Assistant Professor

• 2023 - Present

Postdoctoral Research Fellow

• Sep 2019 - May 2023

Department of Radiology, Harvard Medical School Department of Radiology, Harvard Medical School

Postdoctoral Research Fellow

• Sep 2019 - May 2023

The Martinos Center for Biomedical Imaging, Dept. of Radiology, Massachusetts General Hospital The Martinos Center for Biomedical Imaging, Dept. of Radiology, Massachusetts General Hospital

Co-Founder

• Mar 2021 - Jun 2022

Company history: Deer AI Medical -> Wecover -> Wecover Platforms - Deer AI Medical: We were initially a startup making deep learning-based medical imaging software (e.g., amniotic fluid measurement in fetal ultrasound, image enhancement for dental cone-beam CT, identification of unknown dental implants). - Wecover: Then, we pivoted to insurtech in the area of health insurance. - Wecover platforms: We did another pivoting within insurtech, but in the area of property and casualty, to bridge between insurance and reinsurance companies by developing a platform to automate various time-consuming tasks. The company also works on parametric insurance I had numerous rolls as a co-founder, but I primarily focused on R&D and project management. Show less

Research Collaboration

• Oct 2018 - Jul 2019

- Project: Prediction of patients’ status in anesthesia treatment with single-channel electroencephalogram (EEG) signal from multi-site using different hardware and drugs. Classification among four status: wakefulness, general anesthesia, burst suppression, and recovery of consciousness.- Data: Single-channel EEG signal collected from multiple hospitals with multiple devices.- Responsibility: Modeling of a neural networm that provides high prediction performance, Exploration of input/feature type (input to neural networks) – e.g., raw signal, permutation entropy, sample entropy, spectrogram, response power spectral density, etc.- Keywords: CNN, LSTM, Transfer Learning, Anesthesia, EEG Show less

Research Assistant

• Jan 2019 - Jun 2019

- Advisor: Prof. Anne Sereno- Project: Modeling of task representations in neural networks trained to perform tasks including object localization and detection. As the first step, we try to find differences in the mechanism of object localization and detection between monkeys and artificial neural networks (e.g., Faster RCNN, YOLO, and SSD).- Data: Action potentials recorded extracellularly while monkeys perform behavioral (attention) tasks using microelectrodes.- Responsibility: Network design and modeling, Data mining and analysis. Show less

Ph.D. Student

• Aug 2013 - May 2019

- Medical/Neuro Imaging (DWI, fMRI, EEG, NIRS) & Image Processing- Machine Learning: Deep learning, Transfer learning, Ensemble/Boosting, Support vector machine, Regression analysis, Bayesian classification, Principal component analysis, KNN.- Thesis Topics:1) Predictive modeling of mild traumatic brain injury for collision-sport athletes and patient classification using machine learning.2) Automated quality control/assessment of 3D/4D medical images (e.g., diffusion MR images, T1-weighted images) with convolutional neural networks and transfer learning.- Other Projects:1) Modeling of diffusion tensor (or diffusivity) using resting-state fMRI.2) Investigation of cerebrovascular reactivity alterations due to subconcussive repetitive head trauma in asymptomatic football players using fMRI.3) Designing and optimization of parallel transmit and receive SENSE MRI. Show less

Lab Instructor (TA) – Digital Signal Processing

• Aug 2014 - Dec 2018

- Course: ECE 438 (Digital Signal Processing)- Topics: Frequency analysis, Sampling and reconstruction, Interpolation and decimation, DFT and FFT, Digital filter design, Speech recognition and synthesis, Spectrogram, Image processing, Halftoning, Random processes, Waveform quantization, etc.- Responsibility: Instructing lab sessions, Lab development, Syllabus & Rubric development, Holding office hours, Grading reports, quizzes, and codes.- Worked with Profs. Jan Allebach, Mireille Boutin, and Okan Ersoy Show less

Full-time Instructor – Signals and Systems

• Jun 2015 - Aug 2015

- Course: ECE 301 (Signals and Systems)- Topics: LTI system, Fourier series, Fourier transform, Sampling theory, Z-Transform, etc- Responsibilities: Coordinating course with TA and grader, Lecturing, Development of exams, homework, and quizzes, and Proctoring and grading exams, and Holding office hours.

Data Analyst

• Apr 2014 - Nov 2014

- Advisor: Dr. Bridget Walsh- Project: Characterization and identification of brain signals of stuttering children compared to healthy controls using functional near-infrared spectroscopy (fNIRS).- Responsibility: Signal processing and analysis of the collected fNIRS data.

Deep Learning SW Intern

• May 2018 - Aug 2018

- Team: Autonomous Driving R&D - Projects: 1) Generate physically possible randomized trajectories based on recorded data/trajectories then viewport transform the recorded videos accordingly. This data augmentation allows the self-driving car to cope with situations that have never/rarely been seen. 2) Extend the current training architecture to add feedback connections using LSTM to memorize circumstances/features that may be important at an unknown time in the future and cope with unexpected situations. 3) Re-implement and improve the neural network’s data prep & training infrastructure to make it ready for rapid iterations with a massive amount of data. - Other responsibilities: Algorithm enhancement, Data preprocessing, Interfacing with sensors, SW development (continuous integration, writing tests, code review, etc.) - Language: C++ (90%), Python/Tensorflow (10%) Show less

Research Intern (Startup company)

• Jun 2016 - Aug 2016

- Team: Custom Products R&D - Goal: Analysis of hearing aids' photodetector and light tip emitter alignment in the ear when patients talk, smile, or yawn, to estimate and optimize the performance and power efficiency of the system. - Responsibilities: 1) Computer Vision & Image Processing. e.g., edge and feature detection, point tracking, line fitting, rigid-body tracking, camera calibration, un-distortion using intrinsic parameters. 2) 3-D Geometry Calculation. e.g., estimation of camera motion (translation and rotation) from consecutive 2-D images taken by a camera. 3) Data Collection & Analysis. e.g., statistical analysis of experimental results. 4) Algorithm Automation. e.g., Matlab, Python, Grasshopper (3-D modeling/analysis tool). Show less

Research Intern

• Feb 2013 - Jul 2013

Research Intern --- advisor: Prof. Dong-hyun Kim - Research on enhanced image reconstruction of electric conductivity from MRI - Research on total variation de-noising on MR images Research Intern --- advisor: Prof. Dong-hyun Kim - Research on enhanced image reconstruction of electric conductivity from MRI - Research on total variation de-noising on MR images