- Which Laboratory Information System data interfaces do you support?
Standard HL7, Triple G Ultra, and Kestral PLS. A detailed Interface Specification is available on request.
- How do you charge for LabWizard™?
There are charges for initial installation and for the provision of customer support and training. The billing is then on a usage basis: either per patient episode or per interpretation made by each Knowledge Base.
- What types of pathology test results are suitable for interpretation by LabWizard?
Typically 80% of test panels ("test batteries") are amenable for LabWizard interpretation, depending on the specialties of the particular laboratory. An example of a pathology domain that is NOT readily amenable to LabWizard would be anatomical pathology where images or slides need to be examined by the Pathologist. Our customers currently use LabWizard in these 18 Pathology domains. In general, any Pathology domain where the numeric or textual test results are available electronically, is amenable to LabWizard interpretation.
- Who builds the rules?
Normally Pathologists or senior scientists.
- Do you have starter knowledge bases?
Yes. PKS can provide you with a useful starter knowledge base for most panels. The content can then be refined over time by your own staff.
- Will LabWizard™ replace pathologists or scientists?
No! Human experts are critical in the LabWizard process workflow. They provide all the expertise contained in the interpretations, and have the responsibility for signing off each interpretation before it is released. LabWizard simply enables them to work more effectively - providing a higher level of service to the primary care physicians than they would have been able to do otherwise.
- How long does it take to build a rule?
Instead of the industry average of 2-5 rules per day, LabWizard users can add 1 rule per minute.
LabWizard performance reports for the Sex Hormones knowledge base created and maintained by one of our customers.
Figure 1: The pathologists created a 217 rule knowledge base (y axis). Note that few refinements (new rules) were required after the first 6 weeks of operation.
Figure 2: Note that of the 35 week period shown, only 3 hours and 47 minutes (y axis) were spent on adding/modifying rules. This is an average time of 63 seconds spent on the addition of each rule.
- How large can the knowledge bases get before they are unmanageable?
The RippleDown Rule technololgy is designed to be scalable: in fact this is one of the key differentiators between RDR and other commercialised AI techniques.
LabWizard performance reports for the Lipid knowledge base created and maintained by one of our customers.
Figure 3: The pathologists created a 7,000 rule knowledge base (y axis) over the course of 30 months.
Figure 4: Note that the average time to add each rule is similar to that of the 217 rule knowledge base: 6673 rules / 100 hours is an average of 54 seconds to add/modify each rule.
- Why add interpretations to "normal" results?
Inclusion of interpretive comments is an opportunity to build a closer relationship with the client doctor:
- Why the results are "normal" can be clearly stated, for the benefit of the client doctors and patients
- Advice on how to keep the condition being investigated under control can be specified
- Appropriate and timely follow-up testing can be specified
- How can I prevent comments going to specialists?
A high degree of control over the content of the released interpretations can be defined by adding rules to the knowledge base to define the nature of the reports. In particular, rules can easily be added so that a different report (or no report) is sent to a Specialist, whereas the full report for that patient is sent to the primary care physician.
