Postdoc on Predictive Monitoring for Alarm-Limiting AlgoRithm-based Monitoring
Postdoc on Predictive Monitoring for Alarm-Limiting AlgoRithm-based Monitoring
You cannot apply for this job anymore (deadline was 31 Mar 2018).
Browse the current job offers or choose an item in the top navigation above.
Job description
Background
Medical care for very premature infants in developed countries is provided in a Neonatal Intensive Care Unit (NICU), where infants spend the first weeks or even months of their lives in a vulnerable state in which life-threatening complications can occur. Patient monitoring equipment is used to continuously monitor vital signs, such as heart rate, respiration and oxygen saturation. Alarms are used to alert the caretaker in case of a critical situation. Unfortunately, current systems too often raise false alarms while at the same time sometimes life-threatening situations are easily missed.
The ALARM project aims to develop advanced methods for improved patient monitoring by fusing the measured vital signs and signals acquired with additional video monitoring to reduce false alarms and improve detection of relevant events. The project will employ data analytics and machine learning methods to detect deterioration earlier, using the vital signals and physiology-based algorithms. In addition, it shall use video derived signals for robust patient-motion detection and unobtrusive vital sign monitoring as a further resource to raise the appropriate alarms, and lower the false alarms.
Project description
This postdoc project is part of the larger ALARM project for which a grant was obtained from NWO-TTW will focus on predictive monitoring. The postdoc is expected to develop machine learning methods for predictive monitoring in an intensive care setting and to design the setup needed for adequate machine learning. The PostDoc will help to guide two PhD students. For one of these PhD students and the postdoc, the following phases in the project need to be addressed:
In the first phase, an integrated algorithm will be developed to predict adverse outcome using a multimodal sensing approach. Acquired data will be used to search for patterns relating to deterioration, using the collected 'anchor points' that indicate clinical diagnosis of the disease. Starting from trend analysis and simple correlation analysis, methods will proceed towards physiology-based advanced data mining
In a second phase, the algorithm will be used to develop parameters that can function as 'early warning signs'. Data mining techniques are expected to result in indicators that are not directly insightful to clinicians. To bring this back into the clinical domain we intend to combine data mining with physiological models in order to help the physician to understand which factors are the main contributors to the indicators that the patient is deteriorating.
Departments and collaborators
Eindhoven University of Technology (TU/e) is a world-leading research university specializing in engineering science & technology, with a strategic collaboration with Philips and several regional hospitals, like Maxima Medical Center, who are both involved in the current project. The TU/e is the world's best-performing research university in terms of research cooperation with industry (#1 since 2009).
The School of Medical Physics and Engineering Eindhoven (SMPE/e) specializes in engineering and implementation of technology in healthcare and has a long track record of educating healthcare engineers and supporting healthcare projects in hospitals and other healthcare institutions, in clinical physics, medical engineering and clinical informatics.
The ALARM project team is designed to combine extensive knowledge in the key fields (neonatal care, physiological monitoring, system and signal analysis, data analytics and implementation), with an elaborate clinical infrastructure. There is long-standing research collaboration between Eindhoven University of Technology (TU/e), Máxima Medical Center (MMC) and Philips on perinatal and neonatal projects, formalized in the so-called IMPULS Perinatology flagship. This project will be part of that collaboration. As part of this collaboration, the PhD candidate will work part of the time at Philips and carry out experiments at Maxima Medical Center.
Medical care for very premature infants in developed countries is provided in a Neonatal Intensive Care Unit (NICU), where infants spend the first weeks or even months of their lives in a vulnerable state in which life-threatening complications can occur. Patient monitoring equipment is used to continuously monitor vital signs, such as heart rate, respiration and oxygen saturation. Alarms are used to alert the caretaker in case of a critical situation. Unfortunately, current systems too often raise false alarms while at the same time sometimes life-threatening situations are easily missed.
The ALARM project aims to develop advanced methods for improved patient monitoring by fusing the measured vital signs and signals acquired with additional video monitoring to reduce false alarms and improve detection of relevant events. The project will employ data analytics and machine learning methods to detect deterioration earlier, using the vital signals and physiology-based algorithms. In addition, it shall use video derived signals for robust patient-motion detection and unobtrusive vital sign monitoring as a further resource to raise the appropriate alarms, and lower the false alarms.
Project description
This postdoc project is part of the larger ALARM project for which a grant was obtained from NWO-TTW will focus on predictive monitoring. The postdoc is expected to develop machine learning methods for predictive monitoring in an intensive care setting and to design the setup needed for adequate machine learning. The PostDoc will help to guide two PhD students. For one of these PhD students and the postdoc, the following phases in the project need to be addressed:
In the first phase, an integrated algorithm will be developed to predict adverse outcome using a multimodal sensing approach. Acquired data will be used to search for patterns relating to deterioration, using the collected 'anchor points' that indicate clinical diagnosis of the disease. Starting from trend analysis and simple correlation analysis, methods will proceed towards physiology-based advanced data mining
In a second phase, the algorithm will be used to develop parameters that can function as 'early warning signs'. Data mining techniques are expected to result in indicators that are not directly insightful to clinicians. To bring this back into the clinical domain we intend to combine data mining with physiological models in order to help the physician to understand which factors are the main contributors to the indicators that the patient is deteriorating.
Departments and collaborators
Eindhoven University of Technology (TU/e) is a world-leading research university specializing in engineering science & technology, with a strategic collaboration with Philips and several regional hospitals, like Maxima Medical Center, who are both involved in the current project. The TU/e is the world's best-performing research university in terms of research cooperation with industry (#1 since 2009).
The School of Medical Physics and Engineering Eindhoven (SMPE/e) specializes in engineering and implementation of technology in healthcare and has a long track record of educating healthcare engineers and supporting healthcare projects in hospitals and other healthcare institutions, in clinical physics, medical engineering and clinical informatics.
The ALARM project team is designed to combine extensive knowledge in the key fields (neonatal care, physiological monitoring, system and signal analysis, data analytics and implementation), with an elaborate clinical infrastructure. There is long-standing research collaboration between Eindhoven University of Technology (TU/e), Máxima Medical Center (MMC) and Philips on perinatal and neonatal projects, formalized in the so-called IMPULS Perinatology flagship. This project will be part of that collaboration. As part of this collaboration, the PhD candidate will work part of the time at Philips and carry out experiments at Maxima Medical Center.
Specifications
- max. 38 hours per week
- Eindhoven View on Google Maps
:fill(white)/logos/tue-en-wide_acOLN2b.png)
Eindhoven University of Technology (TU/e)
Requirements
We are looking for candidates that match the following profile: A PhD degree in a subjected related to data science / machine learning / deep learning. With a background in Applied Physics, Computer Science, Electrical Engineering, Biomedical Engineering or related disciplines with excellent grades. A good mix between an interest in basic science and applied engineering. Solid programming skills (e.g., in Matlab or Python and C or C++).A team player that enjoys to work in multidisciplinary teams (including scientists, engineers, physicians and nurses). Good communication and organization skills. Excellent English language skills (writing and presenting). Proficiency in Dutch or the willingness to learn Dutch.
Conditions of employment
We offer a fixed-term, 2 year postdoc position in a research consortium with an excellent reputation. Salary and benefits are in accordance with the Collective Labour Agreement for Dutch Universities, including: A gross monthly salary between € 3111 and € 3358 (depending on experience). Additionally, 8% holiday and 8.3% end-of-year annual supplements. Additional benefits, including excellent technical infrastructure, child care, holiday savings schemes, and sports facilities.
Assistance for finding accommodation is offered for foreign postdocs.
Assistance for finding accommodation is offered for foreign postdocs.
