Background
We started our journey to build a clinical decision support AI system for low back pain in 2018 . We recruited 16 Physios and GPs and undertook an elicitation exercise to find out:
We gathered a great deal of information from those participants. In order to break the task down in to manageable chunks, we focussed the later stages of elicitation and AI system building on serious spinal pathology. We are now revisiting the initial information that we collected from them. We need your input to turn the information we have into an AI system.
The AI system we want to use is called a Bayesian Network. The benefit of using a Bayesian Network is that they can be built from expert knowledge. The videos below introduce the project and give an overview of how a Bayesian Network can represent the clinical reasoning process.
- The most important factors to assess in low back pain
- What judgements clinicians make in low back pain
- What treatments are appropriate across the spectrum of low back pain
We gathered a great deal of information from those participants. In order to break the task down in to manageable chunks, we focussed the later stages of elicitation and AI system building on serious spinal pathology. We are now revisiting the initial information that we collected from them. We need your input to turn the information we have into an AI system.
The AI system we want to use is called a Bayesian Network. The benefit of using a Bayesian Network is that they can be built from expert knowledge. The videos below introduce the project and give an overview of how a Bayesian Network can represent the clinical reasoning process.
Project overview
|
Explanation of Bayesian Networks
|
Questionnaire
For the first part of this study, we are asking you to complete a questionnaire. The questionnaire responses and the follow up discussion will allow us to start building the Bayesian Network. For a more detailed explanation, please see the video below.
You can access the questionnaire here: https://qmul.onlinesurveys.ac.uk/back-pain-elicitation-stage-1v2
You can access the questionnaire here: https://qmul.onlinesurveys.ac.uk/back-pain-elicitation-stage-1v2
Relationships in the Bayesian Network
|
Tips for the questionnaire
|
Reference resources
Here are some of the resources created by the participants of our previous study. These should help you if you have any queries about the questionnaire contents. Here you can find definitions of the variables as agreed by our previous study participants. They provided a context question that might be asked of the patient to gain the information during a consultation. They also provided a unit or method of measurement, such as a scale or patient reported outcome measure, where appropriate.
Skip to:
Skip to:
Variable |
Question to the patient |
How is this measured? |