Supervise students for Professional Experience projects as industry advisor.

Algorithmic mapping between health classifications - ICD-10, ICD-10-AM, ACHI, ICHI. Specializing in Natural Language Processing (NLP) and machine learning.

AR-DRG Classification System Development & Refinement Services at ACCD Lidcombe. My work includes algorithmic medical codes mapping between diseases/procedure classifications (ICD, ACHI, ICHI). This is achieved by text pre-processing, normalisation and tokenisation using Spacy with the help of cTakes for POS and dependency parsing. Elasticsearch technique is used to find the closest description in order to map between ICD-10 and ICD-10-AM. The same technique is applied for ACHI to ICHI classifications. As there are no mappings exist to verify accuracy, a web application(vue.js) is created as an easy-to-use interface for coder specialists to input mappings manually. These maps are then used as training data to refine machine learning model (Word2vec). Sentence embedding models like BERT is also used. These model are utilised to increase accuracy of mappings for the rest of the classifications.