Cipher

Cipher

Machine Learning applied to the early detection of diseases.
Challenge

Challenge

The aim of this I+D draft report is to cretae a platform so that, by means of an Expert System and ML analysis algorithms and Big Data, can serve in a predictive way to detect diseases, especially ictus.

Andalusian Health Service

Andalusian Health Service

Created in 1986, is one of the most important Health Services in Europe, servicing more than8 millions of people.

SAS (Andalusian Health Service) has a support of 92.934 professionals (annual average) in order to assist its welfare network: 24.436 for primary medical attention and 68.498 for hospital care. Diraya is the system used for the Andalusian Public Health System as support for the electronic medical history.

Diraya is the system used for the Andalusian Public Health System as support for the electronic medical history.

Our solution

Data Collection Techniques
Data Collection Techniques
Information on Standards Techniques associated to medical ontologies such as SNOMED CT and linked analysis of the relationships coming from this connection.
Data mining and KDD application of standard processes applied to the construction of cerebrovascular diseases categorizers according to the HCE.
Machine learning Research
Machine learning Research
Pattern research findings in patients simptomatology and clinic data. Correlation analysis between risk factor and stroke case progress probability,as well as procedure and prescription effectiveness used for treatment. Farm of non-relational database. Semantic algorithm and self-learning detection of stroke probability.
Machine learning Challenge
Machine learning Challenge
Automatic learning scalability to medical pathology detection algorithm,based on theoretic models, tree diagrams and rules that are already verified in other existing medical models. Construction of functional model of pathology learning,such as genetic algorithm, artificial neuronal networks, gathering algorithms or Bayesian networks.

State of the arte enhancements

Results

Results

Creation of a expert system of sanitary software based on Big data, Machine learning and automatic learning Algorithms,capable of discovering hidden simptomatology in electronic records that are not considered or properly valuated nowadays.

This detection must be used to feed a Decision Support System ( DSS ) that warns the doctor when it detects a hazardous situation.

Technologies

Talend
Hadoop
MongoDB
Hortonworks
Spark
Solr
Elastic Search
Apache Hbase
Hive
Mahout
MapReduce

The project has been co-financed by the Ministry of Energy, Tourism and Digital Agenda.

Project number: EXP-00095962 / IDI-20161013
We want to help you achieve your digital objectives. Let's talk!
base pixel px
Convert
Enter PX px
or
Enter EM em
Result