patients who were initially strictly diagnosed as UD during scanning and then transformed to BD during follow-up (tBD) were enrolled in the study. Firstly, to exclude the effects of age and gender, we constructed a normative model of brain changes with age and gender based on a large publicly available dataset of HC (N=1112). Then, the individual deviation of each patient in brain structure (GMV) and function (ALFF) could be obtained from these normative models (Figure 1c). Secondly, we included age of onset, family history of psychosis, number of episodes, and five subfactor scores of HAMD-17 as clinical characteristics. Here, we combined sparse multivariate canonical correlation analysis (smCCA) with joint independent component analysis (joint ICA) to identify latent brain structure-function patterns that correlate clinical characteristics and whole-brain personalization deviations (z-scores) across all patients (Figure 1d). Finally, we compared the distribution of these latent patterns across the three groups of patients.
·Figure 1. Schematic of the study. (a) All participants in this study underwent structural and functional MRI scans. (b) The images were preprocessed using CAT12 and DPABI. (c) Normative model of brain
Results:
Our analysis revealed two latent patterns (Figure 2): 1) Pattern 1 was associated with age at onset. This pattern mainly involved the high ALFF deviation of limbic and subcortical network in functional brain; low GMV deviation of visual network and high GMV deviation of executive control network and temporoparietal network in structural brain. 2) Pattern 2 was associated with retardation, cognitive impairment and anxiety/somatization. This pattern mainly involved high ALFF deviation of visual and dorsal attention network, and low ALFF deviation of ventral attention and subcortical network in functional brain; high GMV deviation of default mode network A, subcortical network and low GMV deviation of visual network and default mode network B. The results of one-way ANOVA showed that there were significant differences in the subject weights of pattern GMV_1 (p = 0.0015), GMV_2 (p = 0.0067) and ALFF_2 (p = 0.0369) among the three groups above (FDR correction). Post-hoc analysis results showed that there were significant differences in these three modes between BD and UD group (p = 2.8e-04, p = 0.0048, p = 0.0143, respectively). Furthermore, we found that there is a significant difference in pattern GMV_2 between the tBD and UD group (p = 0.0339), but no difference between the tBD and BD group.
·Figure 2. The two latent patterns. (a) The contribution of the clinical phenotypes. (b) The contribution of brain structural-functional images were summarized by 18 networks and their corresponding su
Conclusions:
We revealed two latent brain structural-functional patterns associated with clinical phenotypes in patients with UD and BD. Our study showed that these patterns can help identify the differences between UD and BD patients, and hold the promise of timely identification of BD patients early in the course of the disease.
Disorders of the Nervous System:
Psychiatric (eg. Depression, Anxiety, Schizophrenia) 1
Modeling and Analysis Methods:
Multivariate Approaches 2
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
Anatomy and Functional Systems
Keywords:
Affective Disorders
FUNCTIONAL MRI
Machine Learning
Psychiatric Disorders
STRUCTURAL MRI
1|2Indicates the priority used for review
Provide references using author date format
de Almeida, J. R. C., & Phillips, M. L. (2013). Distinguishing between Unipolar Depression and Bipolar Depression: Current and Future Clinical and Neuroimaging Perspectives. Biological Psychiatry, 73(2), 111-118. doi:10.1016/j.biopsych.2012.06.010
Faedda, G. L., Baldessarini, R. J., Marangoni, C., Bechdolf, A., Berk, M., Birmaher, B., . . . Correll, C. U. (2019). An International Society of Bipolar Disorders task force report: Precursors and prodromes of bipolar disorder. Bipolar Disord, 21(8), 720-740. doi:10.1111/bdi.12831
Phillips, M. L., & Kupfer, D. J. (2013). Bipolar disorder diagnosis: challenges and future directions. Lancet, 381(9878), 1663-1671. doi:10.1016/S0140-6736(13)60989-7
Siegel-Ramsay, J. E., Bertocci, M. A., Wu, B., Phillips, M. L., Strakowski, S. M., & Almeida, J. R. C. (2022). Distinguishing between depression in bipolar disorder and unipolar depression using magnetic resonance imaging: a systematic review. Bipolar Disorders, 24(5), 474-498. doi:10.1111/bdi.13176