Cutting Garden 2023: Los Angeles Garden!

As part of the CuttingGardens2023 conference.

The Los Angeles (LA) Garden will be held at the Ming Hsieh Department of Electrical and Computer Engineering (EEB) at the University of Southern California (USC). This event is scheduled to be an in-person gathering, but there will also be an option to broadcast it to all Gardens.

The LA Garden will span across two days. On the first day, we have scheduled a Brainstorm workshop and a Brainsuite Demo. The second day will feature local speakers presenting cutting-edge methods on Deep Neural Networks and Machine learning techniques applied to EEG data.

Furthermore, we are arranging a live stream of the Global Program, allowing anyone interested to join remotely. External attendees are more than welcome to participate, but prior registration will be required for attendance.

WEDNESDAY 18th

5:30 – [Global] Reproductible processing pipelines and multiverses

8:30 – Registration & Continental Breakfast

9:15 – Richard Leahy (University of Southern California):

Opening of the LA Garden

9:30 – David Shattuck (University of California, Los Angeles):

• Introduction and Overview of Brainsuite

10:15 Coffee Break

10:30 – Anand Joshi (University of Southern California):

• Brainsuite Tools & Discussion

11:00 – Richard Leahy (University of Southern California):

• Introduction to EEG/MEG Analysis

11:45 – Cameron Sacks (Wearable Sensing):

• EEG Live Demo & Discussion

12:00 – Lunch Break

12:30–17:30 – Tutorial – Hands-On Brainstorm

• Raymundo Cassani (McGill University) & Takfarinas Medani (University of Southern California)

12:30–13:00 Onsite assistance in installing the material for the training session

13:00–13:30 Introduction to Brainstorm (lecture) - Youtube Video -

13:30–14:35 Loading anatomy and recordings

• Set anatomy
• Review Raw recordings
• Import events

14:35–15:35 Pre-processing

• Frequency filters
• Artefact detections
• Artifact correction with SSP

15:30–15:45 Coffee Break

15:45–16:20 Analysis sensor level

• Import recording

• Review trials

• Trial averages

16:20–16:55 Source estimation

• Forward Model (aka Head model)

• Noise covariance matrix

• Source estimation (from EEG and MEG recording)

16:55–17:15 Analysis source level

• Cortex parcellations: Atlases and scouts

• Noise covariance matrix

• Source estimation (from EEG and MEG recording)

THURSDAY 19th

5:30 – [Global] Deep Neural Network (DNN) analysis for MEEG data

8:30 – Continental Breakfast

9:00 – 17:30 – [Local] Machine Learning and EEG

Session 1: Computational Tools and Pipelines for ML analysis; session chair: Richard Leahy.

• 9:00–9:25- Arnaud Delorme (University of California, San Diego): “Machine learning and the BIDS EEG data format”

• 9:25–9:50- Tim Mullen (Intheon Labs):

• “Creating Deployable Workflows for EEG Signal Processing and ML/DL Using NeuroPype”

• 9:50–10:15– Ivan Tashev (Microsoft Research):

• “Workload estimation using brain- and bio- signals for adaptive training system”

• 10:15–10:40– Bin He (Carnegie Mellon University):

• “AI/ML Enhances Dynamic Brain Imaging from EEG/MEG”

Coffee Break (10:40–11:00)

Session 2: Self Supervised Leaning; session chair: Takfarinas Medani

• 11:00–11:25– Dominique Duncan (University of Southern California):

• “Unsupervised Multivariate Time-Series Transformers for Seizure Identification on EEG”

• 11:25–11:50– Wenhui Cui (University of Southern California):

• “Neuro-GPT: A Foundation Model Pretrained on Large-Scale EEG Data”

Lunch Break (12:00–13:00)

Session 3: Machine Learning for brain computer interfaces; session chair: Shrikanth Narayanan

• 13:00–13:30- Alexander Silva (University of California, San Francisco):

• “A high performance neuroprosthesis for speech decoding and avatar control“

• 13:30–14:00– Maryam Shanechi (University of Southern California):

• “AI-powered next-generation neurotechnologies”

• 14:00–14:30– Ludovic Bellier '(University of California, Berkeley) :

• “Reconstructing Pink Floyd from human auditory cortex”

Panel Discussion

• 14:40–15:10– Moderator: Shrikanth Narayanan and Kristina Lerman (University of Southern California):

• “The Role of Foundational Models in Spontaneous and Event-Related EEG“

Coffee Break(15:10–15:40)

Session 4: Machine Learning for neurological disorders; session chair: Kristina Lerman

• 15:40–16:05– Srikantan Nagarajan '(University of California, San Francisco):

• “Machine learning algorithms for electromagnetic brain imaging in dementia”

• 16:05–16:30– Dimitrios Pantazis (Massachusetts Institute of Technology):

• “Graph representation learning of MEG signals opens a window to aging trajectories and Alzheimer’s disease”

• 16:30–16:55– Jason da Silva Castanheira (McGill University):

• “Inter-individual differences in neurophysiology vary with age and disease”

Closing and final remarks