AI Models for Medical Images, Region Skåne Call-off 2865
Liam Tran
HR Administrator
Efterfrågade kompetenser
Om tjänsten
Role Summary
An experienced AI specialist is required for a consulting assignment supporting research at Clinical Physiology and Nuclear Medicine at Skåne University Hospital in Malmö and Lund. The assignment focuses on developing AI-based image biomarkers and deep learning models for medical image segmentation and image enhancement for research purposes.
The primary focus will be the development of robust models for PET/CT image segmentation, including whole-organ segmentation, identification of individual skeletal bones, and tumour delineation in CT and PET images. The models will support research involving fluorid-PET/CT for prostate cancer, FDG-PET/CT for lymphoma and lung cancer, Dotatate-PET/CT for neuroendocrine tumours, and PSMA-PET/CT for prostate cancer.
Key Responsibilities
- Design, develop, train and evaluate deep learning models for segmentation and enhancement of medical images.
- Develop PET/CT-focused segmentation models covering all organs, individual bones in the skeleton, and tumours within relevant organs.
- Work with imaging data from computed tomography (CT), positron emission tomography (PET), and magnetic resonance imaging (MRI).
- Handle medical images and segmentation files in accordance with the DICOM standard.
- Develop models suitable for research use and ensure that the research group receives unrestricted access to the delivered models for research purposes.
- Contribute scientific and technical text for research publications.
- Collaborate with researchers in clinical physiology, nuclear medicine, and medical imaging research.
Technical Stack & Requirements
- Strong practical experience developing AI and deep learning models for medical imaging applications.
- Extensive experience with deep learning-based segmentation of PET and CT images.
- Experience with PET/CT image analysis, including skeletal segmentation and tumour delineation.
- Strong knowledge of DICOM-formatted CT and PET images.
- Experience managing segmentation files compliant with the DICOM standard.
- Knowledge of image segmentation and image enhancement techniques for medical images.
- All servers and computers used for the assignment must be located in Sweden.
Qualifications & Experience
The consultant should have substantial prior experience in AI tool development for medical imaging, with particular expertise in deep learning for PET and CT image segmentation. A completed PhD in a relevant field is highly advantageous. Fluency in both Swedish and English is required, including the ability to contribute to scientific publications.
Work Logistics & Location
This is a consulting assignment for an initial term of 18 months, with an option to extend for an additional 18 months. The work supports research activities at Skåne University Hospital in Malmö and Lund. Technical infrastructure used within the assignment must be hosted in Sweden.