This document describes a relevance feedback algorithm inspired by quantum detection. It proposes using quantum probability and detection concepts to re-weight query terms and re-rank retrieved documents. The algorithm projects the query vector onto a subspace spanned by eigenvectors that maximize the distance between distributions of relevance and non-relevance probabilities. The system is designed to compute a new query vector using a linear combination of the original, relevant document, and non-relevant document vectors. Diagrams are included showing the system design using various Unified Modeling Language diagrams.