Optimization for Spiking Neural Networks
An exploration of optimization methods for spiking neural networks
High-Performance Simulation of Spiking Neural Network on GPUs
Exploit sparsity of data flow in spiking neural network simulations
Tracing the role of clock neurons in brain function
The mammalian "master clock" consists of 20,000 neurons in the hypothalamus controlling all aspects of diurnal behaviour and physiology. Using techniques ranging from optogenetics to single-cell -omics technologies, our laboratory tries to understand neural circuits therein and their function.
Engineering the Diabetic Tendon Microenvironment
This project aims to engineer a synthetic hydrogel-based cell niche that mimics key aspects of the diabetic tendon extracellular matrix. Patient-derived tendon fibroblasts will be used to construct a more physiologically relevant three-dimensional (3D) model.
Manipulation of nanowires for engineered strain sensors used in biomedical implants.
Fabrication of biocompatible strain sensors, resistive or capacitive to be used in-vivo to measure deformation of different tissues. The project is quite broad so many backgrounds will suit. Electrical Engineers, Material Scientist, Biomedical Engineers, among others.
PostDOC position in child neurodevelopment.
We look for a PhD neuro-/psychologist to join our team as a senior researcher for studies focused on the neurodevelopment of human milk fed infants with a special interest on family/environmental dynamics.
The role of tendon matrix composition in regulating tenocytes fate: A tissue engineering approach.
The goal of this project is to create a 3D in vitro model to study tendon cell-ECM interaction using collagen gels and primary rat tenocytes.
Tendon-on-a-Chip: Developing and characterizing a multi-channels microfluidic device for studying tendon biology
This goal of this project is to refine and characterize a novel approach in designing microfluidic networks that allow to host tendon-like micro-tissue structures for long-term culturing experiments under mechanical stimulation.
Modelling cardiac mechanics and function
The aim of the project is an in-silico replica of a patient’s heart starting from Magnetic Resonance (MR) imaging which will guide clinical evaluation of cardiac function, predict disease progression and augment MR data with low resolution and/or missing details.
PhD Position in Computational Biomechanics
We are looking for a qualified PhD candidate for computational musculoskeletal modeling and simulation. Deadline for applications: October, 31
Home-based wound monitoring with mobile app
Periodic wound measurement and documentation are important to monitor wound progression and healing. we run a study with patients in Systematic Sclerosis with digital ulcers. Wound images were collected by patients using our developed mHealth tool at home.
Quantification of Cardiac Metabolism Using Machine Learning
Magnetic resonance imaging (MRI) of hyperpolarized 13C-pyruvate allows studying cardiac metabolism. Quantification of cardiac metabolism based on advanced kinetic models is desired. As an alternative to conventional parameter fitting, machine learning can be used to predict reliable parameters.
Development of novel biosensor
I have a number of Master Thesis projects on Focal Molography. Molography is a novel method for the analysis of biomolecular interactions with unprecedented robustness and sensitivity.
Reinforcement learning for car racing in GTS
Reinforcement learning for car racing in Gran Turismo SPORT (GTS) for the Sony Playstation 4
Learning-Guided MPC Flight
Train a neural network to predict and intermediate representation that can be used by an MPC.
Optoacoustic (photoacoustic) imaging has been the fastest growing imaging modality in the last decade. Our group has pioneered 5D (3D + real-time + multi-spectral) optoacoustic tomography and shown unique capabilities in biomedical research. We offer projects on this topic.
A Practical Application of Neuromorphic Computing: Detecting Human Epileptic Seizures From Short-Term Intracranial Electroencephalograms (iEEG)
This project aims to build brain-inspired computing hardware to implement always-on analysis of iEEG recordings to optimize diagnostics and therapies for epilepsy patients, by developing Spiking Neural Network models and validating them using neuromorphic prototype chips
Ultrasound-coupling holder design for optoacoustic tomography applications
Optoacoustic tomography is a novel technique that can be used to image small animals for preclinical research and patients for clinical research. Water coupling between ultrasound detectors and inspected tissue is required for propagation of ultrasonic waves. An easy-to-use design is required.
DNA Aptamer-Based Electronic, Plasmonic, and Optical Biosensors
Integrating DNA-based biorecognition elements termed aptamers that recognize chemical targets with high specificity and selectivity into next-generation electronic/plasmonic biosensors.
Neuromorphic chip interface with a microcontroller
This project will allow to explore input/output solutions of spikes generated by (or sent to) neuromorphic chips using the Teensy 4.0 platform (ARM Cortex M7), and to characterize temporal (1us resolution) and bandwidth requirements.
Stiffness Perception with Different Grasp Type Tools
The goal of this work is to perform a pilot study that assesses human stiffness perception when a material is explored with a tool with different grasp type handles. You will be involved in all steps of the pilot study: study design, study conduction, data analysis, and evaluation.
Development und design of a multifunctional Laparoscope
Most of the currently existing tools have exactly one function and are single use devices. Therefore, we conceptualized a multifunctional, multiuse laparoscopic device. Are you excited to build things that matter?
Machine Learning for Advanced Magnetic Resonance Perfusion Quantification
Magnetic Resonance (MR) perfusion imaging is a clinically established technique to detect coronary artery disease by visual assessment. However, quantitative evaluation of myocardial blood flow is desirable as it promotes objective diagnosis. Here, ML may help to avoid expensive fitting procedures.
Parameter and Uncertainty Estimation for MR Relaxometry
The primary aim of the project is to adapt and characterize an existing framework for Bayesian parameter and uncertainty estimation that utilizes machine-learning for posterior estimation.
Can Smart Textile Sleepwear using far infrared technology improve the Sleep Quality and hence Recovery of Athletes?
All living organisms permanently emit energy into their surroundings. Carefully chosen minerals, embedded in or printed on textiles reflect this energy in the form of far infrared rays. Far infrared rays can improve the oxygenation of cells, which enables the body to fall asleep quicker, stay in deep sleep mode longer, lower the muscle tonus and helps the body to perform maintenance and preservation tasks while you sleep. Dagsmejan, a smart textile start-up based in St. Gallen, has developed sleepwear that uses this far-infrared technology to help, in particular athletes, experience a more restful sleep, to enhance recovery and to improve athletic performance.
Design of a tendon-based robotic device for locomotion training
The proposed project aims at designing a novel tendon-based parallel robotic device for locomotion training that can guide or resist the patients’ movements during training.
Deep Learning Enabled Segmentation of Brain Tissue in MRI
This project aims to create deep learning based image segmentation algorithm for high-field mouse MRI data for advanced neuroimaging applications.
Smart Teaching - Master Thesis in collaboration with CSEM
Industrial grasping and pick-and-place highly benefit from robotic systems, allowing fast and precise execution of highly repetitive tasks. The common and widely established approach implies each single action of each single robot to be manually preprogrammed within a teaching phase, and simply repe
Boosting Bayesian models of human brain connectivity
The aim of this project is to implement recent advances in boosting black box variational inference in Pyro and apply the resulting method to estimate models of effective brain connectivity in human functional MRI data.
Development of a Fiber-Based Optoacoustic Microscope
Optoacoustic microscopy can image vasculature with single capillary resolution in 3D and at depth. This new imaging approach holds great potential to provide better insights into cerebrovascular function and facilitate efficient studies into neurological and vascular abnormalities of the brain.
Probabilistic System Identification of a Quadrotor Platform
This work investigates the usage of Gaussian Processes for uncertainty-aware system identification of a quadrotor platform.
Why so complicated? Global localization without fuss
Achieve comparable or better place recognition performance than NetVLAD, with a simpler network architecture.
Data-Driven Visual Inertial Odometry for Quadrotor Flight
Investigate the usability of data-driven methods to improve the performance of a VIO pipeline on a resource-constrained platform.
Sleep Traits and Age: does our ‘Brain Fingerprint’ Remain?
From studies conducted with young adults and teenagers, we know that brain activity during sleep varies from person to person, but is very repeatable from night to night within one person. We want to know whether this “brain fingerprint” is also visible later in life.
Learning to Deblur Images with Events
Explore machine learning based approaches for deblurring of images with events
Simulation to Real World Transfer
The project aims to develop techniques based on machine learning to have maximal knowledge transfer between simulated and real world on a navigation task.
Unsupervised Obstacle Detection Learning
In this project, we aim to build a self-supervised depth estimation and segmentation algorithm by embedding classic computer vision principles (e.g. brightness constancy) into a neural network.
Target following on nano-scale UAV
This project focuses on the object tracking, (i.e., target following) on nano-UAVs (few centimeters in size).
Ground Segmentation for Landing
In order for UAVs to fly and especially land autonomously we need evaluate the safety of landing spots. Having semantic information about the scenery, eg. road, forest, roof, etc, enables the UAV to pick a safe landing spot.
Stereo matching using CUDA
Dense stereo matching is crucial for creating depth maps. However, it is very computationally expensive on CPUs. As a result the update rate is low, which makes it unusable for many tasks, eg. avoidance in dynamic environments. Especially on drones the computation power is limited. Through the introduction of the Nvidia Jetson series computers we got access to lightweight embedded GPUs. Running the stereo matching on a GPU can potentially make it significantly faster.
Biomedical Software Engineering - focusing on Continuous Integration, Distribution, and Linux Dependency Management
We develop Python-based Open Source pipelines which enable cutting-edge modelling and visualization of functional brain imaging data. As our pipelines grow in scope, we are looking to meet more stringent software engineering standards, and improve our distribution model with regard to accessibility.
VIO in dynamic environments
Visual-Inertial Odometry is a great solution for drone-navigation in GPS-denied environments. Its ability to provide centimeter-level precision in local navigation makes it a suitable choice in many commercial applications like last centimeter drone delivery. Conventional VIO algorithms work well in static environments. However, when the environment is dynamic, i.e most of the visual features come from a moving environment, for instance a moving platform, the VIO does not perform reliably. This problem can be attributed to the unreliable initialisation phase of the VIO pipeline, which is the most critical phase. Most initialisation algorithms are based on structure-from-motion, which assumes that the environment is static. In such a scenario the initialisation algorithm needs to be adapted to take into account the motion of the features.
Optical ‘PET’ - development of a portable fluorescence tomography system
Fluorescence molecular tomography (FMT) provides molecular information on tracer bio-distribution in the organism similar to positron emission tomography (PET). Instead of using radioactive tracers in PET, FMT non-invasively resolves the three-dimensional distribution of fluorescent probes in vivo. Thus, it can be considered an optical version of PET (but only 1% cost of PET). Although both image reconstruction algorithms and instrumentation for FMT have evolved over the past decade, none of the setups has been widely accepted as a standard molecular imaging tool for routine biomedical research. A miniaturized FMT with compact design and easy handling is certainly attractive. This calls for the integration of MEMS and customized components facilitated with 3D printer technology. Previously we have developed a compact FMT standalone system and a FMT/MRI hybrid system, which serves as a starting point for the project.
Elucidating the etiopathology of neuronal development and maturation in attention-deficit hyperactivity disorder (ADHD) using iPSC-induced neuronal modelling
The Master candidate will learn the method generating iPSC following by neuronal generation to study maturation and differentiation alterations in attention-deficit hyperactivity disorder (ADHD) patients compared to controls. Morphological, molecular and functional techniques will be applied.
Dopamine transporter trafficking following methylphenidate (Ritalin) treatment in human dopaminergic neurons
The student will investigate the trafficking of the dopamine transporter between intracellular and synaptic localization in a human dopaminergic cell line via live cell imaging following pharmacological treatment.
Building a photorealistic rendering engine
Using photorealistic game engine, such as Unity or Unreal Engine, for vision-based AI research has become increasingly popular in the robotics community since it is faster to generate and automatically annotate high-quality synthetic image data. Existing drone simulators, such as RotorS or Flightgoggles, are either not providing high-quality synthetic images or not supporting accurate quadrotor dynamics.
Building a simulator for quadrotors
Building a simulator that combines the photorealistic image rendering engine with the ROS framework could greatly help the robotics research community to develop algorithms with the simulator. For example, both two popular open-source simulators, CARLA and AirSim, are supporting ROS.
PhD on host-microbe interactions
PhD position in the LeibundGut-lab on "Skin commensal fungi at the interface between health and disease"
MSc thesis on host-microbe interactions
Skin commensal fungi at the interface between health and disease. This MSc project in immunology is hosted by the Section of Immunology (located at the Vetsuisse Faculty of UZH)
Establishment of a Novel Experimental System for AFM-based Biophysical Membrane Fusion and Curvature Measurements
• Discover the basic working principles of cells and synapses. • Develop new tools to study them.
Multimaterial 3D Printing at the Microscale
3D printing is a family of emerging techniques for fabrication of functional devices. We are currently exploring methods to expand the range of printable materials, such as metals and polymers. Extending the range of materials will automatically lead to new opportunities for fabrication of functiona
3D Printing of a Microelectrode Array for Neural Applications
3D printing is a family of emerging techniques for fabrication of functional devices. Here, we make use of basics of electrochemistry and scanning probe methods to deliver metal ions locally and transform them into solid metal features. This is achieved by using glass capillaries with ultrasmall opening diameters with dimensions down to a few nanometers.
Design of a Multichannel Microfluidic AFM System
Design, Fabrication and Analysis of a Multichannel Hollow Cantilever Probe for the Fluid Force Microscope (FluidFM).
Machine Learning in Stroke Imaging
The objective of this project is to develop and implement various machine learning methods to the recognition, segmentation, and diagnosis of brain strokes, using a large database of computer tomographic (CT) and magnetic resonance (MR) images.
Development of a pulsatile flow test bench for in-vitro studies on endothelial cells
In this project, a test bench is to be developed which allows the generation of accurately controllable pulsatile flows for in-vitro studies on endothelial cells.
Pushing hard cases in tag detection with a CNN
Try to handle failures in april/aruco tag detection using deep learning.
Teach and Aggressive Repeat
Teach and repeat, but try to be as fast as possible on repeat. Goal is to deploy this on a quad.
Doktorandenstelle Digitale Beratung (Bankberater-Kunde, Arzt-Patient) Universität Zürich
Wir transformieren Beratung mit IT in Banken, öffentlichem Sektor und von Ärzten. Wollen Sie dazu promovieren und sprechen gut Deutsch? Dann melden Sie sich!
Masterarbeit an der KJPP/PUK (Developmental Neuroimaging Group): Psychometrische Testung der Leseleistung
Zu vergeben sind zwei Masterarbeiten in der Developmental Neuroimaging Group am Zentrum für kinder- und jugendpsychiatrische Forschung der Psychiatrischen Universitätsklinik (PUK) in Zürich.
Masterarbeit an der KJPP/PUK (Developmental Neuroimaging Group): EEG/fMRT-Studie zu Dyslexie
zu vergeben sind zwei Masterarbeiten in zwei Projekten der Developmental Neuroimaging Group am Zentrum für kinder- und jugendpsychiatrische Forschung der Psychiatrischen Universitätsklinik (PUK) in Zürich.
Are we done getting taller?
Trends in height and growth in Switzerland over the past 200 years are analysed. We combine monitoring data (size at birth, schoolchildren, conscripts) as well as survey and family data. Of interest are inter-generational effects as well as socioeconomic/regional differences.
Control Algorithm Development for a Series Elastic Actuated Exoskeleton
Robot-assisted therapy of stroke patients is a promising approach to improve the therapy outcome and to tackle the challenges of demographic ageing. Using a novel 6 DoF exoskeleton prototype, we want to investigate/develop new control strategies for the next generation of these robots.
VIAN Usability Study
Support a usability study with a controlled group of external VIAN users. Task is to coordinate the test results, adapt the specifications and manage the evaluation in close consultation with the developer of VIAN, Gaudenz Halter.
The effect of psychosocial and craving-induced stress on transcriptomics in cocaine users: a longitudinal approach
The Master candidate will assess the gene-expression (transcriptomics) alterations as a result of psychosocial stress and cocaine drug craving-related stress in cocaine users and matched healthy controls. Transcriptomic and statistical techniques will be applied.
Cardiac Magnetic Resonance Image Reconstruction Using Machine Learning
Dynamic Magnetic Resonance (MR) imaging offers exquisite views of cardiac anatomy and function. The objective of this project is to develop and implement methods that allow learning a data model from large sets of training data to be used in nonlinear data recovery from highly undersampled MR data.
Development of Automatized Finite Element Models for Instrumented Spine Simulations.
Implementation of different surgical interventions, such as posterior instrumentation and cage placement into an automatic FE-Element pipeline, for functional outcome prediction.
Combining Calcium Imaging and fMRI: Understanding the Neurophysiology of the Blood Oxygen Level Dependent Response
Advances in MR technology are rapidly increasing our imaging capabilities, however we still don't know what the BOLD signal actually represents. In the lab we try to gain a better understanding of the properties of brain circuits to increase the capabilities of fMRI.
Building of a binary machine learning classifier for radiographic images using the Google Seedbank platform
Potential applications for predictive models providing a binary classification based on radiographic images are manifold. A modular machine learning application that can be trained with different radiographic data sets without any programming skills would, therefore, be desirable.
Optimisation of acquisition and reconstruction parameters for a portable cone-beam computed X-ray tomography system
At our Institute, we have recently developed a light, portable, and highly configurable X-ray computed tomography system. To further optimize the quality of our scans, we want to systematically test various system settings during image acquisition and image reconstruction for different sample types.

Powered by  SiROP - the academic career network