There was a fascinating presentation given by Dr Albert Gatt, on 'Automatic Summarisation of Neonatal Intensive Care Data' to The Malta Group of Professional Engineering Institutions on 10th December 2009.
Dr. Gatt is a Lecturer at the Institute of Linguistics, University of Malta, and a Research Fellow at the Department of Computing Science, University of Aberdeen. His research focuses on various aspects of language and its relation to cognition and computation.
In the drive towards achieving better patient care, contemporary Neonatal Intensive Care Units (NICUs) collect very large volumes of data about patients, both through sensors (such as ECG and blood pressure monitors) and via manual. The data is essential for effective clinical decision-making by medical staff. However, the sheer volume of information makes it difficult for staff to assimilate what is important and utilise the information efficiently. Current technologies for information presentation, which typically rely on visualisation (such as time series plots), have been shown to suffer from a number of limitations.
The so-called 'BabyTalk' Project, is a collaboration between the Universities of Aberdeen and Edinburgh, and the Royal Infirmary of Edinburgh (NHS Lothian, is developing technology for the automatic summarisation of patient data, using techniques from Artificial Intelligence and Natural Language Generation (NLG). The aim is to process the raw data intelligently, producing a coherent summary in English of the clinically important events that occurred in relation to a patient over a specific period.
This talk described the challenges involved in developing such technology, with particular reference to two NLG systems and their evaluation with clinicians, both on-ward and off-ward.
Joe Camilleri, Assistant Honourary Secretary
Malta Group of Professional Engineering Institutions