Experience

• Leading a team of 6 developers in the areas of ML, DS, and CV

• Designing and implementing a specialized image processing system for Volastra Therapeutics, utilized for cell counting, classification, and detection

• Building a COVID-19/pneumonia classification model that utilizes guided attention for the Beth Israel Deaconess Medical Center

• Developing a COVID-19 scoring pipeline for the Beth Israel Deaconess Medical Center

Media: Article 1, Article 2, Article 3

• Analyze time-series data on blood flow index to derive actionable insights

• Develop advanced machine learning and deep learning models for noninvasive intracranial pressure forecasting

• Mentor PhD students in ML algorithm development for time series analysis

• Hyperspectral data processing and analysis

• Obtaining hyperspectral images while performing experimental surgery on animals

• Development of advanced ML models for hyperspectral detection and clustering

• Developing a model for obscuring faces and ears on 3D MRI/CT data for Biospective

• Creation of a contrast classification model using CT data for Bristol Myers Squibb

• Retraining and implementing the CLOVA OCR for medical text recognition

• Developing ML models for recognizing body parts, utilizing both MRI and CT data for Bristol Myers Squibb

• Implementing the CRAFT (Character-Region Awareness) text detector

• Supervising a group of 5 developers in the creation of CV algorithms and ML models

• Designing a specialized deep learning model for segmenting medical devices for Boston Children’s Hospital

• Building an automatic fire detection system utilizing machine learning techniques for the Incom Group

• Developing a Multi-Task Learning-based system for tracking aorta/catheter keypoints for Kemerovo Cardiology Center

Media: System for visualization of minimally invasive surgery (Article 1, Article 2)

• Development of a stenosis detection and scoring system

• Leverage Amazon Web Services for both model training and deployment

• Utilizing VAEs and GANs to generate new medical data

Media: Research statement, Reference letter by professor Alejandro F. Frangi

• Building a model to classify and detect pulmonary emphysema

• Developing an algorithm to generate semi-synthetic ultrasound data

• Implementing an MRI reconstruction technique using partial k-space data

• Utilizing unsupervised machine learning methods for biomedical image segmentation

Media: Internship certificate, Project evaluation, Reference letter by professor Maria J. Ledesma-Carbayo

• Utilization of feature engineering techniques for extracting features from raw data

• Development of a feature selection algorithm and its comparison with other methods

Media: Mobility certificate, Reference letter by professor Farid Melgani

• Implementation of process control systems within petrochemical pipelines

• Maintenance and repair of measurement devices, as well as automated process control systems such as Siemens S300/400, Yokogawa Centum, Numerik PS2000 and Remicont