Database Basic
- Relational databases (e.g. MySQL) are probably the most common. Essentially they store data in a set of 2D tables (relations) that follow certain rules of normalization and can be linked to each other via relational algebra (mostly set theory type functions).
- To address problems encountered when trying to interface a RDBMS with an object-oriented programming language, alternative non-relational database structures (e.g. XML, NoSQL, Hierarchical, Network) are available and may have advantages in certain situations.
- Tables (Relations/Relvars) are made up of Rows (Tuples/Records) and Columns (Attributes/Fields) where the intersection at a certain cell is called a value.
- Attributes contain values of the same data type (domain), e.g. integer, string, boolean.
- Each Tuple should have a primary key (normally an arbitrary integer) that identifies a unique tuple. Natural keys (those based on attributes) are possible as well but less optimized.
- Databases can be indexed to streamline queries.
Useful Wikipedia Pages
Working Prototype
Disease |
Risk Factors |
Associated Conditions |
Complications |
Symptoms |
Vitals |
Signs |
Labs |
Imaging |
Studies
|
Primary Key |
LIST of epidemiologic risk factors (age, sex, race, family history) |
problem LIST at time of presentation |
LIST of possible complications (screening and anticipatory guidance) |
LIST of symptoms (keep it broad: chest pain rather than substernal acute pressure-like chest pain with radiation) |
LIST of vital abnormalities |
LIST of pertinent positive/negative physical exam findings |
LIST relevant lab abnormalities |
LIST of useful imaging findings |
LIST relevant echo/ECG/PFT/sleep/etc studies
|
Base Table
Attribute |
Domain |
Cardinality
|
Disease (PK) |
text |
--
|
Risk factors |
text |
M2M
|
Associated Conditions |
text |
M2M
|
Complications |
text |
M2M
|
Symptoms |
text |
M2M
|
Vitals |
text |
M2M
|
Signs |
text |
M2M
|
Labs |
text |
M2M
|
Imaging |
text |
M2M
|
Studies |
text |
M2M
|
Each primary key (disease) can be associated with multiple values in each attribute. For example, CAD is associated with multiple risk factors (age, male gender, smoking, etc) and multiple symptoms (angina, dyspnea, nausea, etc). At the same time, each of these risk factors and symptoms are associated with other diseases besides CAD. For example, smoking is a risk factor for CAD, lung cancer, PVD, COPD, head and neck cancer, etc. In order to handle this many-to-many cardinality of the data, multiple base tables should be created and an associative table (junction table) can then be created as needed. The base tables in the above example would be:
Base Table 1
|
Disease (PK)
|
Risk factors
|
Disease < Symptom < Disease