Consultant for Machine Learning and Data Science in Neuroimaging
I’m Taha Parsayan, a machine learning and data science engineer in Neuroimaging (PhD). I provide consultancy and coaching for Neuroimaging projects. I am also open to collaborations, where I implement the data analysis and image processing aspects of your project.
Project Design & Scientific Writing
Designing a project starts with defining the pipeline, writing a proposal, and preparing an ethical application, and ends with writing a paper, thesis, or report.
Project Design
Learning how to do pre-processing of the MRI and PET images using OPETIA, SPM12, and FreeSurfer.
Additionally, learning Python programming can help with the automation of some of the data processing steps.
Image Processing
Processing of PET and MRI images such as co-registration, normalization to the standard space, tissue segmentation, and ROI definition.
Feature Extraction
Quantification of MRI and PET images and extraction of features such as cerebral volume, SUV, and SUVR from anatomical ROIs across the brain.
Advanced Statistical Analysis
Statistical analysis implementation on the data, such as one/two-sample t-test, effect size measurements, Cohen's D measurement, linear regression, ANOVA, ANCOVA, etc.
Machine Learning / Deep Learning
Application of deep models for classification of the data, such as SVM, NN, CNN, MLP, ViT, ResNet, MONAI, etc.
📬Contact Information
Feel free to reach out via email or linkedin to discuss your project or collaboration ideas.
When contacting me, please include a brief overview of:
Your project goals
The type of data you're working with
Any preferred tools or methods
How you think I can support you
💻 Neuroimaging Toolkits: Teaching & Tutorials
I offer teaching and tutorials on leading neuroimaging tools, including SPM12, FreeSurfer, and my own software, OPETIA. Whether you're new to these platforms or looking to deepen your expertise, I provide step-by-step guidance tailored to your project needs, from pre-processing and segmentation to statistical analysis and machine learning integration.
🧠Preprocessing of PET/MRI Images
Robust pre-processing is essential to ensure accuracy and reproducibility. I offer end-to-end support for the following key preprocessing steps:
Brain Extraction(skull removal)Co-registrationof PET images to T1-weighted MRI spaceSegmentationinto gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF)Normalizationto a standard template space for group-level analysisRegion of Interest (ROI) Segmentationtailored to your study design
📊 Feature Extraction
In most classification and pattern recognition pipelines, it's crucial to convert neuroimaging data into meaningful, quantitative features. Key features include:
Cerebral Volumemeasurements for each region of interest (ROI)Standardized Uptake Values (SUV)derived from PET imaging
📈 Advanced Statistical Analysis
Once preprocessing and feature extraction are complete, statistical analysis becomes the cornerstone of your neuroimaging project, revealing meaningful patterns and validating your findings. I can assist with a wide range of statistical methods, including:
T-teststo identify regions with significant group differencesLevene’s Testto assess equality of variancesShapiro–Wilk Testfor normality checksAnalysis of Variance (ANOVA)andAnalysis of Covariance (ANCOVA)for multi-group comparisonsBonferroni Correctionaccounts for multiple comparisons
🤖 Machine Learning & Classification
To uncover diagnostic patterns and predictive insights, I offer end-to-end machine learning pipelines tailored to neuroimaging data. I can help with:
Logistic RegressionSupport Vector Machines (SVM)Multilayer Perceptrons Neural Networks (MLPs)Computer vision (
ViT,ResNET,MONAI)Model Evaluation using
cross-validation,confusion matrices,ROC curves, and moreInterpretability
feature importanceand region-level impact analysis
🧾 Project Design & Scientific Writing
Beyond technical implementation, I provide comprehensive research consulting to support your project from concept to publication. With my academic and industry experience, I can assist in:
Study Design & Data Collection Planning– including defining inclusion/exclusion criteria, choosing modalities, and aligning data with your research questionsSample Size Calculation– using tools like G*Power or Python-based approaches to ensure statistical power for group comparisons or classification modelsEthical Applications– preparing submissions for institutional review boards (IRB/REC), ensuring clarity and complianceGrant Proposals & Research Proposals– helping you articulate objectives, methodology, expected impact, and budget justificationsManuscript Writing & Publication Support– from structuring scientific papers to polishing content and figures for peer-reviewed journals
I'm Taha Parsayan — a PhD researcher specializing in machine learning and data science in Neuroimaging at the University of Southern Denmark.I developed OPETIA as a full-stack developer. OPETIA is a toolbox for automatic processing of PET and MRI images and extraction of image features.My PhD project included developing and training deep learning models (MLP and MLR) for multi-class classification of Alzheimer's Disease (AD) vs. Mild Cognitive Impairment (MCI) vs. Normal Cognition (NC) using [18]F-FDG-PET and MRI T1-W images from the ADNI database.
Rates & Services
| Title | Fee per hour |
|---|---|
| Consultation | $20 |
| Consultation for students | $15 |
| Ethics, proposal, thesis, manuscript reviewing and editing | $20 |
| Image pre-processing | $25 |
| Statistical analysis | $25 |
| Visualization of data, results, etc | $20 |
| Developing and training machine learning | $30 |
| End-to-end pipeline (preproc + ML + report) | $50 |