1
|
Weng T, Zheng Y, Xie Y, Qin W, Guo L. Diagnosing schizophrenia using deep learning: Novel interpretation approaches and multi-site validation. Brain Res 2024; 1833:148876. [PMID: 38513996 DOI: 10.1016/j.brainres.2024.148876] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/04/2023] [Revised: 02/28/2024] [Accepted: 03/18/2024] [Indexed: 03/23/2024]
Abstract
Schizophrenia is a profound and enduring mental disorder that imposes significant negative impacts on individuals, their families, and society at large. The development of more accurate and objective diagnostic tools for schizophrenia can be expedited through the employment of deep learning (DL), that excels at deciphering complex hierarchical non-linear patterns. However, the limited interpretability of deep learning has eroded confidence in the model and restricted its clinical utility. At the same time, if the data source is only derived from a single center, the model's generalizability is difficult to test. To enhance the model's reliability and applicability, leave-one-center-out validation with a large and diverse sample from multiple centers is crucial. In this study, we utilized Nine different global centers to train and test the 3D Resnet model's generalizability, resulting in an 82% classification performance (area under the curve) on all datasets sourced from different countries, employing a leave-one-center-out-validation approach. Per our approximation of the feature significance of each region on the atlas, we identified marked differences in the thalamus, pallidum, and inferior frontal gyrus between individuals with schizophrenia and healthy controls, lending credence to prior research findings. At the same time, in order to translate the model's output into clinically applicable insights, the SHapley Additive exPlanations (SHAP) permutation explainer method with an anatomical atlas have been refined, thereby offering precise neuroanatomical and functional interpretations of different brain regions.
Collapse
Affiliation(s)
- Tingting Weng
- School of Medical Imaging, Tianjin Medical University, Tianjin 300203, China
| | - Yuemei Zheng
- Department of Medical Imaging, Affiliated Hospital of Jining Medical University, Shandong 100038, China
| | - Yingying Xie
- Department of Radiology and Tianjin Key Laboratory of Functional Imaging, Tianjin Medical University General Hospital, Tianjin 300052, China
| | - Wen Qin
- Department of Radiology and Tianjin Key Laboratory of Functional Imaging, Tianjin Medical University General Hospital, Tianjin 300052, China
| | - Li Guo
- School of Medical Imaging, Tianjin Medical University, Tianjin 300203, China.
| |
Collapse
|
2
|
Momota Y, Bun S, Hirano J, Kamiya K, Ueda R, Iwabuchi Y, Takahata K, Yamamoto Y, Tezuka T, Kubota M, Seki M, Shikimoto R, Mimura Y, Kishimoto T, Tabuchi H, Jinzaki M, Ito D, Mimura M. Amyloid-β prediction machine learning model using source-based morphometry across neurocognitive disorders. Sci Rep 2024; 14:7633. [PMID: 38561395 PMCID: PMC10984960 DOI: 10.1038/s41598-024-58223-3] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/12/2023] [Accepted: 03/26/2024] [Indexed: 04/04/2024] Open
Abstract
Previous studies have developed and explored magnetic resonance imaging (MRI)-based machine learning models for predicting Alzheimer's disease (AD). However, limited research has focused on models incorporating diverse patient populations. This study aimed to build a clinically useful prediction model for amyloid-beta (Aβ) deposition using source-based morphometry, using a data-driven algorithm based on independent component analyses. Additionally, we assessed how the predictive accuracies varied with the feature combinations. Data from 118 participants clinically diagnosed with various conditions such as AD, mild cognitive impairment, frontotemporal lobar degeneration, corticobasal syndrome, progressive supranuclear palsy, and psychiatric disorders, as well as healthy controls were used for the development of the model. We used structural MR images, cognitive test results, and apolipoprotein E status for feature selection. Three-dimensional T1-weighted images were preprocessed into voxel-based gray matter images and then subjected to source-based morphometry. We used a support vector machine as a classifier. We applied SHapley Additive exPlanations, a game-theoretical approach, to ensure model accountability. The final model that was based on MR-images, cognitive test results, and apolipoprotein E status yielded 89.8% accuracy and a receiver operating characteristic curve of 0.888. The model based on MR-images alone showed 84.7% accuracy. Aβ-positivity was correctly detected in non-AD patients. One of the seven independent components derived from source-based morphometry was considered to represent an AD-related gray matter volume pattern and showed the strongest impact on the model output. Aβ-positivity across neurological and psychiatric disorders was predicted with moderate-to-high accuracy and was associated with a probable AD-related gray matter volume pattern. An MRI-based data-driven machine learning approach can be beneficial as a diagnostic aid.
Collapse
Affiliation(s)
- Yuki Momota
- Department of Neuropsychiatry, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-Ku, Tokyo, 160-8582, Japan
- Department of Functional Brain Imaging Research, Institute for Quantum Medical Science, National Institutes for Quantum Science and Technology, 4-9-1 Anagawa, Inage-Ku, Chiba-Shi, Chiba, 263-8555, Japan
| | - Shogyoku Bun
- Department of Neuropsychiatry, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-Ku, Tokyo, 160-8582, Japan.
| | - Jinichi Hirano
- Department of Neuropsychiatry, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-Ku, Tokyo, 160-8582, Japan.
| | - Kei Kamiya
- Department of Neuropsychiatry, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-Ku, Tokyo, 160-8582, Japan
| | - Ryo Ueda
- Office of Radiation Technology, Keio University Hospital, 35 Shinanomachi, Shinjuku-Ku, Tokyo, 160-8582, Japan
| | - Yu Iwabuchi
- Department of Radiology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-Ku, Tokyo, 160-8582, Japan
| | - Keisuke Takahata
- Department of Neuropsychiatry, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-Ku, Tokyo, 160-8582, Japan
- Department of Functional Brain Imaging Research, Institute for Quantum Medical Science, National Institutes for Quantum Science and Technology, 4-9-1 Anagawa, Inage-Ku, Chiba-Shi, Chiba, 263-8555, Japan
| | - Yasuharu Yamamoto
- Department of Functional Brain Imaging Research, Institute for Quantum Medical Science, National Institutes for Quantum Science and Technology, 4-9-1 Anagawa, Inage-Ku, Chiba-Shi, Chiba, 263-8555, Japan
| | - Toshiki Tezuka
- Department of Neurology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-Ku, Tokyo, 160-8582, Japan
| | - Masahito Kubota
- Department of Neurology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-Ku, Tokyo, 160-8582, Japan
| | - Morinobu Seki
- Department of Neurology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-Ku, Tokyo, 160-8582, Japan
| | - Ryo Shikimoto
- Department of Neuropsychiatry, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-Ku, Tokyo, 160-8582, Japan
| | - Yu Mimura
- Department of Neuropsychiatry, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-Ku, Tokyo, 160-8582, Japan
| | - Taishiro Kishimoto
- Psychiatry Department, Donald and Barbara Zucker School of Medicine, Hempstead, NY, 11549, USA
- Hills Joint Research Laboratory for Future Preventive Medicine and Wellness, Keio University School of Medicine, Mori JP Tower F7, 1-3-1 Azabudai, Minato-ku, Tokyo, 106-0041, Japan
| | - Hajime Tabuchi
- Department of Neuropsychiatry, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-Ku, Tokyo, 160-8582, Japan
| | - Masahiro Jinzaki
- Department of Radiology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-Ku, Tokyo, 160-8582, Japan
| | - Daisuke Ito
- Department of Physiology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-Ku, Tokyo, 160-8582, Japan
- Memory Center, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-Ku, Tokyo, 160-8582, Japan
| | - Masaru Mimura
- Center for Preventive Medicine, Keio University, Mori JP Tower 7th Floor, 1-3-1 Azabudai, Minato-ku, Tokyo, 106-0041, Japan
| |
Collapse
|
3
|
Deng L, Wei W, Qiao C, Yin Y, Li X, Yu H, Jian L, Ma X, Zhao L, Wang Q, Deng W, Guo W, Li T. Dynamic aberrances of substantia nigra-relevant coactivation patterns in first-episode treatment-naïve patients with schizophrenia. Psychol Med 2024:1-11. [PMID: 38523252 DOI: 10.1017/s0033291724000655] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 03/26/2024]
Abstract
BACKGROUND Although dopaminergic disturbances are well-known in schizophrenia, the understanding of dopamine-related brain dynamics remains limited. This study investigates the dynamic coactivation patterns (CAPs) associated with the substantia nigra (SN), a key dopaminergic nucleus, in first-episode treatment-naïve patients with schizophrenia (FES). METHODS Resting-state fMRI data were collected from 84 FES and 94 healthy controls (HCs). Frame-wise clustering was implemented to generate CAPs related to SN activation or deactivation. Connectome features of each CAP were derived using an edge-centric method. The occurrence for each CAP and the balance ratio for antagonistic CAPs were calculated and compared between two groups, and correlations between temporal dynamic metrics and symptom burdens were explored. RESULTS Functional reconfigurations in CAPs exhibited significant differences between the activation and deactivation states of SN. During SN activation, FES more frequently recruited a CAP characterized by activated default network, language network, control network, and the caudate, compared to HCs (F = 8.54, FDR-p = 0.030). Moreover, FES displayed a tilted balance towards a CAP featuring SN-coactivation with the control network, caudate, and thalamus, as opposed to its antagonistic CAP (F = 7.48, FDR-p = 0.030). During SN deactivation, FES exhibited increased recruitment of a CAP with activated visual and dorsal attention networks but decreased recruitment of its opposing CAP (F = 6.58, FDR-p = 0.034). CONCLUSION Our results suggest that neuroregulatory dysfunction in dopaminergic pathways involving SN potentially mediates aberrant time-varying functional reorganizations in schizophrenia. This finding enriches the dopamine hypothesis of schizophrenia from the perspective of brain dynamics.
Collapse
Affiliation(s)
- Lihong Deng
- Department of Neurobiology, Affiliated Mental Health Center & Hangzhou Seventh People's Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China
- Nanhu Brain-computer Interface Institute, Hangzhou, Zhejiang, China
- Liangzhu Laboratory, MOE Frontier Science Center for Brain Science and Brain-machine Integration, State Key Laboratory of Brain-machine Intelligence, Zhejiang University, Hangzhou, Zhejiang, China
- NHC and CAMS Key Laboratory of Medical Neurobiology, Zhejiang University, Hangzhou, Zhejiang, China
- Mental Health Center and Psychiatric Laboratory, State Key Laboratory of Biotherapy, West China Hospital of Sichuan University, Chengdu, Sichuan, China
| | - Wei Wei
- Department of Neurobiology, Affiliated Mental Health Center & Hangzhou Seventh People's Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China
- Nanhu Brain-computer Interface Institute, Hangzhou, Zhejiang, China
- Liangzhu Laboratory, MOE Frontier Science Center for Brain Science and Brain-machine Integration, State Key Laboratory of Brain-machine Intelligence, Zhejiang University, Hangzhou, Zhejiang, China
- NHC and CAMS Key Laboratory of Medical Neurobiology, Zhejiang University, Hangzhou, Zhejiang, China
| | - Chunxia Qiao
- Mental Health Center and Psychiatric Laboratory, State Key Laboratory of Biotherapy, West China Hospital of Sichuan University, Chengdu, Sichuan, China
| | - Yubing Yin
- Mental Health Center and Psychiatric Laboratory, State Key Laboratory of Biotherapy, West China Hospital of Sichuan University, Chengdu, Sichuan, China
| | - Xiaojing Li
- Department of Neurobiology, Affiliated Mental Health Center & Hangzhou Seventh People's Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China
- Nanhu Brain-computer Interface Institute, Hangzhou, Zhejiang, China
- Liangzhu Laboratory, MOE Frontier Science Center for Brain Science and Brain-machine Integration, State Key Laboratory of Brain-machine Intelligence, Zhejiang University, Hangzhou, Zhejiang, China
- NHC and CAMS Key Laboratory of Medical Neurobiology, Zhejiang University, Hangzhou, Zhejiang, China
| | - Hua Yu
- Department of Neurobiology, Affiliated Mental Health Center & Hangzhou Seventh People's Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China
- Nanhu Brain-computer Interface Institute, Hangzhou, Zhejiang, China
- Liangzhu Laboratory, MOE Frontier Science Center for Brain Science and Brain-machine Integration, State Key Laboratory of Brain-machine Intelligence, Zhejiang University, Hangzhou, Zhejiang, China
- NHC and CAMS Key Laboratory of Medical Neurobiology, Zhejiang University, Hangzhou, Zhejiang, China
| | - Lingqi Jian
- Mental Health Center and Psychiatric Laboratory, State Key Laboratory of Biotherapy, West China Hospital of Sichuan University, Chengdu, Sichuan, China
| | - Xiaohong Ma
- Mental Health Center and Psychiatric Laboratory, State Key Laboratory of Biotherapy, West China Hospital of Sichuan University, Chengdu, Sichuan, China
| | - Liansheng Zhao
- Mental Health Center and Psychiatric Laboratory, State Key Laboratory of Biotherapy, West China Hospital of Sichuan University, Chengdu, Sichuan, China
| | - Qiang Wang
- Mental Health Center and Psychiatric Laboratory, State Key Laboratory of Biotherapy, West China Hospital of Sichuan University, Chengdu, Sichuan, China
| | - Wei Deng
- Department of Neurobiology, Affiliated Mental Health Center & Hangzhou Seventh People's Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China
- Nanhu Brain-computer Interface Institute, Hangzhou, Zhejiang, China
- Liangzhu Laboratory, MOE Frontier Science Center for Brain Science and Brain-machine Integration, State Key Laboratory of Brain-machine Intelligence, Zhejiang University, Hangzhou, Zhejiang, China
- NHC and CAMS Key Laboratory of Medical Neurobiology, Zhejiang University, Hangzhou, Zhejiang, China
| | - Wanjun Guo
- Department of Neurobiology, Affiliated Mental Health Center & Hangzhou Seventh People's Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China
- Nanhu Brain-computer Interface Institute, Hangzhou, Zhejiang, China
- Liangzhu Laboratory, MOE Frontier Science Center for Brain Science and Brain-machine Integration, State Key Laboratory of Brain-machine Intelligence, Zhejiang University, Hangzhou, Zhejiang, China
- NHC and CAMS Key Laboratory of Medical Neurobiology, Zhejiang University, Hangzhou, Zhejiang, China
| | - Tao Li
- Department of Neurobiology, Affiliated Mental Health Center & Hangzhou Seventh People's Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China
- Nanhu Brain-computer Interface Institute, Hangzhou, Zhejiang, China
- Liangzhu Laboratory, MOE Frontier Science Center for Brain Science and Brain-machine Integration, State Key Laboratory of Brain-machine Intelligence, Zhejiang University, Hangzhou, Zhejiang, China
- NHC and CAMS Key Laboratory of Medical Neurobiology, Zhejiang University, Hangzhou, Zhejiang, China
| |
Collapse
|
4
|
King VL, Lahti AC, Maximo JO, ver Hoef LW, John S, Kraguljac NV. Contrasting Frontoparietal Network Connectivity in Antipsychotic Medication-Naive First-Episode Psychosis Patients Who Do and Do Not Display Features of the Deficit Syndrome. Schizophr Bull 2022; 48:1344-1353. [PMID: 35869578 PMCID: PMC9673254 DOI: 10.1093/schbul/sbac081] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 12/14/2022]
Abstract
BACKGROUND The deficit syndrome is a clinical subtype of schizophrenia that is characterized by enduring negative symptoms. Several lines of evidence point to frontoparietal involvement, but the frontoparietal control network (FPCN) and its subsystems (FPCNA and FPCNB) proposed by Yeo et al. have not been systematically characterized at rest in patients with the deficit syndrome. METHODS We used resting-state fMRI to investigate the FPCN and its subnetworks in 72 healthy controls and 65 antipsychotic medication-naive, first-episode psychosis patients (22 displayed deficit syndrome features, 43 did not). To assess whole-brain FPCN connectivity, we used the right posterior parietal cortex as the seed region. We then performed region of interest analyses in FPCN subsystems. RESULTS We found that patterns of FPCN dysconnectivity to the whole brain differed in patients who displayed deficit syndrome features compared with those who did not. Examining the FPCN on a more granular level revealed reduced within-FPCN(A) connectivity only in patients displaying deficit features. FPCNB connectivity did not differ between patient groups. DISCUSSION Here, we describe a neurobiological signature of aberrant FPCN connectivity in antipsychotic-naive, first-episode patients who display clinical features of the deficit syndrome. Importantly, frontoparietal subnetwork connectivity differentiated subgroups, where the FPCNA is selectively involved in patients with deficit features. Our findings add to the growing body of literature supporting a neurobiological distinction between two clinical subtypes of schizophrenia, which has the potential to be leveraged for patient stratification in clinical trials and the development of novel treatments.
Collapse
Affiliation(s)
- Victoria L King
- Department of Psychiatry and Behavioral Neurobiology, University of Alabama at Birmingham, Birmingham, AL, USA
| | - Adrienne C Lahti
- Department of Psychiatry and Behavioral Neurobiology, University of Alabama at Birmingham, Birmingham, AL, USA
| | - Jose O Maximo
- Department of Psychiatry and Behavioral Neurobiology, University of Alabama at Birmingham, Birmingham, AL, USA
| | - Lawrence W ver Hoef
- Department of Neurology, University of Alabama at Birmingham, Birmingham, AL, USA
| | - Sooraj John
- Department of Psychiatry and Behavioral Neurobiology, University of Alabama at Birmingham, Birmingham, AL, USA
| | - Nina V Kraguljac
- To whom correspondence should be addressed; Department of Psychiatry and Behavioral Neurobiology, University of Alabama at Birmingham, SC 501, 1720 7th Ave S, Birmingham, AL 35294-0017, USA; tel: 205-996-7171, e-mail:
| |
Collapse
|