Shirin Vafaei

Hello! I’m currently a PhD student at Osaka University supervised by Takufumi Yanagisawa.

My goal is to develop accurate brain decoding and encoding algorithms, with the aspiration that my research can significantly benefit individuals facing challenges in communication or movement. To achieve this, I employ machine learning techniques to explore and understand the processes by which the human brain interprets information. This includes investigating the latent space and the underlying geometric representations of visual, auditory, and sensory information within the brain. My goal is to identify and refine algorithms that accurately represent and interpret these stimuli, thereby enhancing our ability to decode brain activity effectively.

Prior to my PhD, I completed my master’s degree at Kyoto University. During my master’s studies, I developed a method to reconstruct images from brain activity patterns.

Besides research work, I like to play volleyball, go cycling, hiking, and running and have finished running a marathon twice. I used to play double dutch with a team during my master’s studies at Kyoto University, and we had several performances at Kyoto University festivals. I sometimes play the violin during my free time, and once played in an orchestra.

news

Sep 18, 2024	We have released my first first-author preprint title brain-aligning of semantic vectors improves neural decoding of visual stimuli
Jan 21, 2024	My study was accepted to presented as poster at Cosyne 2024

selected publications

Arxive preprint

Brain-aligning of semantic vectors improves neural decoding of visual stimuli

Shirin Vafaei, Ryohei Fukuma, Takufumi Yanagisawa, and 10 more authors

2024

Abs HTML

The development of algorithms to accurately decode of neural information is a long-standing effort in the field of neuroscience. Brain decoding is typically employed by training machine learning models to map neural data onto a preestablished vector representation of stimulus features. These vectors are usually derived from image- and/or text-based feature spaces. Nonetheless, the intrinsic characteristics of these vectors might be fundamentally different than those encoded by the brain, limiting the ability of algorithms to accurately learn this mapping. To address this issue, here, we propose a representation learning framework, called brain-aligning of semantic vectors, that fine-tunes pretrained feature vectors to better align with the structure of neural representations of visual stimuli in the human brain. We trained this model with functional magnetic resonance imaging (fMRI) data representing 150 visual stimulus categories; then, we performed zero-shot brain decoding on 1) fMRI, 2) magnetoencephalography (MEG), and 3) electrocorticography (ECoG) data reflecting neural representations of visual stimuli. By using fMRI-based brain-aligned vectors, the zero-shot decoding accuracy all three neuroimaging datasets increased. This finding underscores the potential of leveraging a richer array of brainderived features to increase the performance of brain decoding algorithms.
The effect of publicly available non-invasive brain stimulation on attention functions in healthy adults: A systematic review and meta-analysis

Toru Takahashi, Shirin Vafaei, Ikko Kimura, and 2 more authors

2023

HTML