The study not only confirms that PD significantly affects emotional perception, but also demonstrates that EEG can be used to identify these differences with high accuracy. The use of machine learning techniques and advanced EEG descriptors allows both the differentiation between PD and healthy individuals and the understanding of the emotional deficits specific to PD.
Parkinson’s disease (PD) is a neurodegenerative condition that affects the central nervous system, mainly compromising motor control. Patients with PD often present with tremors, muscle stiffness and slowness of movement.
In addition to these motor symptoms, the disease also causes cognitive, behavioral and emotional impacts, affecting the quality of life of more than 10 million people worldwide.
This condition is associated with the progressive loss of dopaminergic neurons in the substantia nigra, an area of the brain responsible for regulating movement, but which also plays important roles in emotional and cognitive perception.
Research has investigated how PD influences emotional perception and processing. Among the tools used, the electroencephalogram (EEG) has gained prominence.
EEG measures the brain's electrical activity in real time, using sensors placed on the scalp. It is widely recognized as a non-invasive technique, with high temporal precision and easy to apply.
Traditionally, EEG is performed in a resting state, where the patient remains still, often with their eyes closed in a controlled environment. However, recent studies suggest that using EEG during more natural and engaging tasks, such as watching movies or listening to music, can provide more representative insights into patients' emotional perception.
The ability to recognize emotions is essential for communication and social interactions. Previous studies indicate that patients with PD have difficulty identifying emotions, both positive and negative, in facial expressions and tone of voice (prosody).
In addition, they have a reduced emotional response to intense visual stimuli, such as emotionally charged images. These limitations directly affect the social interactions and quality of life of these patients, making it difficult to understand emotions not expressed verbally.
In this study, researchers sought to explore how PD patients process emotions and how these responses can be used to diagnose the condition. The dataset comprised EEG signals from 20 non-demented PD individuals (10 males/10 females) and 20 controls (9 males/11 females) from Hospital Universiti Kebangsaan Malaysia,
Data collection involved analyzing EEGs while they performed emotional tasks, such as watching movies or listening to music. These stimuli were chosen because they are effective in eliciting a variety of emotions, allowing for a more detailed assessment of patients’ emotional perception.
Rather than relying solely on resting-state stimuli, which can be artificial and limited, researchers opted for more realistic contexts in which emotional responses are more spontaneous and representative.
The collected EEG data was processed using specific descriptors that help capture brain patterns associated with emotions. One of the methods used was the analysis of spectral power vectors (SPVs), which measures the intensity of brain waves in different frequency ranges, such as alpha, beta and gamma waves, each associated with different emotional and cognitive states.
Another approach was the use of common spatial patterns (CSPs), which highlight the differences in brain signals between the groups studied: PD patients and healthy controls (HCs).
In addition, the researchers created visual maps called topomaps, which show the spatial distribution of electrical activity in the brain, and also analyzed how this activity evolves over time during the tasks, creating descriptors of EEG “movies”. To interpret the data, the researchers applied advanced machine learning techniques, including deep neural networks.
These algorithms were trained to identify specific patterns in the brain signals that could differentiate PD patients from healthy individuals and recognize the emotional responses associated with different categories, such as happiness, sadness or fear.
This analysis also allowed us to identify broader emotional dimensions, such as valence, which refers to the pleasant or unpleasant nature of an emotion, and arousal, which measures emotional intensity.
The results revealed that patients with PD have significant difficulty processing emotions related to valence, that is, identifying whether something is positive or negative.
In terms of specific emotional categories, these patients demonstrated greater accuracy in recognizing sadness, but had difficulty identifying emotions such as fear, disgust, and surprise. These confusions were confirmed by analyses that showed frequent errors in interpreting opposing emotions, such as mistaking positive for negative stimuli.
In addition, brain signals captured by EEG allowed almost perfect distinction between PD patients and healthy controls. This suggests that EEG may be an effective tool for diagnosing PD, especially because it detects implicit emotional responses, those that do not rely on patients’ self-reports, which can be subjective or limited.
This study demonstrates that EEG can be used as a practical and sustainable tool for diagnosing PD and understanding its emotional implications.
Examining EEG in more realistic scenarios, such as during emotional tasks, offers a more ecological and representative approach, avoiding the limitations of studies based solely on resting state.
These findings not only highlight the potential of EEG to identify PD-related emotional deficits, but also pave the way for the use of this technology in more accurate clinical diagnostics and in monitoring the progression of the disease.
READ MORE:
Exploring Electroencephalography-Based Affective Analysis and Detection of Parkinson’s Disease
RAVIKIRAN PARAMESHWARA, SOUJANYA NARAYANA, MURUGAPPAN MURUGAPPAN, IBRAHIM RADWAN, ROLAND GOECKE, and RAMANATHAN SUBRAMANIAN
INTELLIGENT COMPUTING. 17 Oct 2024. Vol 3. Article ID: 0084
DOI: 10.34133/icomputing.0084
Abstract:
While Parkinson’s disease (PD) is typically characterized by motor disorder, there is also evidence of diminished emotion perception in PD patients. This study examines the utility of electroencephalography (EEG) signals to understand emotional differences between PD and healthy controls (HCs), and for automated PD detection. Employing traditional machine learning and deep learning methods on multiple EEG descriptors, we explore (a) dimensional and categorical emotion recognition and (b) PD versus HC classification from multiple descriptors characterizing emotional EEG signals. Our results reveal that PD patients comprehend arousal better than valence and, among emotion categories, fear, disgust, and surprise less accurately, and sadness most accurately. Mislabeling analyses confirm confounds among opposite-valence emotions for PD data. Emotional EEG responses also achieve near-perfect PD versus HC recognition. Cumulatively, our study demonstrates that (a) examining implicit responses alone enables (i) discovery of valence-related impairments in PD patients and (ii) differentiation of PD from HC and that (b) emotional EEG analysis is an ecologically valid, effective, practical, and sustainable tool for PD diagnosis vis-à-vis self-reports, expert assessments, and resting-state analysis.
Comments