Diagnosing Fungal Infections By Artificial Intelligence

Professor Teng, Lee Lee   (Project Coordinator (PC))

Professor Yiu Siu Ming   (Co-Investigator)

Dr Schnieders Dirk   (Co-Investigator)

Professor Lau Susanna Kar Pui   (Co-Investigator)

Professor Woo Patrick Chiu Yat   (Co-Investigator)

Dr Tsang Chi Ching   (Co-Investigator)

36

2019-04-01

2022-03-31

3499997

Diagnosing Fungal Infections By Artificial Intelligence

Artificial Intelligence, Diagnosing, Fungal Infections

Others - Medicine, Dentistry and Health

MRP/026/18

Midstream Research Programme for Universities (MRP)

2018

Completed

The objectives of this proposal are to develop a novel method of fungal identification by artificial intelligence through the development of user-friendly databases and software for fungal identification by morphology recognition and sequence matching respectively. For morphology recognition, photos of typical morphologies of pathogenic fungi will be input to an image recognition software in order to ""teach the computer"". The ""computer that has learned"" will be able to recognize and identify new morphology image inputs of fungal isolates recovered from patients. For sequence matching, a software and database that contain the internal transcribed spacer region sequences of pathogenic fungi will be developed. The software will be able to match new sequence inputs of fungal isolates recovered from patients to sequences of pathogenic fungi in the database. Both software will be extremely user-friendly and the training time will be less than two hours. Identification of fungi will be rapid, timely and inexpensive. These developments will have tremendous impact on the diagnosis of fungal infections and hence will greatly benefit the elderly population who are the main sufferers of fungal infections. The project is highly feasible because we have the successful experience of developing a similar database and software for bacterial identification.