Unsupervised Detection of Sub-Territories of the Subthalamic Nucleus During DBS Surgery with Manifold Learning
Authors:
Ido Cohen,
Dan Valsky,
Ronen Talmon
Abstract:
During Deep Brain Stimulation(DBS) surgery for treating Parkinson's disease, one vital task is to detect a specific brain area called the Subthalamic Nucleus(STN) and a sub-territory within the STN called the Dorsolateral Oscillatory Region(DLOR). Accurate detection of the STN borders is crucial for adequate clinical outcomes. Currently, the detection is based on human experts, guided by supervise…
▽ More
During Deep Brain Stimulation(DBS) surgery for treating Parkinson's disease, one vital task is to detect a specific brain area called the Subthalamic Nucleus(STN) and a sub-territory within the STN called the Dorsolateral Oscillatory Region(DLOR). Accurate detection of the STN borders is crucial for adequate clinical outcomes. Currently, the detection is based on human experts, guided by supervised machine learning detection algorithms. Consequently, this procedure depends on the knowledge and experience of particular experts and on the amount and quality of the labeled data used for training the machine learning algorithms. In this paper, to circumvent the dependence and bias caused by the training data, we present a data-driven unsupervised method for detecting the STN and the DLOR during DBS surgery. Our method is based on an agnostic modeling approach for general target detection tasks. Given a set of measurements, we extract features and propose a variant of the Mahalanobis distance between these features. We show theoretically that this distance enhances the differences between measurements with different intrinsic characteristics. Then, we incorporate the new features and distances into a manifold learning method, called Diffusion Maps. We show that this method gives rise to a representation that is consistent with the underlying factors that govern the measurements. Since the construction of this representation is carried out without rigid modeling assumptions, it can facilitate a wide range of detection tasks; here, we propose a specification for the STN and DLOR detection tasks. We present detection results on 25 sets of measurements recorded from 16 patients during surgery. Compared to a competing supervised algorithm based on a Hidden Markov Model, our unsupervised method demonstrates similar results in the STN detection task and superior results in the DLOR detection task.
△ Less
Submitted 23 August, 2022;
originally announced August 2022.
