MENU
BMI
Multiple Organ System Failure
SOB on
Myocardial Infarction
Angina
Arrhythmia
Cardiac Stents Within Last Yr
Shocked
TIA or CVA
Sepsis
DIC
Major Trauma
Leaking AAA
Ischaemic Bowel (CT confirmed)
Diabetes Mellitus
Pregnant
Valvular HD
LV Ejection Fraction
Right Heart Failure
Implanted Cardiac Device
Widespread Malignancy
Seizures
Hepatic Encephalopathy
Obstructive Sleep Apnoea

ACS     1.0     Reset

SYSTEM IN DEVELOPMENT.
DO NOT USE FOR CLINICAL DECISION-MAKING.


The Anaesthetic Comorbidity Score (ACS) is a high-resolution, modern equivalent to the ASA Classification which has been an international standard for over 60 years. The 24 clinical inputs are processed in real time using an Artificial Neural Network trained by Consultant Anaesthetists at the QEUH in Glasgow.

The ACS ranges from 1 (fit and well) to 5 (moribund) and is displayed at a resolution of 0.1

Why use a Neural Network? There are 13,824 possible combinations of comorbidities in the ACS system. While it would be theoretically possible to write a large number of complex formulae to provide a realistic score for each possible combination, a neural solution combines accuracy and generalisation in an organic way. The Neural Network implemented here grows increasingly more accurate by continuing to learn from inputs by experienced anaesthetists. Neural networks of the type used here are known to be able to duplicate any mathematical formula no matter how subtle and complex. An Artificial Neural Network solution has a clear advantage over a traditional scoring system when the task involves a complex and subtle interplay of clinical factors.




Real-time Training Access Code