In order to write a thesis at KIS*MED, interested students can send an e-mail to one of the team members <lastname>@kismed.tu-darmstadt.de. Please send your transcript of records (Leistungsspiegel) and a short letter of motivation. All topics can be customised.
Possible topics and the relevant contact persons are listed below:
Electrical Impedance Tomography (EIT)
With electrical impedance tomography (EIT), cross-sectional images of the human body can be generated by injecting small amounts of electrical current and measuring the resulting voltages. Unlike computed tomography (CT), no harmful X-rays are used, and the system is also significantly more cost-effective than an MRI. The potential applications of EIT are diverse: it can be used for respiratory monitoring, for examining the musculoskeletal system, and even in a miniaturized form for monitoring cell growth.
In EIT, both simulated measurements and measurements conducted on test tanks or human subjects are used. Our research focuses on how we can collect large datasets from real measurements and reconstruct them into actual cross-sectional images. These images, for example, could be used to monitor fracture healing.
Possible research topics include:
- Development of new reconstruction algorithms, for example, using machine learning
- Further development of the existing measurement setup
- Investigation of potential application areas for EIT
- Simulation of EIT measurements in Matlab
- Comparison between simulation and real-world data
While knowledge of Python and/or Matlab is recommended, it is not strictly required. Prior experience in machine learning/deep learning can be beneficial for certain topics.
Trustworthy and Explainable AI Solutions for Medical Diagnosis and Treatment Planning
Artificial Intelligence (AI) has immense potential to transform healthcare by improving diagnosis accuracy and enabling personalized treatment plans. However, the lack of trust and transparency in AI models, especially in the medical field, limits their widespread adoption. We aim to explore the development of trustworthy and explainable AI solutions for medical data (CT scans, MRIs, OCT etc) that enhance healthcare professionals' confidence in using AI for medical decision-making.
Possible topics are:
• Explainable AI (XAI): Develop and integrate explainable AI techniques, to make model decisions transparent and understandable to healthcare professionals.
• Trustworthy AI: Ensure the AI model adheres to principles like fairness, accountability, and transparency, addressing biases and providing reliable results.
• Synthetic Data Generation: Use advanced synthetic data generation methods (e.g., GANs, data augmentation) to create realistic, privacy-preserving medical datasets that supplement real data for training AI models.
• Patient Emotion Analysis: Integrate emotion detection techniques, such as facial recognition, speech analysis, and text mining from patient interactions, to assess emotional states and improve treatment plans and communication strategies.
Using tools such as Python, Matlab, classical signal, and image processing.
Robust multimodal medical time series analysis
The quality of treatment for various diseases (such as epilepsy and cardiovascular diseases) has been significantly improved by their early detection using medical measurement methods (such as ECG, EEG, EMG, PPG, CT, MRI, etc.). Unfortunately, these measurement methods often suffer from artifacts and missing sensors, so that they only provide noisy and incomplete data. We are therefore researching robust algorithms that can deal with these artifacts and supplement missing information, for example by fusing with other modalities.
Possible topics for processing are
Robust machine learning & interpretability
Methods for sensor fusion (cameras, wearables, …)
Unsupervised / self-supervised learning for time series
Transfer learning between different modalities
Knowledge of Python and signal processing is recommended. Theoretical knowledge of machine learning / deep learning can be an advantage for some topics.
Vital parameters such as pulse and respiration are important indicators of various diseases and the general state of health of a person. The gold standard methods, such as electrocardiography for monitoring the heart, usually require contact with the patient, e.g. in the form of adhesive electrodes. We are researching ways to replace these with cameras where possible.
Possible topics include
- Machine Learning & Interpretability
- Multiphysics simulations and rendering of synthetic video data
- Methods for sensor fusion (cameras, wearables, …)
- Photoplethymography imaging
- Robust estimation of heart rate variability
Knowledge of Python (or alternatively Matlab) and signal processing is recommended. Knowledge of Blender or Unreal Engine may be an advantage for some topics.
Stress is a common cause of sleep problems, concentration disorders, depression or even cardiovascular disease. While stress is often assessed purely subjectively, we are working on recording vital signs in order to make a quantitative statement about a person's well-being. The vital signs are primarily to be recorded non-invasively in order to enable the end user to be as comfortable as possible. In addition, stress can also be recorded in people who are unable to articulate this independently, such as newborn babies or people with disabilities.
While research is mainly focussed on the development of stress caused by acoustic stimuli such as noise, it is not limited to this. Possible areas of application include contactless parameter recording at the workplace, in the car or in bed. The parameters are to be analysed primarily with the help of machine learning algorithms.
Possible topics:
- Non-invasive vital parameter estimation based on cECG, thermographic camera and PPGI
- Sensor fusion for robust parameter estimation
- Quantitative analysis of stress
- Application of machine learning with a focus on computer vision and signal processing
- Implementation of virtual reality environments for test setups
Programming skills in Python or Matlab are recommended. If you are looking for a topic in the field of VR, it is recommended that you have knowledge of C# and Unity programming.
Physiotherapy is an important part of the treatment of many injuries and illnesses. Ideally, physiotherapy is carried out under the supervision of a healthcare professional who can provide personalised and immediate feedback. But even without direct supervision, exercises at home can be beneficial to the healing process. At the same time, incorrect execution, misjudgement of your own fitness level and overexertion can also lead to inefficient training or, worse still, serious injury.
To mitigate these problems, an automated scoring system can be used to assess the quality of exercise execution and reduce the need for human supervision. Building on advances in the field of computer vision, we are researching the development of such support systems using video-based estimation of human posture.
Possible topics include
- Natural language processing for the automated generation of training data from videos with subtitles
- Analysis of sensor data (EMG, force plates, IMUs) and relation to camera recordings
- Development of biomechanical models of human movement
- Development of methods for the quantitative evaluation of human movement
- Machine learning & computer vision
- Transfer learning & domain adaptation (transfer to infrared cameras)
Knowledge of Python (or alternatively Matlab) and signal processing is recommended. Knowledge of Blender can be an advantage for some topics.
Video-based Human State Analysis refers to the use of video footage to assess and interpret the physical, emotional, or psychological states of individuals. This involves analyzing visual data captured by cameras to derive insights about a person's behavior, emotions, actions, or overall well-being.
Possible topics:
- Methods for detecting/tracking human limb movements to correlate with the individual's current state
- Methods for human reflex analyses
- Quantitative analysis of human emotional state
- Machine learning and computer vision
Using tools such as Python, Matlab, classical signal, and image processing
Exceptional Theses
In this section, we present theses that we found to be exceptionally good. With this, we aim to both recognize the excellent work of the authors and provide examples for future students.
Titel | Year | Art | Author |
---|---|---|---|
(opens in new tab) Multimodale Registrierung und Analyse von Infrarot- und Tiefenkameradaten | 2022 | Bachelor | Larissa Werner |
(opens in new tab) Monocular 3D Human Pose Estimation From Thermographic Images Using Neural Networks | 2023 | Bachelor | Julian Imhof |
(opens in new tab) Machbarkeitsanalyse zur Bildgebung mittels miniaturisierte Elektrische Impedanz-Tomographie | 2024 | Bachelor | Emily Reinhold |
(opens in new tab) Improving feature extraction for camera-based motion analysis using trajectory similarity | 2024 | Master | Malte Mai |
(opens in new tab) Development and Optimization of a motion sensitive adaptive ROI extraction for remote photoplethysmography | 2024 | Master | Philipp Witulla |
Exceptional final presentations
Titel | Year | Art | Author |
(opens in new tab) Sensorfusion zur Bewegungsanalyse von physiotherapeutischen Übungen | 2024 | Bachelor | Luise Herrmann |
Finished Theses
Topic | Bachelor / Master | Status | Supervisor |
---|---|---|---|
Pattern Recognition in the Capacitive Electrocardiogram and Reconstruction of the Reference Electrocardiogram |
Master | finished | Maurice Rohr |
Machine Learning in der EKG-Diagnostik – Auswirkungen von typischen Fehlern auf Klassifikatoren |
Bachelor | finished | Maurice Rohr |
Robust Non-Contact Vital Sign Monitoring of Sleep Lab Patients Using Image Fusing and Deep Learning | Master | finished | Maurice Rohr |
Entwicklung und Quantifizierung einer videobasierten Analyse von Tremor |
Bachelor | finished | Christoph Hoog Antink |
Analysis and Optimization of Photoplethysmography Imaging Methods for Non-Contact Measurement of Heart Variability Parameters | Master | finished | Maurice Rohr |
KI-gestützte Bewegungsanalyse basierend auf Thermographie- und RGB-Bildgebung | Master | finished | Sebastian Dill |