A measurement concept defining how outcomes are modeled and how state descriptions are updated after an outcome is recorded. It governs outcome probabilities, information extraction, and the disturbance introduced by the measurement interaction. It does not yield reliable inference without adequate calibration, sufficient data, and appropriate estimation procedures. It supports reconstruction and validation of state and process descriptions from experimental statistics. The concept is generally stable, though practical implementations and estimation methods evolve over time.

Principle

Tomography exploits the linear relation between probabilities and the underlying operator representation (Born rule). Informationally complete measurement sets and well-posed inversion or constrained estimation (positivity, trace constraints) are required so that observed statistics determine the unknown operator within experimental uncertainty.

Demonstration

State tomography of a single qubit: perform measurements in three Pauli bases (X,Y,Z), collect counts for each outcome, and use linear inversion or maximum-likelihood estimation to produce a best-fit density matrix ρ that is positive semidefinite and trace-one. Process tomography example: prepare a basis of input states, apply the channel, perform state tomography on outputs, then reconstruct the channel’s Choi matrix.

Misapplication

Performing tomography with an informationally incomplete measurement set, neglecting positivity and trace constraints (yielding unphysical density matrices), or ignoring state-preparation-and-measurement (SPAM) errors and finite-sample uncertainty. Treating the tomographic estimate as exact rather than as an estimator with uncertainty is another common error.

Consequence

When correctly implemented, tomography yields operational characterizations used for device calibration, error diagnostics, and theoretical model testing. It also quantifies uncertainties and reveals nonideal behavior, enabling improved control and validation of quantum hardware.

Reversal

Reversing the idea — viewing tomography as a single-shot, definitive readout rather than as a statistical estimation problem — eliminates the role of sampling error and model assumptions and overstates confidence in the reconstructed object.

Boundary

Scope: statistical estimation procedures to reconstruct quantum states, channels, or measurement operators from experimental data. Excludes single-shot projective readouts, direct fidelity estimation protocols that avoid full reconstruction, and device characterization methods that do not produce an operator estimate (e.g., randomized benchmarking summaries).

Semantic Tension

Tension exists between full tomographic reconstruction (complete operator estimation) and targeted characterization methods (fidelity estimation, randomized protocols, compressed sensing). There is also a tension between model-based parametric fits and model-free linear inversion.

Synthesis

Quantum tomography is the set of experimental designs and statistical inversion techniques that, using the Born-rule relation between operators and outcome probabilities, reconstruct approximate representations of quantum states, processes, or measurements from repeated, informationally designed experiments while accounting for physical constraints and statistical uncertainty.