This work is being done in collaboration with Pakistan Institute of Medical Sciences (PIMS), Islamabad under the supervision of Dr Lubna Naseem. Congenital malformations (CM) are abnormalities of structures arising during the prenatal development and hampering body.
CM is one of the most important causes of infant mortality in the developing countries. In Pakistan 6-9% of the perinatal deaths are attributed to CM, but a comprehensive nation-wide data on the prevalence, nature and dynamics of CM are largely missing. Hence, the aim of present study is to forecast the prevalence of CM in the multiethnic and multilinguistic population of Rawalpindi/Islamabad through an inference engine. Inference engine helps in formulating new conclusions about the data that is provided to the inference engine and stored in the knowledge base of the inference engine.
This pilot engine presents a comprehensive overview of neonatal and maternal parameters and highlights the potential risk factors associated with CM by formulating new conclusions. Additionally, this inference engine would be helpful to establish the dynamics of CM in our society. It is anticipated that such project conducted on a country-wide sample could be highly beneficial in guiding our national health policy, resource allocation and management of CM
User interface is also required to translate the input suitable for engine to interpret and apply on set of production rules. Inference engine is the heart of the expert-system. It tracks the ordering of rules in which they are allowed to fire.
Expert System is developed as a C# .net desktop application and using the local database of visual studio for storing the production rules.
The GUI of the prototype is shown in Figure 1:
Figure 1: Prototype of Expert System
The Inference Engine was tested for the validity of its prediction using a testing dataset that is obtained from the PIMS Hospital.
Figure 2: Working of the system