Neurosalience #S2E15 with Pedro Valdes-Sosa - EEG analysis: Past, present and future
Jan 26, 2022
auto_awesome
Pedro Valdes-Sosa, an expert in EEG analysis, discusses his early career, creative work in EEG signal interpretation, his visiting position in Chengdu, China, the growing EEG database, and his international consortium. He briefly touches on medical care in Cuba and Cuba's response to COVID-19. They also explore topics such as universal screening for hearing loss, challenges in source analysis for EEG and MEG, intelligence and brain structure in Latin America, sharing multinational EEG data, integrating science and public health, COVID-19 longitudinal study and brain-induced disorders, the impact of the pandemic on science inequity, and challenges and optimism in spreading science.
Pedro Valdes-Sosa has been a pioneer in the clinical use of EEG since the 70s and continues to innovate in interpreting EEG signals today.
He is actively involved in developing a growing EEG database and leading an international consortium dedicated to improving science development in less developed countries.
The challenges in EEG analysis include artifact removal, automated methods development, and establishing consistent annotation protocols for improving accuracy and reliability.
Deep dives
Crystallizing Fluid Intelligence is Predicted by White Matter Microstructure
This study explored the relationship between white matter microstructure and fluid intelligence. Using fractional anisotropy, the researchers analyzed specific white matter tracts and found a significant positive correlation between the microstructure of these tracts and fluid intelligence. This indicates that more ordered white matter tracks are associated with higher levels of fluid intelligence. The findings highlight the importance of white matter integrity in cognitive function.
Investigating the Alpha Rhythm and Cortical Circuits
This study delved into the characteristics of the alpha rhythm and its relation to cortical circuits. The results suggested that the alpha rhythm may be a result of cortical circuits rather than a global standing wave. The research used EEG measurements across different age ranges and explored the changes in the alpha peak frequency over time. The findings shed light on the complex nature of the alpha rhythm and its possible underlying mechanisms.
Pooling EEG Data for Biomarker Discovery
Pooling EEG data has become an important approach in biomarker discovery. Researchers are increasingly utilizing large-scale datasets to identify EEG-based biomarkers for various applications. International initiatives, such as EEGNet and the Global Brain Concern, aim to promote data sharing, open science, and the development of robust biomarkers. Annotated data and expert labeling are essential for removing artifacts and improving the accuracy of EEG-based biomarker identification.
Addressing Challenges in EEG Annotation and Artifact Removal
One of the major challenges in EEG analysis is the presence of artifacts. These can arise from various sources, such as eye movements or head motion, and can significantly affect the interpretation of EEG data. Efforts are underway to develop automated methods for artifact removal, but visual inspection by certified EEG technicians remains crucial. Additionally, establishing consistent annotation protocols and addressing the ambiguity of certain EEG features, such as graph elements, are essential for improving the accuracy and reliability of EEG analysis.
The Importance of Biomarker Research and Unsupervised Machine Learning
Biomarker research and the use of unsupervised machine learning techniques are crucial for advancing healthcare. A recent paper in GigaScience explores the search for biomarkers unsupervised, which is an exciting development. Unsupervised machine learning can identify signals and patterns in data without prior knowledge or guidance. By comparing results and gaining the trust of clinical professionals, population modeling with machine learning can enhance measures of mental health. This collaborative approach between artificial intelligence and medical experts is essential in revolutionizing healthcare.
Harmonization and Batch Correction in Brain Research
Harmonization and batch correction are critical in brain research, particularly in studies involving different sites and vendors. The Global Brain Consortium has been working on a multinational normative dataset to address these challenges. Using simple measures like cross-spectrum analysis from EEG data, researchers can identify outliers and correct data inconsistencies. The importance of harmonization is evident in the improved classification performance achieved with batch correction techniques. Ultimately, cooperation and the development of a vendor dictionary can lead to more accurate and reliable brain research outcomes.
In this episode, we discuss what was important to Pedro early in his career. He describes his first forays into clinical use of EEG back in the 70s and then we go on to discuss some of his highly creative work in deeply interpreting EEG signals today. Later we discuss his current visiting position in Chengdu, China and a growing EEG database as well as his international consortium. We touch briefly on the current state of medical care in Cuba as well as how Cuba has dealt with COVID-19. This episode was recorded on October 22nd 2021.
Guest:
Pedro Valdes-Sosa is the General Vice-Director for Research of the Cuban Neurosciences Center, which he co-founded in 1990. He studied medicine at the University of Havana, and graduated in 1972. He also studied Mathematics in 1973. He obtained his Ph.D. in 1978. In 1979 he did a PostDoc on "Neurometrics and Computational Techniques" and "Biophysical Modeling of brain electrical activity" with Prof. E. Roy John at the Brain Research Lab of New York University. He is a full member of the Cuban Academy of Sciences, and the Latin American Academy of Sciences, associate member of the International Center for Theoretical Physics.
Pedro is known not only for his innovation and rigor in EEG analysis but also for his highly collaborative work and passion to improve science development, communication and dissemination in less developed countries. He’s currently flying back and forth between Havana and Chengdu, China where he is developing pooled databases for quantitative EEG.
Remember Everything You Learn from Podcasts
Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
