BioMiner Features

The BioMiner assists in the surveillance of emerging influenza outbreaks for the world-wide network of Navy and Marine Corps medical facilities. Its surveillance system is based on data currently received by the EPI Data Center, including:

  • Lab Results
  • Pharmacy Transaction

The BioMiner identifies patients that have a high probability of influenza prior to their being tested. It also provides an alert based system used to notify personnel to the existence of these patients and provide any observed evidence regarding their current situation. It also collects statistics on the accuracy of the surveillance model being currently implemented into the system.


BioMiner conducts surveillance, collects performance statistics, and generates alerts with a minimum of human intervention. Replacing the often labor-intensive manual tasks of biosurveillance with automated processes saves time and money and allows you to concentrate on patient care.

BN Archive

All trained Bayesian Networks are stored in the BN Archive. Users can review the library of networks created with the software and can track their performance through time.

Data Statistics

The BioMiner provides a quick overview of the data in your database. Get statistics on the quantity and quality of your data, and use this information to make informed decisions. Information about individual patient cases is available at your fingertips. Advanced feature selection algorithms provide feedback on the quality of data features and which features to use in your Bayesian Networks.

Easy To Use

The BioMiner is designed to be simple to operate and to be a time save for busy medical professionals. It features easy-to-use wizards to perform various functions throughout the software. This streamlines user operations and creates a consistent user interface for performing tasks.


After training is complete, the resulting Bayesian Networks can be selected to be part of a Surveillance Model Profile. Evaluation lets you examine how these models perform before selecting them for live patient surveillance. The evaluation process produces a series of graphs and statistics to help you make an informed decision.

Featured Selection

The BioMiner employs advanced algorithms for performing feature selection. By analyzing training data, the BioMiner can determine which features are the greatest differentiators for your data set, thereby providing you with a Bayesian Network producing the most useful results with minimal noise.


Comprehensive configuration management allows BioMiner to quickly adapt to changing environments. Support for multiple data sources allow you to take advantage of all your aggregated data together in one place.

Job Engine

All CPU-intensive operations, such as training or evaluation, are submitted to a central job engine. This allows users to queue multiple jobs thereby freeing them to perform other duties. All job results and output documents are associated with the job and are easily found through the user interface when the job completes.


Surveillance examines patient records on a case-by-case basis. If a patient's case is deemed suspicious by the BioMiner, users can be contacted using various external messaging mechanisms. Through e-mail or desktop alerts, medical professionals can be notified with all the relevant data allowing them to make an informed decision.


Training the BioMiner allows it to learn the characteristics of your data and to find interrelationships between features. Training is made simple with training wizards, and made flexible with multiple training profiles. All experience learned from training is captured in a Bayesian Network which can then be used to perform surveillance.


Continual surveillance of patients on a case-by-case basis and support for external messaging ensures that no cases slip through the cracks. Upon detection of anomalous cases, users are notified directly on their desktops with the relevant data they need to make informed decisions and take decisive action.