Maral Kasiri, Ph.D. Email & Phone Number

Los Angeles, Ca, Us

Salt Lake City, Utah, Us

Cambridge, Ma, Us

Cambridge, Ma, Us

Irvine, Ca, Us

Abstracts:-Thalamic and pallidal LFP power correlates with voluntary movement in children with dystonia., Society for Neuroscience, San Diego, California. 2022-Low frequency thalamic and pallidal local field potential can predict voluntary movement in children with dystonia.,International Conference on Neurology and Brain Disorders, Orlando, Florida. 2022-Investigation of low frequency thalamic and pallidal LFP correlation with voluntary movement in children with dystonia.,UC Systemwide Symposium on Bioengineering and Biotechnology Industry Showcase, Santa Barbara; AGS symposium, Irvine, 2nd Place Winner. 2022-Pallidothalamic Signal Transmission Model in Human Brain: A pilot study through deep brain stimulation., International Brain Stimulation Conference, Charleston, South Carolina. 2021-Dynamic causal modeling predicts the basal ganglia- thalamus loop neurotransmission in children with secondary dystonia undergoing deep brain stimulation., Society for Neuroscience, Chicago, Illinois. 2021

Orange, California, Us

Chicago, Il, Us

- Established and assessed PostgreSQL + PGVector vector database for storing ProT-VAE+BioNeMo LLM (Variational auto-encoder built on NVIDIA BioNeMo) protein sequence embedding vectors + Protein embedding dimensionality reduction and quantization analysis for optimized storage requirements- Assessed the protein embeddings efficiency in protein sequence representation via amino acid contact prediction using XGboost, linear regression, and a dilated residual neural network model - Evaluated its prediction power by comparing the embeddings with ESM embeddings (Meta protein language model) and one-hot encoded representation of proteins (positive and negative control).- Kept detailed and precise records to track project progress and outcomes

Los Angeles, Ca, Us

Abstracts:Artifact removal in rodent electroencephalogram for accurate high frequency oscillation detection, Kasiri, M,. et al. (American Epilepsy Society 2020)

Long Beach, Ca, Us

Teaching Associate (BME 211- Biomechanics), California State University, Long Beach 2018• Teaching the course and leading a teaching assistant and a grader to deliver the graded assignments and respond to students intime and holding office hours• Preparing and developing the course syllabus based on the references and textbooks• Designing the course project and assignments and motivating students by using active learning techniques

Long Beach, Ca, Us

Human Performance and Robotics Lab at CSULB: "EMG signal classification: Hand gesture prediction using CNN and gradient boosting":I classified 8 different hand gestures using the EMG recordings from 8 muscles of 32 male with amputated arms using two different methods of learning. I showed that gradient boosting yielded higher testing accuracy (92%) for long periods of recordings compared to a CNN model; however for the shorter segments, a 2 layer convolutional network classified the hand gestures better, with 94% accuracy on the testing set. The models were tested with both the raw EMG signals and the wavelet features extracted from EMG. There was no significant difference between the two feature sets. Tools: Python (TensorFlow, Keras).

Los Angeles, Ca, Us

-Collaborated in NIH funded project to create a synthetic neuromorphic spinal cord capable of controlling tendon-driven, anatomical and robotic limbs-Analysis and visualization of the Neuromorphic System Data by Matlab-Project: "Muscle models and their differences in application": I implemented three different computational muscle models in MATLAB and and compared their responses to different perturbations. The virtual computational muscle model, introduced by Gerald Loeb, was the best to simulate the response of a muscle to an external force compared to other lumped models (Hill type muscle model). Tools: MATLAB-Implementing a User Interface using JavaScript and HTML to display the results and plots -Simulation of a single muscle's contractions in different muscle models using MATLAB-Simulation of sensory feedback (PID), GTO, and muscle spindle in Virtual Muscle to provide position control to the muscle using MATLAB

Los Angeles, Ca, Us

• Math245 (Differential Equations )

• Regulating and calibrating medical devices and systems based on ISO 13485/ IEC 62363• Guiding and preparing documents as part of the hospital workflow• Communicating the issues related to calibrations through progress reports on a regular basis

Doctor Of Philosophy - Phd, Biomedical Engineering/ Neuroscience

Doctor Of Philosophy - Phd, Biomedical/Medical Engineering

Master’S Degree, Biomedical/Medical Engineering

Bachelor'S Degree, Biomedical/Medical Engineering

High School Diploma