The detection of involuntary and usually imperceptible “horror tranquility” hand is crucial for early diagnosis and effective treatment of neurological diseases such as Parkinson’s disease. Until recently, the diagnosis was mainly subjective, as the estimate of the frequency and intensity of the phenomenon was based on the notion of treating. Furthermore, clinical symptom recording conditions using accelerometers alters the nature of the condition, and enters the invasive agent in the examination of the patient.
The objective criteria, continuous and non-invasive recording of hand motion meets interaction with the mobile phone. The application of appropriate techniques of signal processing and machine learning to signals recorded by epitanchynsiometro mobile phone can automate the process of rest tremor existence assessment in an objective and quantified way.
The “multimedia understanding group” of laboratory of “Information Processing and Computation” Department of Electrical and Computer Engineering from the Polytechnic School of Thessaloniki, under the direction of Prof. Anastasios Delopoulos, after presenting the predominant signal processing techniques for the detection of terror, would “cause” attendees in an “hackathon” to implement automatic rest tremor detection algorithms from such recordings.
Wednesday 7 December 2016 | Aristotle University’s Research Dissemination Center (KEDEA) Conference Hall II
6.00-6.30 pm .: Problem presentation and a proposed sequence of solving it
6.30-8.30 pm .: Conducting «Hackathon» contest
8.30-9.00 pm .: Results Assessment, Announcement of the winner and presentation of the solution
It is required basic familiarity, both theoretical and at the implementation level in Matlab, the following
• Signal Processing Techniques
Signal processing, statistical estimation, autocorrelation / cross correlation, spectrum, etc.
• Technical machine learning (basic knowledge)
Education and use classification models (SVM, ANN, etc.)
Matlab 2014b (or later)
Alternatively, older Matlab version in combination with libSVM
* Entries can be individual or in groups of two persons.
Facebook Event: https://www.facebook.com/events/588314358035547/