Scientific Assistant in Machine Learning for Healthcare

The Biomedical Data Science Lab (BMDS), headed by Prof. Catherine Jutzeler, is looking for a highly motivated and skilled Scientific Assistant to join our interdisciplinary team. In this role, you will work closely with the BMDS team of doctoral and postdoctoral researchers to contribute to a new research initiatives in using deep learning models for time series data captured in intensive care unit environments. In this position, you will be advised by Dr. Lakmal Meegahapola. We particularly value, if you are comfortable addressing cutting-edge research questions in machine learning and deep learning, where you must be creative and motivated in searching and reading relevant literature, and developing novel models to tackle complex real-world problems in healthcare.

Modern intensive care unit (ICU) environments produce real-time data from a plethora of modalities. From the time a patient is admitted to an ICU, real-time physiological time series data about the patient’s vitals (heart rate, blood pressure, temperature, etc.) are captured, supplemented by various lab tests and demographic information. These multimodal data can potentially be used for risk profiling and for detecting and predicting various health outcomes, such as sepsis, mortality, and kidney failure, using machine learning and deep learning models. These models have the potential to help clinicians in decision-making, thereby saving patients' lives.

You will: 

• Review the latest machine learning literature on time series data processing and prediction tasks in intensive care units. 
• Use data from the national project on sepsis and other publicly available ICU datasets such as MIMIC IV, HiRID, and eICU, to develop deep learning models for predicting health outcomes (e.g., sepsis onset, mortality, kidney failure) using multimodal time series data.
• Improve the models with personalization techniques.
• Report the findings in the form of research publications. 

Qualifications:

• Recently completed Master's degree in Computer Science, Data Science, Machine Learning, Electrical and Computer Engineering, Computational Biology and Bioinformatics, Health Sciences and Technology, or related fields.
• Strong programming skills in Python. 
• Experience working with large datasets, high-performance computing environments, and linux operating systems.
• Experience developing machine learning and deep learning models (e.g., Tensorflow, PyTorch).
• Excellent written and oral communication skills in English. 
• Willingness and passion to learn about biomedical applications of machine learning and deep learning models. 
• Ability to work both independently and collaboratively in a team environment.

Preferred Qualifications:

• Project experiences in developing deep learning models.
• Knowledge of model fairness, robustness, and domain adaptation. 
• Experience working with multimodal time series data.
• Publications (papers, posters, etc.) in machine learning or sensor data processing-oriented conferences, workshops, or symposiums (e.g., NeurIPS, ICML, ICLR, AAAI, CHIL, ML4H, IMWUT, IPSN, etc.).
• Swiss/EU citizens and swiss work permit holders preferred due to the tight timeline. 

We offer a 1-year contract (80% rate) starting 1st August 2024, at the BMDS lab that includes: 

• Salary in accordance with ETH Zurich's Scientific Assistant I salary level, pro-rated to 80%.
• Opportunities to engage with different communities bridging machine learning and health research, leading to high-impact publications in machine learning, computer science, and/or healthcare oriented venues. 
• You will be part of a highly motivated, friendly and collaborative team.
• We will support your scientific career.
• You will be able to attend relevant (inter-)national conferences to increase your visibility and present the project outcomes.
• You will be involved in the supervision of undergraduate students and are also offered the possibility to be involved in teaching activities in the lab.
• We offer a flexible hybrid working environment.