Interview with Professor Ioannis Iossifidis, Head of Robotics and BCI Laboratory, Ruhr West University of Applied Sciences
Neurological disorders like Parkinson's are often diagnosed once the disease has already progressed to a later stage. The VAFES project was initiated to facilitate an early detection. Sensor technology and VR are used in the creation of a playful test system.
Professor Ioannis Iossifidis, Head of Robotics and BCI Laboratory, Ruhr West University of Applied Sciences
Professor Iossifidis sat down with MEDICA.de and told us what prompted the project, explained the next steps and revealed its objective.
Professor Iossifidis, what prompted you to launch the VAFES project?
Prof. Ioannis Iossifidis: This project centers on patients with arm dysfunction due to a neurological disorder, with a primary focus on Parkinson's patients. The goal here is to develop a reliable diagnostic test system. This should make it possible to classify the signs and symptoms that indicate the disease. In addition, the project aims to facilitate an automatic follow-up and assessment of possible treatment regimens.
There is a standardized arm test, the so-called Action Research Arm Test (ARAT), which basically consists of a carrying case and some wooden parts, washers, etc. Patients who experience dysfunction when gripping things or have reduced arm movement perform specific exercises under close observation by a doctor. Based on the test, the physician makes a classification to determine the stage of the disease. However, this is not a very accurate method and the scoring is rather subjective, which is also reflected in the assessments.
More topic-related exciting news from the editors of MEDICA-tradefair.com:
The test system should be usable in smaller clinics as well.
We want to add a more customized exercise approach to this diagnostic system. Arm and hand movements are separated in virtual reality (VR). In doing so, detailed information about the movement patterns, speed and acceleration can facilitate an accurate analysis. In doing so, the current state assessment is based on objective data for the very first time ever.
What’s more, as part of the project, we hope to develop models that accurately analyze the arm movements, thus making an early detection of Parkinson's disease possible. At present, a DaTscan, also known as a PET/CT scan, is used as a tool to confirm the diagnosis of Parkinson’s disease. Unfortunately, emission computed tomography is very expensive, making it a rarely performed imaging test. The Bergmannsheil Bochum Hospital, for example, performs a DaTscan only two to three times per year.
What role will virtual reality (VR) play in your project?
Iossifidis: I am confident that after the project has been completed, we will be able to offer a solution that can be implemented as a diagnostic test using a VR headset and the respective gloves. To collect the data, the patients are tested via a kind of virtual reality game. This data can then be used to determine the progression and stage of the patient's disease.
How does the sensor-based smart glove work?
Iossifidis: Our project objective is to develop an active sensor-based smart glove. It is made of fabric and includes tendons and small motors and is intended for people who have difficulties closing and opening their hands. The glove aims to gently compensate for the weak grip strength. This aspect is controlled by a muscle stimulation module. At the same time, trackers measure the movement of the hand.
Over the course of the project, we will minutely measure muscle activity to develop a device that stimulates the antagonist muscles to offset the tremor. The goal is to achieve the same functionality as a brain implant. The latter is frequently used to reduce tremors.
With the help of the test system, an earlier diagnosis should be possible.
What are your next steps?
Iossifidis: We already started the development of the virtual reality components. We initiated the device design process and developed game scenarios and motion-based activities for the respective patient groups we plan to assess.
First, we start by collecting data from healthy patients as a baseline for later comparison. The data is subsequently effectively classified and sorted using machine learning.
Next, we plan to create a generative model for arm movements. The resulting software can detect anomalies in movement and the degree of the anomaly, allowing classification of Parkinson’s disease stages. An approach that combines EEG data and machine learning can find correlations and indicators of the biosignals that are characteristic of the disease.
How can this type of test system impact early detection of the disorder?
Iossifidis: Sadly, patients with suspected Parkinson’s disease typically don’t undergo a computerized tomography scan. In light of that, it is essential to develop this type of test system. Unfortunately, patients with Parkinson's disease don’t seek the support of physicians until symptoms and tumors have become severe, even though the first early signs appear years before the classic motor issues.
Ideally, we will develop a test for medical practices that facilitates the detection of the disorder in the early stages and involves a 10-minute routine exam using a VR headset and a glove.