TAO Toolkit

AWS Sagemaker Hosting

• Cleaned 30,000 rows of breast cancer imaging data in SQL database

• Built React web application for searching of XML files containing breast cancer lesion data

• Used Tesseract OCR to read clinical case report PDFs, cleaned text with Markov chain methodology combined with a TextRank algorithm

• Used processed text to train a TensorFlow model to classify case reports as 'Benign' or 'Malignant'