Parisima Abdali Email & Phone Number

New York, Ny, Us

New York, Ny

New York, United States

1) Compressed Sensing methods for MRI Reconstruction on DCE-MRI dataset.- Developed Magnitude Subtraction Compress Sensing using Primal Dual Splitting Algorithm- Enhanced spatial and temporal resolution of reconstructed dynamic MR images comparing to traditional approaches. - Modified algorithm to reference based MS-CS and improved the reconstruction quality by 27.6%. 2) MR Contrast Image Synthesis using Deep Learning - Adapted Constrained Contrastive Learning for Feature Engineering and enhancing further synthesis task- Developed constraint maps using k-means clustering to guide a self-supervised learning model in feature extraction.- Fine-tuned an encoder-decoder Unet model on BraTS'21 Dataset to generate unavailable image T1CE using other available MR - contrast images.- Conducted statistical analysis using hypothesis testing on SSIM, PSNR, and LPIPS metrics (utilizing AlexNet and VGG) for a test cohort, demonstrating that the resulting p-values provide statistically significant evidence.- Implemented using Tensorflow and optimized computational processes, enabling efficient execution of large deep learning training jobs on distributed HPC systems.

New York, United States

SP’24 CS-GY-6923 Machine Learning Professor Gustavo Sandoval

New York, United States

CS - GY 6953 Deep Learning FA’23 - Professor Gustavo SandovalSP’24 - Professor Chinmay Hegde

Qazvin, Iran

Digital Signal Processing 1

Tehran Province, Iran

Worked on two projects in Deep Learning domain:- Training CNN model on TIMIT dataset for Multi-Speaker Detection up to 5 speakers. - Integrated YOLOv3 and MonoDepth on NYU Depth Dataset using PyTorch and TensorFlow, for social distancing estimation.

Qazvin Province, Iran

Digital Systems Design 1

Master Of Science - Ms, Electrical And Electronics Engineering

Thesis: MR Contrast Image Synthesis Using Deep Learning

Bachelor Of Science - Bs, Electrical And Electronics Engineering

Worked as Teacher Assistant for two courses Studies of Sound Source Localization using Machine Learning techniques for my bachelor’s.