Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Development of Deep Learning Methods for routine Dermatology consultation
Department of Dermatology at the University Hospital Zurich is currently offering various Master and Bachelor thesis projects in clinical applications of deep learning algorithms.
Keywords: Deep Learning, Machine Learning, Artificial Intelligence, Dermatology, Real World Data, Computer Vision, Cancer
Deep learning algorithms have already been proven to outperform board-certified dermatologists in skin cancer identification tasks in certain situations. Such studies, however, did not consider real world data representative of the reality encountered by the dermatologists in routine consultations. In order to design effective diagnosis-aid tools the Dermatology clinic is acquiring a dedicated dataset used to training state-of-the-art neural networks based classifiers. The final goal of such tools is to be deployed during routine consultation, so the design requirements are guided by the requirements stated by experienced dermatologists.
Deep learning algorithms have already been proven to outperform board-certified dermatologists in skin cancer identification tasks in certain situations. Such studies, however, did not consider real world data representative of the reality encountered by the dermatologists in routine consultations. In order to design effective diagnosis-aid tools the Dermatology clinic is acquiring a dedicated dataset used to training state-of-the-art neural networks based classifiers. The final goal of such tools is to be deployed during routine consultation, so the design requirements are guided by the requirements stated by experienced dermatologists.
Your main tasks will include:
- Contribute to the curation of the real world dataset continuously acquired at the USZ Dermatology clinic.
- Develop and optimize deep learning based models for clinical diagnosis according to dermatology considerations, in particular self-supervised methods for skin lesion screening.
- Depending on the progress, help to deploy the tools during real consultation, evaluate the impact as diagnosis support tools and implement feedback from dermatologists
Your main tasks will include:
- Contribute to the curation of the real world dataset continuously acquired at the USZ Dermatology clinic. - Develop and optimize deep learning based models for clinical diagnosis according to dermatology considerations, in particular self-supervised methods for skin lesion screening. - Depending on the progress, help to deploy the tools during real consultation, evaluate the impact as diagnosis support tools and implement feedback from dermatologists
_Your profile_ - You have a technical background (computing science, physics or engineering) and ideally previous experience with machine learning algorithms. Interest and motivation to work with clinical images in a multi-disciplinary team is a must.
_What we offer_ - Interesting topics in the machine learning domain with real world applications, in a young group with many master and doctoral students. As well, collaborations with others Swiss hospitals for the development of the algorithms is foreseen.
_Contact_ - Dr. sc. nat. Javier Barranco Garcia javier(dot)barrancogarcia(at)usz(dot)ch
_Your profile_ - You have a technical background (computing science, physics or engineering) and ideally previous experience with machine learning algorithms. Interest and motivation to work with clinical images in a multi-disciplinary team is a must.
_What we offer_ - Interesting topics in the machine learning domain with real world applications, in a young group with many master and doctoral students. As well, collaborations with others Swiss hospitals for the development of the algorithms is foreseen.
_Contact_ - Dr. sc. nat. Javier Barranco Garcia javier(dot)barrancogarcia(at)usz(dot)ch