In recent years, Deep Brain Stimulation (DBS) techniques using electrical signals have been studied. These techniques allow us to reduce certain brain diseases’ symptoms, as is the case in Parkinson’s disease, which causes heavy trembling in patients’ limbs.
Marketed systems use stimulation signals with a fixed
shape, which may entail certain secondary effects and an increase in the
long-term symptoms, despite of the continuous use of these methods. That is why
DBS adaptative algorithms are being developed. These algorithms can adapt to
the optimal stimulating current depending on the patient’s state. They operate
in one way or another depending on the information provided by various
A biomarker is a tool used for obtaining different
biometric data from the patient. The most common biomarker when treating with
DBS is the local field potential (LFP), whose signal amplitude is closely
related to the symptoms being experienced.
One of the most widely known adaptative algorithms, the dual threshold algorithm, makes use the of the LFP signal as input data. This algorithm consists of varying the amplitude value of a PWM stimulation signal depending on the data provided by the biomarker. In this project the goal is to develop an algorithm which may serve as an alternate method to the dual threshold algorithm. This algorithm seeks to balance between efficiency (data looked for in the biomarkers) and energy consumption, as reducing the latter may result in a longer lifespan for the device running it. As being an implantable system requires it to have the longest lifespan possible
Nowadays, artificial neural networks (ANNs) are computational models that, while they have solved many different problems, they require a large amount of memory to execute those solutions. Therefore, their implementation is more common in systems with high-performance capabilities, such as data centers or servers. However, there is an increasing interest in developing these solutions on devices with fewer resources, such as personal computers, mobile devices, and microcontrollers. This situation has led to the proliferation of multiple techniques and tools to reduce the requirements of these models and allow their implementation on those platforms.
The main objective of this project
is the evaluation of a tool for the implementation of artificial neural
networks in microcontroller-based systems. For this
purpose, a practical use case has been defined.
To achieve this objective, the classification of types of tremors in Parkinson’s patients was
chosen as the practical use case. In addition, the Arduino Nano 33 BLE Sense
has been also chosen as the microcontroller-based platform to implement the
solutions of the practical use case.
Later, a dataset has been generated
with real measurements from a sensor on that platform. With this dataset
various experiments have been carried out to determine how the different
structures of artificial neural networks deal with the chosen use case.
Then, based on those experiments,
some appropriate ANNs for classifying types of tremors have been designed, trained,
and evaluated in a system with graphics and tensor processing units and in a microcontroller-based
Finally, based on the results
obtained, the trade-offs between classification accuracy, memory footprint and
inference latency involved in the implementation of ANNs solutions in
microcontroller-based systems have been determined. In addition, a discussion
of the strengths and weaknesses of the selected tool for the implementation of artificial
neural networks in microcontroller-based systems has been presented.