Quantum Theranostics Overview

• Quantum theranostics (QTX) is an emerging field that integrates quantum-enabled computation, sensing, and nanomaterials to co-design diagnostics and therapies.
• It leverages technologies like quantum computing, nuclear imaging, and NV-diamond biosensing to enhance precision medicine and individualized treatment design.
• QTX applications span from computational genomics and radiotherapy-integrated imaging to quantum biosensing, while addressing challenges in biocompatibility and system verification.

Searching arXiv for papers on quantum theranostics and related quantum medicine frameworks. Quantum theranostics (QTX), as used across the cited literature, denotes the integration of diagnostics and therapeutics by means of quantum-enabled computation, quantum sensing, quantum imaging, or quantum nanomaterials. In this usage, QTX is not restricted to a single modality. It includes quantum computing and quantum machine learning for precision medicine and drug discovery, cell-centric therapeutic design, radionuclide imaging schemes such as $\beta^+\gamma$ PET, radiotherapy-coupled entangled-photon imaging, nitrogen-vacancy diamond biosensing, and nanosystems such as spasers and quantum dots (Bertl et al., 25 Feb 2025, Maniscalco et al., 2022, Basu et al., 2023, Matulewicz, 2021, Chi-Durán et al., 15 Aug 2025, Galanzha et al., 2015). The unifying objective is theranostic coupling: diagnostic inference and therapeutic intervention are treated as parts of a single pipeline, with quantum methods contributing speed, expressivity, sensing contrast, or multimodal integration.

1. Conceptual scope and domain structure

In precision-medicine papers, QTX is framed as an extension of theranostics into quantum-native computational and sensing regimes. One strand emphasizes quantum computing (QC) and quantum machine learning (QML) for faster genomic analysis, molecular simulation, disease prediction, and individualized treatment design (Bertl et al., 25 Feb 2025). A second strand places QTX within a multiscale “quantum network medicine” program in which disease modules, protein structures, and drug–target quantum chemistry are analyzed in a feedback loop (Maniscalco et al., 2022). A third strand shifts the emphasis from computation to instrumentation, using nuclear, spin, optical, or plasmonic physics to combine imaging and therapy in a single platform (Matulewicz, 2021, Chi-Durán et al., 15 Aug 2025, Galanzha et al., 2015).

QTX stratum	Representative mechanism	Representative source
Precision medicine	QC, QML, formal methods, molecular simulation	(Bertl et al., 25 Feb 2025)
Network- and cell-centric therapeutics	quantum walks, VQE, QCNNs, quantum OT, QTDA	(Maniscalco et al., 2022, Basu et al., 2023)
Nuclear imaging theranostics	$\beta^+\gamma$ PET with prompt-$\gamma$ radionuclides	(Matulewicz, 2021)
Radiotherapy-integrated quantum imaging	MV-beam-generated entangled 511 keV photon pairs	(Olivera et al., 4 Sep 2025)
Quantum biosensing	NV-center spin relaxometry on a multiplexed diamond microarray	(Chi-Durán et al., 15 Aug 2025)
Quantum nanomaterials	spasers and quantum dots for imaging, sensing, and therapy	(Galanzha et al., 2015, Biswas et al., 21 May 2025)

A common misconception is that QTX denotes only an imaging modality. Across the cited work, it instead functions as an umbrella term for architectures in which diagnosis and treatment are co-designed, sometimes computationally, sometimes instrumentally, and often across multiple biological scales.

2. Computational QTX in precision medicine and systems biology

The computational literature treats QC as advantageous for biomedical problems with superpolynomial complexity, highlighting superposition, entanglement, and tunneling as the operative physical resources. In the precision-medicine survey, Grover’s algorithm is presented as a canonical example for accelerating unstructured search, with

$$T_{classical}(W)= t \frac{N}{2}, \qquad T_{quantum}(W)= t \sqrt{N},$$

and the same source classifies QML into four paradigms: QC, CQ, QQ, and CC (Bertl et al., 25 Feb 2025). Within this framework, QSVMs are described as using Grover’s algorithm for faster hyperplane discovery, while quantum annealing and QPU topologies such as D-Wave’s Chimera graph are discussed in relation to deep-learning-style model encoding (Bertl et al., 25 Feb 2025).

For drug discovery, the dominant computational motif is quantum chemistry. Variational Quantum Eigensolver (VQE) and Quantum Phase Estimation (QPE) are identified as routes to solving the electronic Schrödinger equation for drug–target interactions,

$\hat{H}|\Psi\rangle = E|\Psi\rangle,$

with protein folding and virtual screening further linked to quantum Monte Carlo, quantum annealing, and Grover-style search (Bertl et al., 25 Feb 2025). In the network-medicine formulation, these quantum-chemical calculations are embedded in a multiscale workflow: disease-module identification, structural modeling, quantum simulation of prioritized drug–target pairs, and iterative refinement of the network model (Maniscalco et al., 2022). The same source highlights continuous-time quantum walks for interactome analysis, with classical diffusion summarized as $\sigma^2 \propto t$ and quantum diffusion as $\sigma^2 \propto t^2$, and gives the transition probability

$$P_{i \rightarrow j}(t) = |\langle j|e^{-i \hat{H} t}|i\rangle|^2.$$

Cell-centric therapeutics extends this computational program from molecules to tissues and perturbational dynamics. Four domains are emphasized: cell engineering, tissue modeling, perturbation modeling, and bio-topology. QCNNs and QNNs are proposed for combinatorial cell-design tasks such as CAR T cell engineering; hybrid classical–quantum GNN pipelines are proposed for spatial omics and tumor microenvironment cell-graphs; quantum conditional optimal transport is introduced for modeling drug- or patient-specific state transitions; and quantum topological data analysis (QTDA) is proposed for persistent homology, Betti numbers, and higher-order interaction discovery (Basu et al., 2023). In the same literature, Burch et al. are cited for quantum tensor decomposition in multi-omics integration, and quantum cumulant computation is linked to the cumulant generating function $K(t)=\log \mathbb{E}[e^{tX}]$ (Bertl et al., 25 Feb 2025, Basu et al., 2023).

A concrete gene-regulatory example is given by “Alz-QNet,” a quantum regression network for Alzheimer’s disease theranostics (Konar et al., 6 Aug 2025). Each gene is represented as a qubit; an encoder layer initializes qubits via $R_Y(\theta_{k,k})$ with $\beta^+\gamma$0; regulation layers use controlled $\beta^+\gamma$1 gates; and the symmetry $\beta^+\gamma$2 reduces the entangling gate count from $\beta^+\gamma$3 to $\beta^+\gamma$4 (Konar et al., 6 Aug 2025). The implementation described there uses $\beta^+\gamma$5 genes from the $\beta^+\gamma$6 entorhinal cortex dataset and a KL-divergence-based objective to recover a weighted, directed regulatory network with edges such as YY1 repressing PLD3 and PLD3 repressing APP and SREBF2 (Konar et al., 6 Aug 2025). This suggests that QTX can operate not only at the level of drug candidates or images but also at the level of disease-specific regulatory inference.

3. Imaging-centered QTX: from $\beta^+\gamma$7 PET to MV-beam entangled-photon platforms

In nuclear-medicine usage, QTX includes $\beta^+\gamma$8 PET, in which a radionuclide emits a positron and a prompt additional $\beta^+\gamma$9-ray. Standard PET reconstructs a line of response (LoR) from the two 511 keV annihilation photons. In the $\gamma$0 scheme, the third $\gamma$1 yields a Compton cone, and the intersection of that cone with the LoR enables event-by-event localization of the decay vertex (Matulewicz, 2021). The generic decay sequence is

$\gamma$2

$\gamma$3

followed by

$\gamma$4

The paper identifies $\gamma$5Sc, $\gamma$6Mn, and $\gamma$7Cu as the three leading radionuclides on the basis of the product of $\gamma$8 branching and prompt-$\gamma$9 emission, while also noting $$T_{classical}(W)= t \frac{N}{2}, \qquad T_{quantum}(W)= t \sqrt{N},$$0V, $$T_{classical}(W)= t \frac{N}{2}, \qquad T_{quantum}(W)= t \sqrt{N},$$1Co, $$T_{classical}(W)= t \frac{N}{2}, \qquad T_{quantum}(W)= t \sqrt{N},$$2Ga, $$T_{classical}(W)= t \frac{N}{2}, \qquad T_{quantum}(W)= t \sqrt{N},$$3Tc, and $$T_{classical}(W)= t \frac{N}{2}, \qquad T_{quantum}(W)= t \sqrt{N},$$4I as other options (Matulewicz, 2021). The theranostic rationale is explicit: some isotopes are suitable for PET and also useful for therapy, with $$T_{classical}(W)= t \frac{N}{2}, \qquad T_{quantum}(W)= t \sqrt{N},$$5Sc/$$T_{classical}(W)= t \frac{N}{2}, \qquad T_{quantum}(W)= t \sqrt{N},$$6Sc and $$T_{classical}(W)= t \frac{N}{2}, \qquad T_{quantum}(W)= t \sqrt{N},$$7Cu/$$T_{classical}(W)= t \frac{N}{2}, \qquad T_{quantum}(W)= t \sqrt{N},$$8Cu offered as diagnostic–therapeutic pairings (Matulewicz, 2021).

A distinct high-energy formulation appears in the work on clinical megavoltage radiotherapy beams as entangled-photon sources (Olivera et al., 4 Sep 2025). There, GEANT4 simulations of water-equivalent phantoms with 2 cm spherical tumors loaded with gold nanoparticles (AuNPs, 10 mg/mL) show that 6, 10, and 15 MV beams can generate 511 keV photon pairs with yields reaching about $$T_{classical}(W)= t \frac{N}{2}, \qquad T_{quantum}(W)= t \sqrt{N},$$9 pairs per Gy per cm$\hat{H}|\Psi\rangle = E|\Psi\rangle,$0 in AuNP-loaded tumors, with a more specific value of $\hat{H}|\Psi\rangle = E|\Psi\rangle,$1 pairs / Gy / cm$\hat{H}|\Psi\rangle = E|\Psi\rangle,$2 listed in the summary table (Olivera et al., 4 Sep 2025). A typical 2 Gy fraction produces $\hat{H}|\Psi\rangle = E|\Psi\rangle,$3 entangled pairs per 1 cm$\hat{H}|\Psi\rangle = E|\Psi\rangle,$4 tumor, and the depth-dependent voxel SNRs are reported as 158 at 5 cm, 119 at 10 cm, and 86 at 15 cm (Olivera et al., 4 Sep 2025). Correlation fidelities are given as $\hat{H}|\Psi\rangle = E|\Psi\rangle,$5, $\hat{H}|\Psi\rangle = E|\Psi\rangle,$6, and $\hat{H}|\Psi\rangle = E|\Psi\rangle,$7 at the same depths.

The same study assigns different biomarker roles to temporal and spectral observables. Positronium lifetime shifts of about 100 ps are linked to oxygenation and reactive oxygen species; Doppler broadening of $\hat{H}|\Psi\rangle = E|\Psi\rangle,$8–3 keV is linked to electron density and tissue structure; AuNP-induced line broadening of 0.5–1 keV marks nanoparticle uptake; and Compton shifts between 13–171 keV are associated with tissue composition and ROS (Olivera et al., 4 Sep 2025). This yields a dual-domain model of QTX: lifetime-resolved measurements provide functional contrast, while energy-resolved measurements provide structural and spectroscopic contrast during therapeutic irradiation.

4. Quantum biosensing and microarray implementations

Quantum biosensing introduces a different theranostic logic: biomolecular recognition is converted directly into a quantum sensor signal. The multiplexed diamond platform based on nitrogen-vacancy (NV) centers integrates the first multiplexed DNA microarray onto a subnanometer antifouling diamond surface (Chi-Durán et al., 15 Aug 2025). The oxygen-terminated diamond is functionalized by covalent grafting of biotin-PEG-silane in a 15-minute, $\hat{H}|\Psi\rangle = E|\Psi\rangle,$9C reaction, yielding a PEG monolayer with thickness $\sigma^2 \propto t$0 nm as measured by AFM (Chi-Durán et al., 15 Aug 2025). Streptavidin adsorption and binding of biotinylated ssDNA then produce a reported surface density of $\sigma^2 \propto t$1 molecules/$\sigma^2 \propto t$2.

The microarray itself is fabricated by a non-contact dispensing robot delivering 300-picoliter droplets to create a $\sigma^2 \propto t$3 array on a $\sigma^2 \propto t$4 mm$\sigma^2 \propto t$5 chip, with 49 distinct analyte-specific regions and 150 $\sigma^2 \propto t$6m-diameter spots (Chi-Durán et al., 15 Aug 2025). Molecular recognition is implemented through target-induced strand displacement. Each spot contains a surface-anchored DNA substrate hybridized to a shorter complementary strand labeled with multiple Gd$\sigma^2 \propto t$7-DOTA paramagnetic tags; target hybridization displaces this incumbent strand and thereby removes the paramagnetic labels from the NV near-field (Chi-Durán et al., 15 Aug 2025).

The quantum readout is NV spin relaxometry. The governing relation is given as

$\sigma^2 \propto t$8

with the transverse field variance

$\sigma^2 \propto t$9

Here, the Gd$\sigma^2 \propto t^2$0 ensemble decreases $\sigma^2 \propto t^2$1 through fluctuating magnetic noise; displacement of the label restores $\sigma^2 \propto t^2$2, producing a binary quantum readout (Chi-Durán et al., 15 Aug 2025). Experimentally, immobilization of Gd$\sigma^2 \propto t^2$3-labeled duplexes reduces NV $\sigma^2 \propto t^2$4 by 47% for 3.4 Gd$\sigma^2 \propto t^2$5 per strand and 70% for 8.5 Gd$\sigma^2 \propto t^2$6, while complementary invader cDNA restores $\sigma^2 \propto t^2$7 to 93–95% of its original value (Chi-Durán et al., 15 Aug 2025). No recovery occurs with non-complementary strands, and hybridization assays demonstrate less than 6% cross-reactivity.

Although the reported assay is a DNA microarray, the source explicitly states that the platform can, in principle, be extended to proteins and small molecules via aptameric probes (Chi-Durán et al., 15 Aug 2025). A plausible implication is that QTX biosensing need not depend on radioactive tracers or quantum processors; it can also be realized as a massively parallel, label-free or displacement-based front end for molecular diagnostics that can be coupled to therapeutic decision-making.

5. Quantum and plasmonic nanosystems for theranostic function

At the nanoscale, QTX includes both coherent plasmonic devices and semiconductor nanocrystals. The spaser—surface plasmon amplification by stimulated emission of radiation—is presented as a nanoscale analogue of a laser that amplifies and emits coherent surface plasmons rather than photons (Galanzha et al., 2015). The implementation described there uses a gold nanoparticle core as plasmonic resonator and a surrounding gain-medium shell with soluble, biocompatible uranine dye. Experimentally, the spaser exhibits a stimulated-emission spectral width of 0.8 nm and an emission intensity more than $\sigma^2 \propto t^2$8-fold better than quantum dots as the best conventional fluorescent nanoprobes (Galanzha et al., 2015). The source also reports no significant photobleaching, no emission saturation with increasing pumping intensity, imaging of single cancer cells through 1 mm of blood, and single-pulse photothermal cell ablation via transient vapor nano/microbubbles (Galanzha et al., 2015). In the same construct, the plasmonic nanocore serves as a photoacoustic and photothermal contrast agent, making diagnosis and therapy physically co-localized.

Quantum dots (QDs) constitute a broader materials class for theranostics. They are described as semiconductor nanocrystals typically 1–10 nm in diameter, with size- and composition-dependent quantum confinement effects (Biswas et al., 21 May 2025). Their bandgap is summarized as

$\sigma^2 \propto t^2$9

supporting tunable emission from UV-visible to near-infrared, while quantum yield is given by

$$P_{i \rightarrow j}(t) = |\langle j|e^{-i \hat{H} t}|i\rangle|^2.$$0

The review reports quantum yields up to 85–90%, with N-doped TiO$$P_{i \rightarrow j}(t) = |\langle j|e^{-i \hat{H} t}|i\rangle|^2.$$1 QDs at 85.9%, and emphasizes high photostability, multicolor multiplexing, and extensive surface functionalizability (Biswas et al., 21 May 2025). Examples include CdSe/ZnS QDs conjugated to antibodies for HER2 targeting, magnetic or rare-earth-doped QDs for MRI/X-ray/fluorescence multimodality, and drug-delivery constructs such as folic-acid-modified chitosan-encapsulated Mn:ZnS QDs, N-doped carbon QDs loaded with gemcitabine and quinic acid, and pH-sensitive ZnO QDs loaded with doxorubicin (Biswas et al., 21 May 2025).

The distinction between spasers and QDs is important. Spasers couple stimulated emission and strong absorption in a single plasmonic object; QDs are primarily treated as tunable fluorescent nanocarriers, biosensors, and multimodal imaging agents (Galanzha et al., 2015, Biswas et al., 21 May 2025). Both fit the theranostic criterion, but they instantiate different physical mechanisms and different translational constraints.

6. Verification, limitations, and open technical questions

A persistent theme in QTX is that capability claims are bounded by verification, hardware, and biocompatibility. In computational precision medicine, the principal concern is reliability of QC/QML outputs. Formal methods are proposed as the mechanism for addressing this problem: formal specification languages define desired algorithmic behavior, model checking explores all reachable states, theorem proving establishes correctness with mathematical proof, and formal optimization reduces resource usage such as qubit count and gate depth (Bertl et al., 25 Feb 2025). The optimization target is written as

$$P_{i \rightarrow j}(t) = |\langle j|e^{-i \hat{H} t}|i\rangle|^2.$$2

and the same source argues that such methods are especially relevant for genomic marker identification and quantum diagnostic predictors (Bertl et al., 25 Feb 2025).

The imaging and sensing literature identifies different bottlenecks. For $$P_{i \rightarrow j}(t) = |\langle j|e^{-i \hat{H} t}|i\rangle|^2.$$3 PET, production feasibility, radionuclide purity, radiopharmaceutical chemistry, and clinical protocols remain active concerns, even though $$P_{i \rightarrow j}(t) = |\langle j|e^{-i \hat{H} t}|i\rangle|^2.$$4Sc and $$P_{i \rightarrow j}(t) = |\langle j|e^{-i \hat{H} t}|i\rangle|^2.$$5Cu can be efficiently produced in current medical cyclotrons (Matulewicz, 2021). For MV-beam entangled-photon QTX, TES detectors require cryogenic cooling around 100 mK, tissue-induced decoherence still requires validation beyond simulation, and the data rates imply a need for QML-assisted real-time reconstruction (Olivera et al., 4 Sep 2025). For NV-diamond microarrays, the core proof-of-principle is strong, but extension from DNA displacement to broader analyte classes still depends on robust probe chemistry and operation in complex biological environments (Chi-Durán et al., 15 Aug 2025).

Nanomaterial-based QTX introduces toxicity and stability constraints. The QD review emphasizes bioaccumulation, toxicity, and short-term stability as major hurdles; cadmium-containing QDs may accumulate in liver, kidney, spleen, brain, and reproductive organs, with metal-ion leaching, ROS generation, and photodegradation as key mechanisms of harm (Biswas et al., 21 May 2025). Mitigation strategies include core-shell engineering, biocompatible coatings, cadmium-free compositions, concentration control, and ultrasmall designs for renal clearance (Biswas et al., 21 May 2025). These concerns do not negate theranostic utility, but they delimit the circumstances under which QTX can become clinically credible.

A second misconception is that “quantum advantage” in medicine is already a settled clinical fact. The surveyed literature is more cautious. Some contributions are conceptual frameworks, some are simulation studies, and some are experimentally validated but preclinical platforms (Maniscalco et al., 2022, Olivera et al., 4 Sep 2025, Chi-Durán et al., 15 Aug 2025). This suggests that QTX should presently be understood as an emerging research program whose modalities are heterogeneous: some, such as PET isotope selection, are close to established nuclear-medicine workflows; others, such as entangled-photon radiotherapy imaging or large-scale QML-driven regulatory-network inference, remain developmental. What unifies them is not a single device class, but the attempt to bind diagnosis, mechanism inference, and therapy into a single quantum-informed operational loop.