Medical Physics

Group leader: Dr. Manuel Blessing

Projects:

  • kV-MV CBCT imaging
    • Commonly used acquisition times of 30s for patient setup with cone-beam CT (CBCT) in radiotherapy of lung tumors are beyond the breath-hold capacity of most patients, resulting in motion artifacts. To reduce the rotation interval to 90° for an accelerated data acquisition both megavoltage (MV) and kilovoltage (kV) photon sources can be used thereby minimizing acquisition time to 15s. To achieve this goal a clinical linac with kV CBCT and MV portal imaging capability was modified with dedicated hardware to enable synchronized and simultaneous data acquisition. An application program was developed which communicates with various linac components via serial port to manage MV detector readout and MV projection angle capture online. Projection and reconstruction based MV to kV data conversion methods were investigated. Currently risk analysis is done as preparation for clinical application.

  • Decision tool for an optimized Cancer therapy

    • The tumor board is a meeting of physicians from different disciplines (radiologist, pathologist, radiation oncologist, etc.) where the current cancer patient cases are discussed. To decide for a certain therapy for a given patient the physicians rely on their experience, i.e. the decision may not be objective. Especially for less-experienced physicians or in the case of a seldom cancer type the optimal solution may not be found. The goal of this project is the development of a decision tool based on the medical record of patients from the past: The case-based reasoning (CBR) method is used to grab the most similar cases from a database.  Currently our own patient database (University Medical Center Mannheim) is used, although it would be more effective to merge the data from many clinics.

  • Classification of progress versus pseudo-progress after resection of a glioblastoma

    • A glioblastoma is an aggressive brain tumor. The affected patient has a survival perspective of 5 months on average. After resection of the tumor and radiation therapy perfusion MR data is acquired to check whether the tumor starts growing again (progress). However, classical methods do not allow a clear differentiation between progress and an inflammation of the lesion (pseudo-progress). Retrospective patient data from our database is used as training data for a neural network to distinguish between progress and pseudo-progress on a machine learning basis.

  • Telemonitoring system for individualized follow-up of cancer patients
    • Especially geriatric cancer patients require an individualized and intensive follow-up. In this project an individualized patient assessment and telemonitoring system is developed. Physiological and psychological data from the patient are collected via smart devices and submitted to a database. Pattern recognition methods are developed to interpret the data, i.e. patient reported outcome (from measurements and questionnaires) is used to estimate the patient’s objective wellbeing. In case of abnormality an intervention can be initiated, e.g. make an appointment to examine the patient in the clinic. This means that the patient’s follow-up becomes more independent from his social environment and a severe course of the disease might be adverted by a directed intervention at an early stage.
  • Ultra low-dose tumor tracking based on single photon counting detectors

    • Respiratory motion is one of the big challenges in radiation therapy of tumors in the lung, liver or upper abdomen. To track the tumor during the irradiation using the conventional CBCT unit mounted on the linac gantry a high dose would be required which makes this approach unfeasible. The goal of this project is to use ultra low-dose (ULD) source and detector (based on single photon counting technology) to allow for continuous imaging during irradiation. From a 4D-CT of a patient several breathing phases can be defined and serve as a reference. A maximum likelihood approach is used to differentiate between the breathing phases. To potentially improve the robustness of this estimation an image-feature based approach is currently investigated.

  • Gait improvement based on an intelligent orthosis

    • Children with cerebral palsy (CP) suffer from movement disorders, e.g. a deviant and harmful gait. Video-based gait analysis combined with an orthoses that is worn in everyday life is one approach to improve a child’s gait. However, the child might fall back to his habitual gait when there is no corrective action. Therefore, we equip the orthosis with different sensors (e.g. pressure, bending and acceleration) and analyze the data to differentiate normal from disordered gait. The analysis software is based on machine learning. The gait data is acquired in the lab and used to train a neural network. Finally, an app collects the sensor data and the classification can be accomplished directly on the smartphone. This enables a direct feedback to the patient.