In-memory database system for medical data

Project implements a custom database system designed to manage medical records using advanced tree data structures. At its core, it features a generic Binary Search Tree (BST) and a self-balancing AVL Tree.

Binary Search Tree (BST)

Generic tree structure implementing standard operations.

• Supports Insert, Find, Delete, Level Order, Min, and Max.
• In Order and Find Interval operations are implemented using the Morris traversal algorithm (transition is based on creating temporary connections in the tree).
• Uses a nested Node<Key, Data> class that stores references to children, the parent, and a flag indicating whether it is a left or right child. The tree maintains a reference to the root and tracks the total node count.

AVL Tree

A self-balancing extension of the standard binary search tree.

• The class tracks node heights and includes methods to calculate the balance factor.
• Automatically executes single or double rotations (Left/Right rotation) during insertion and deletion operations if the tree becomes unbalanced.
• Inherits properties and methods directly from BSTree, overriding the node creation method to instantiate AVL-specific nodes.

Database

The database logic manages a complex system of 11 parallel trees to enable rapid querying of patients and their test results.

• PCRTest: An object containing the test date, patient ID, test ID, clinic ID, district ID, region ID, and result details (including a note).
• Patient: An object containing the name, surname, date of birth, and a unique patient ID.
• For advanced querying, the system utilizes nested AVL trees. For example, a structure like AVLTree<string, AVLTree<DateTime, AVLTree<int, PCRTest>>> allows the system to find a specific patient's tests, sorted chronologically by date and then by test ID.

Operations

• Inserting the PCR test result into the system
• Search for a test result for a patient and view all data
• List of all PCR tests performed for a given patient, sorted by date and time of their execution
• List of all / only positive positive tests performed during the specified time period for the specified district
• List of all / only positive tests performed for the specified time period for the specified region
• List of all / only positive tests performed for the specified time period
• List of sick people in the district / region as of the specified date, with the person considered sick n days after the positive test
• List of sick people in the district on the specified date, where the person is considered sick n days after the positive test, and sick people are arranged by test value
• List of sick people on the specified date, with the person considered sick n days after the positive test
• List of one sick person on the specified date, with the person considered sick n days after the positive test from each district with the highest test value
• List of districts / regions sorted by number of sick people on the given date, where a person is considered sick n days after a positive test
• List of all tests performed for a specified time period at a given workplace
• Search for a PCR test by its code
• Inserting patient into the database
• Permanent and irreversible deletion of the PCR test result
• Deleting a person from the system including their PCR test results

Testing

• Features an extensive test suite (Insert test, Dump test, Partial delete test, Find test, Min/Max test, and Find Interval test) to verify the correctness of individual operations and tree balance.
• Executes random operations based on set probabilities. Every 10 operations, it validates tree integrity, size, and balance against an auxiliary C# HashSet.
• Empirically compares the custom BST, the AVL tree, and the standard C# Dictionary across 10,000 operations, measuring execution time in microseconds.