Data Structures and Algorithms Flashcards

In speckle-tracking echocardiography, block-matching searches neighboring pixels to find the best-matching tissue pattern. What is the time complexity of an exhaustive block-matching search over a search window of n×n candidates?

Answer: O(n²)

Exhaustive block-matching evaluates every candidate position in the n×n search window, yielding O(n²) complexity.

Which sorting algorithm is most suitable for real-time reordering of a nearly-sorted list of echocardiographic measurement timestamps arriving in slightly out-of-order bursts?

Answer: Insertion sort

Insertion sort runs in near O(n) time on nearly-sorted data, making it ideal for re-sorting small, almost-ordered timestamp lists in real time.

A cardiac image archive indexes studies by patient ID and acquisition date. Which data structure supports efficient range queries such as 'retrieve all studies between two dates'?

Answer: Balanced binary search tree (e.g., B-tree)

Balanced BSTs and B-trees store keys in sorted order, enabling efficient O(log n) range queries by in-order traversal between bounds.

Which graph traversal algorithm would most efficiently determine whether two anatomical landmarks on a segmented cardiac mesh are connected?

Answer: Breadth-first search (BFS)

BFS efficiently answers connectivity queries by exploring all reachable nodes level by level from one landmark until the other is found.

When compressing a large batch of echocardiographic DICOM files for archival, lossless compression algorithms typically exploit which data property?

Answer: Redundancy and repeated patterns in the data

Lossless compression (e.g., RLE, Huffman) achieves size reduction by encoding repeated patterns and statistical redundancies in the data without discarding any information.

A recursive algorithm that computes the fractal dimension of a ventricular endocardial border doubles its input size at each step. If it takes T(n) time, what is its recurrence relation?

Answer: T(n) = 2T(n/2) + O(n)

Splitting the problem into 2 halves of size n/2 with O(n) work per level gives T(n) = 2T(n/2) + O(n), which resolves to O(n log n) by the master theorem.

In echocardiographic software, a Bloom filter might be used to quickly determine if a study ID has already been imported. What is a key characteristic of Bloom filters?

Answer: They allow false positives but never false negatives

Bloom filters are probabilistic: they may report an element as present when it is not (false positive), but they never report a present element as absent (no false negatives).