Medical data can come from heterogenous sources and devices. For example, electronic health records, labs, MRI and EKG. The data can be represented by textual,
numeric or nominal features but it can also be described by continuous graphs and hyperspectral images. In addition, the medical data
is usually high-dimensional and noisy. Hence, data mining in medical data involves sophisticated fusion of the different sources, intelligent pre-processing,
normalization and smart feature selection.
Brainstorm Private Consulting has a range of offerings in mining medical data.
In addition, we provide skills in data mining and knowledge discovery, clustering and classification
(for example, to automatically diagnose patients) and
anomaly detection (for example, early detection of anomalous cells).
We can provide a deep medical data analysis capability blending and fusing a range of innovative techniques and tools to provide
high quality intelligence led solution.
EHR and blood tests analysis
Hospitals and medical institutions maintain large databases of electronic health records and lab tests results. Every record and test
is analyzed manually by a physician or a specialist, based on their experience and knowledge. A record is a collections of textual, numeric and nominal
values and usually its dimension is very high. Hence, manual analysis of medical records is, on one hand, an exhaustive task, but on the other hand,
prone to mis-detection of health problems.
In this case study, we will survey the solutions that we provided to a medical institution, to supply a computer-based diagnosis of patients based on their blood tests.
Psychological questionnaire analysis
Psychologists are using psychological questionnaires to evaluate their patients.
Each questionnaire consists of hundreds of yes/no questions.
The filled questionnaires are processed manually by experienced psychologists
and the patients are diagnosed and classified accordingly.
In this case study, we will survey the solutions that we provided to a medical institution, to supply a computer-based
diagnosis and classification of patients based on their psychological questionnaires.