Azar Kazemi
München
Summary
I am a medical informatician with ten years experience in medical data science. With a background in computer science and medical informatics, I have gained valuable experience in interdisciplinary data management, exploration, and analysis. Understanding of real-world healthcare data, combining with scientific problem solving, data interpretation and methodological skills, has empowered me to achieve significant results in data mining within medical domain. I am a detail-oriented person skilled in interpersonal orientation, ensuring smooth collaborative tasks. As a team player, I am committed to delivering data solutions that meet life science and medicine needs and drive informed decision-making.
Overview
11
11
years of professional experience
1
1
Certification
Work History
Researcher
Technical University of Munich
03.2022 - Current
- Applied deep-learning models for early detection of esophageal cancer and Barrett esophagus disease subtyping using 3600 histopathological images
- Weakly supervised learning: attention-based multiple instance learning and Mamba, and enhanced model performance by performing hyperparameter tuning and managing class imbalances
- Implemented feature extraction by using pre-trained pathology foundation models
- Preprocessing of clinical and image datasets, ensuring high quality, and patching images for model training
- Utilized Python, Docker, and command line interface in HPC environment for large-scale image processing
- Supervised and mentored Master's students
- Transfer learning for lymphocyte infiltration detection in colorectal cancer using a pre-trained breast cancer model
- Employed UNET model with some contributions to image segmentation and cell detection
- Applying multiple instance learning and Bayesian neural network for tissue classification (multi-class: 9 classes)
- Interobserver agreement assessment for quantifying tumor-infiltrating lymphocytes in colorectal cancer tissue and systematically review machine learning approaches for quantifying tumor-infiltrating lymphocytes
- Conducted research and applied statistical analysis methods such as Cohen's Kappa and intraclass correlation to measure agreement between pathologists and utilized data visualization techniques (heatmap, Bland-Altman plots) by using R
- Systematically reviewed papers that have quantified lymphocytes in CRC histopathological images
Research Assistant
Mashhad University of Medical Sciences
09.2018 - 02.2022
- Machine learning based survival prediction in self-Immolation patients
- Implemented and fine tuned machine learning-based models (Gradient Boosting, SVM, Random Forest, MLP, and KNN) to predict survival outcomes in self-immolation patients, aiming to improve treatment management and decision-making in burn centers
- Retrospective analysis and feature selection of self-immolated patients using demographic, clinical, and treatment data
- Machine learning-based prediction of COVID-19 patients requiring intensive care
- Applying machine learning models (decision tree & regression models) to predict COVID-19 ICU admission risk based on early clinical and laboratory data, aiding clinicians in prioritizing care
- Applied feature selection and data preprocessing to improve model performance (biostatistical feature selection)
- Analyzed COVID-19 mortality risk factors using hematological and biochemical marker: a data-driven approach
- Conducted statistical analysis on hospitalized COVID-19 patients, identifying key mortality predictors
- Applied logistic regression and hypothesis testing to evaluate the impact of smoking on disease prognosis
- Clinical data engineering and data preprocessing, handling electronic medical records
University Lecturer
Shiraz University of Medical Sciences
01.2014 - 12.2018
- Lectured data structures and C programming and IT in the medical domain to at least 100 undergraduate students in four semesters
- Developed course materials, lectures, and exams, fostering understanding of key concepts in computer and medical informatics
Research Assistant
Shiraz Transplant Research Center
01.2017 - 09.2018
- Designed and implemented a SQL database for pathology reporting, ensuring compliance with pathology protocols
- Developed ER diagrams, normalized schemas, and SQL querying and reporting
- Predicted Patient Survival after Liver Transplant Using Machine Learning
- Applied a predictive model to identify key factors affecting patient survival after liver transplantation using machine-learning techniques (SVM, Bayesian network, decision tree, artificial neural network, KNN)
- Analyzed patient survival by using Cox regression
Education
PhD candidate - Medical Informatics
Mashhad University of Medical Sciences
01-2025
Master's degree - Medical Informatics
Shiraz University of Medical Sciences
01-2016
Bachelor - Software Engineering
Eshragh Higher Education Institute
01-2010
Associate Degree - Software Engineering
Larestan Azad University
01-2006
Skills
- Python
- Pandas
- NumPy
- Matplotlib
- Scikit-learn
- PyTorch
- TensorFlow
- Keras
- R programming
- SQL
- Supervised learning
- Weakly Supervised learning
- Applied Deep Learning
- Machine Learning
- Computer Vision
- OpenCV
- OpenSlide
- Torchvision
- Segmentation_models_pytorch
- Docker
- Git for Version Control
- HPC
- Containerization
- Command Line Interface
- Survival Analysis
- Cox Regression
- Logistic Regression
- Predictive Modeling
- Kaplan-Meier Survival Curves
- SPSS
- Data Management
- Data Engineering
- Data Modeling
- Feature Selection
- Feature Engineering
- Visualization
- Data Optimization
- Digital Pathology
- Whole Slide Imaging
- Pathology Concepts
- Handling Clinical Records
- Medical Terminologies
- Coding Systems
- ICD
- SNOMED CT
- Healthcare Standard
- Regulations
- HL7 FHIR
- DICOM
- GCP
- Therapeutic Fields
- Designing Digital Health Solutions
- Evaluation of Digital Health Solutions
- Medical Software
- Team Spirit
- Collaboration in Cross-Functional Teams
- Problem-Solving
- Detail-Oriented
- Creativity
- Lecturing
- Machine learning
- Data preprocessing
- Statistical analysis
- Image segmentation
- Clinical data management
- Hyper-parameter tuning
- Research methodology
- Data visualization
- Technical writing
- Detail orientation
- Critical thinking
- Deep learning
- Hyperparameter tuning
- Feature extraction
Certification
- Generative AI with Diffusion Models
- Fundamentals of Accelerated Data Science
- R programming
- SQL
Accomplishments
- Awarded a six-month research scholarship for work in Germany., 2021
- Top-ranked PhD student with a GPA of 19.21/20.
Languages
Persian
English
German
Personal Information
Title: Medical Data Scientist
Timeline
Researcher
Technical University of Munich
03.2022 - Current
Research Assistant
Mashhad University of Medical Sciences
09.2018 - 02.2022
Research Assistant
Shiraz Transplant Research Center
01.2017 - 09.2018
University Lecturer
Shiraz University of Medical Sciences
01.2014 - 12.2018
PhD candidate - Medical Informatics
Mashhad University of Medical Sciences
Master's degree - Medical Informatics
Shiraz University of Medical Sciences
Bachelor - Software Engineering
Eshragh Higher Education Institute
Associate Degree - Software Engineering
Larestan Azad University
Similar Profiles
Siddharth JadejaSiddharth Jadeja
Research Trainee at Technical University Of MunichResearch Trainee at Technical University Of Munich
Amen ShahpalAmen Shahpal
Researcher in the field of photocatalysis at Technical University of MunichResearcher in the field of photocatalysis at Technical University of Munich
GRACE KINNEGRACE KINNE
Postdoctoral Researcher at Technical University of MunichPostdoctoral Researcher at Technical University of Munich