04.03.2026 Salary: 14 000 – 16 000 PLN gross
Position Objective
Conduct systematic research and development work on new methods of ophthalmic image analysis and integration of imaging data with descriptive and clinical data using deep learning models, including Large Multimodal Models (LMM), in order to develop innovative diagnostic support systems.
Responsibilities
- Design, implementation, and experimental validation of new neural network architectures for the analysis of OCT, fundus, and retinal images, including segmentation, classification, and pathology localization models.
- Investigation of the impact of network topology (CNN, U-Net, Vision Transformers, multiscale models, attention-based architectures) as well as device and acquisition protocol variability on model stability and generalization.
- Development of data processing algorithms.
- Advancement of dynamic spatial coherence theory.
- Design, implementation, and testing of multimodal architectures combining imaging data with clinical data (descriptions, test results, patient history, demographic data) using Large Multimodal Models (LMM).
- Design of experimental solutions integrating visual encoders, Large Language Models (LLMs), and multimodal fusion layers, including research on fine-tuning strategies and adaptation of foundation models to medical data.
- Analysis of the capability of visual and multimodal models to generate pathology descriptions (explainability), support clinical decision-making, and identify inconsistencies between imaging data and medical documentation.
- Optimization of image processing and recognition algorithms.
- Addressing technological uncertainty challenges related to language model hallucinations, medical data bias, and limited interpretability of multimodal predictions.
- Design of validation procedures and comparative experiments (classical models vs. LMM), analysis of misclassifications, and development of quality assessment metrics for predictions and generated descriptions.
- Development of new methods for retinal structure segmentation and detection of subtle biomarkers in OCT/STOC-T images, and advancement of multimodal clinical decision support systems.
- Design of deployment architecture (PACS/EHR), optimization of model inference performance, and investigation of on-premise deployment possibilities in compliance with GDPR requirements.
- Development of software for inference model deployment.
- Documentation of research hypotheses and experimental results, preparation of materials for IP protection (patents, IP Box), presentation of results at seminars, conferences, and in scientific publications, and collaboration with ophthalmologists in defining and interpreting research problems.
Requirements
- Master’s or PhD degree in computer science, artificial intelligence, biomedical engineering, mathematics, physics, bioinformatics, or related fields.
- Experience in computer science, photonic engineering, or biomedical engineering.
- Experience in designing and training deep neural networks (CNN, U-Net, Vision Transformers, attention-based models, multiscale architectures).
- Experience in medical image analysis (preferred: OCT, fundus, retinal imaging).
- Knowledge of segmentation, classification, and object detection methods in images.
- Ability to work in interdisciplinary teams (collaboration with physicians and clinical teams).
- Ability to independently design and implement processes.
- Strong interpersonal and organizational skills.
- Very good command of English (minimum C1).
- Openness to working in an international and multicultural environment.
- Knowledge of libraries: PyTorch, TensorFlow, OpenCV, pandas, NumPy, MONAI.
- Programming skills in: C++/C#, Python, MATLAB, Julia.
- Experience in GPU programming.
- Ability to create numerical simulations related to the OCT eye measurement process.
- Knowledge of dynamic spatial coherence theory and optical coherence tomography.
- Knowledge of optical eye measurement systems.
- Experience in deploying and integrating inference models.
What We Offer
- A collaborative and research-oriented environment.
- Flexible working arrangements (task-based work / occasional remote work).
- Friendly working atmosphere.
- Training and professional development opportunities (courses, job shadowing).
- Private medical care.
How to Apply?
Please send your CV and cover letter to: icter_jobs@ichf.edu.pl
Please include the recruitment reference number in the email subject line: R_6/2026_ICTER
Application deadline: March 11, 2026
