The Musculoskeletal Informatics Group (https://www.musculoskeletal.ai/) in the Department of Orthopedic Surgery at Boston Children’s Hospital and Harvard Medical School has an immediate opening for a motivated Research Assistant (RA) to work with a multidisciplinary team of biomedical engineers, orthopedic surgeons, and computer and data scientists to develop Natural Language Processing (NLP)-based tools for processing medical records. This role contributes to the development of centralized patient database and disease specific registries. This opportunity is primarily focused on translational research with immediate impact on clinical care of patients with a range of musculoskeletal conditions, through publication of novel findings and development of computational tools used by clinicians.

Principal Duties and Responsibilities

• Use machine learning approaches (e.g., natural language processing) to process clinical notes. Develop algorithms to pre-process the clinical data for development of machine learning pipelines. Develop machine learning pipelines to assist risk and outcomes prediction.
• Development of patient selection and text extraction pipelines, and development and fine tuning of negation and new entity recognition algorithms. Conduct literature searches and reviews, compiling information for journals and suggesting uses for new techniques to the P.I. Attends courses, lectures and conferences to further own knowledge of research and laboratory techniques, consulting with other researchers and technicians to exchange information.
• Assist in the process of data collection and analysis using varied instrumentation and software; provide proper documentation and organization to ensure integrity and organization of collected data.
• Apply basic statistical knowledge; understand, interpret, and describe complex data by preparing graphs and tables.
• Participate in the preparation, review, and submission of manuscripts and abstracts for publication.

Qualifications:

• Bachelor’s degree
• Experience in NLP feature engineering and modeling (e.g., text classification, entity recognition, entity extraction, question answering)
• Experience with SoTA modeling techniques (such as transformers, BERT, etc)
• Experience taking an NLP project from concept to production
• Building deep neural networks with modern tools, such as PyTorch or Tensorflow
• Expertise with Python
• One Journal publication or peer-reviewed conference paper or projects with public code repositories
• Research experience and practical knowledge of research principles is required whether through previous internship, work experience, or coursework. Experience with healthcare datasets is a plus
• Excellent critical thinking skills are required.
• Work requires the ability to communicate effectively both orally and in writing.