Clinically grounded
Research questions begin with biological and clinical relevance.
Medical imaging · Neuroimaging · Artificial intelligence
Turning complex brain data into clinically meaningful insight.
I am Mingshi Chen (陈明诗), a PhD researcher at Amsterdam UMC. My work combines medical imaging, machine learning, and clinical knowledge to study brain health and build methods that remain useful beyond the model itself.
01Clinical context
02Quantitative imaging
03Responsible AI
About
A clinical perspective on computational research.
Mingshi Chen
陈明诗
PhD Researcher · Amsterdam UMC
My background spans clinical medicine, medical imaging, and data science. That combination shapes the questions I ask: not only whether a model performs well, but whether its output is interpretable, reproducible, and meaningful for patients and researchers.
At Amsterdam UMC, I work with structural MRI, diffusion MRI, functional MRI, radiomics, brain-age modelling, and longitudinal statistics. I enjoy building complete research pipelines—from imaging quality control and feature engineering to validation, visualization, and scientific communication.
Methodologically careful
Validation, sensitivity analysis, and clear limitations are part of the result.
Built to communicate
Complex analyses should become understandable evidence, not a black box.
Research themes
Questions I am working on.
My projects connect quantitative neuroimaging with cognition, psychiatric treatment, substance exposure, and individual differences.
01
Brain age & individual differences
Evaluating multiple grey- and white-matter brain-age models and asking how language experience relates to apparent brain ageing.
02
Longitudinal neuroimaging
Studying structural brain changes before and after incident XTC exposure, including regional volumes, whole-brain patterns, dose, and memory.
03
Prediction in psychiatry
Testing whether early clinical and task-fMRI signals can improve treatment outcome prediction in major depressive disorder.
Selected work
Research across methods and clinical questions.
2026
Peer-reviewed publication · Scientific Reports
On the value of radiomics in addition to clinical measures in emotional conflict fMRI for predicting sertraline response in major depressive disorder
A preregistered EMBARC analysis comparing clinical, regional mean-activation, radiomic, and whole-brain task-fMRI features for prediction of sertraline and placebo outcomes.
2025
Peer-reviewed publication · NeuroImage: Clinical
Prediction of methylphenidate treatment response for ADHD using conventional and radiomics T1 and DTI features
A secondary analysis of a randomized clinical trial evaluating conventional MRI measures and radiomics with nested cross-validation in stimulant-naive children and adults with ADHD.
2025
Book chapter · Academic Press
What Is beyond the Image? Machine Learning for MR Image Analysis
An overview of radiomics, machine learning, representation learning, and their application to extracting clinically useful information from magnetic resonance imaging.
2024
Peer-reviewed publication · Cancer Imaging
Intracranial metabolic score in primary central nervous system lymphoma using PET/CT and PET/MR
Contributed to multimodal metabolic imaging research in primary central nervous system lymphoma.
2024
Peer-reviewed publication · Cancers
Pontine-white matter score in primary central nervous system lymphoma
Contributed to MRI-based characterization of primary central nervous system lymphoma.
2023
Peer-reviewed publication · Radiotherapy & Oncology
Endonasopharyngeal ultrasound and MRI of recurrent retropharyngeal lymph nodes
Contributed to work comparing complementary imaging approaches in recurrent retropharyngeal disease.
Current
Manuscript
Association of bilingual language use with brain age: MRI evidence from bilinguals, translators, and interpreters
Examines structural MRI-derived brain age across bilingual phenotypes with different levels of professional language experience and language-control demand.
Current
PhD research
Dose-dependent longitudinal whole-brain structural change associated with XTC exposure
A prospective NeXT-cohort study combining behavioral replication, predefined hippocampal and thalamic analyses, baseline-derived PCA, mixed-effects modelling, and cumulative XTC dose.
Current
Collaborative manuscript
MRI radiomics to predict return-to-sport after hamstring injury
A collaborative application of quantitative MRI and predictive modelling to sports-injury outcome.
For the complete and most current publication record, see Google Scholar or download my CV.
Experience & toolkit
From images to evidence.
2022 — present
PhD Researcher · Amsterdam UMC
Neuroimaging, medical AI, psychiatric imaging, longitudinal modelling, scientific writing, and student supervision.
2014 — 2022
Clinical medicine & imaging and nuclear medicine
BMed training in clinical medicine followed by an MClinMed in imaging and nuclear medicine at Sun Yat-sen University and SYSU Cancer Center.
Selected methods and tools
Imaging
Structural MRIDTIfMRI FreeSurferFastSurferFSLData science
PythonPyTorchMONAI scikit-learnstatsmodelsRAnalysis
RadiomicsBrain ageMixed models Cross-validationVisualizationContact
Let’s connect research, imaging, and clinical impact.
I am interested in postdoctoral research and medical-AI roles involving neuroimaging, model validation, clinical translation, or multimodal data.