Pith. sign in

REVIEW 2 major objections 5 minor 30 references

High contrast dual-mode optical and 13C magnetic resonance imaging in diamond particles

T0 review · 2 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Pith's one-line read Diamond particles with NV centers can be imaged by light and by MRI at the same time, from the same green beam.

desk verdict Solid dual-mode imaging proof of concept with a clean background-suppression demo; the Regime III acceleration claims rest on an equal-per-sample-cost assumption that does not survive realistic cost ratios. read the letter →

arxiv 1909.08064 v3 pith:YHE6VCYX submitted 2019-09-04 physics.ins-det cond-mat.mtrl-sciquant-ph

classification physics.ins-detcond-mat.mtrl-sciquant-ph
keywords nitrogen-vacancycentersdiamondmicroparticles13Chyperpolarizationdual-modeimagingopticalfluorescencemagneticresonancebackgroundsuppressionconjugate-spacesampling
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper sets out to show that diamond microparticles rich in nitrogen-vacancy (NV) centers can act as dual-mode imaging agents: the same green laser that makes them fluoresce also polarizes nearby $^{13}$C nuclei, so the particles become visible in both an optical image and a magnetic resonance image. The authors demonstrate fluorescence imaging and hyperpolarized $^{13}$C MRI of a diamond phantom, reporting about 0.3% $^{13}$C polarization in 40 seconds under 1 W of light (roughly 280-fold signal enhancement over thermal polarization at 7 T). They show that switching either the applied magnetic field (for optics) or the microwave sweep direction (for MRI) modulates the signal on demand, letting them recover diamond signals buried under a fluorescent dye and under $^{13}$C-methanol, with demonstrated background suppression factors of about 2 and 5. They further argue that because optics and MRI sample Fourier-conjugate spaces, a hybrid protocol that takes a few MRI k-space points and then raster-scans only the pixels those points flag can accelerate sparse-image acquisition by an order of magnitude or more, with a proportional cut in optical power.

What carries the argument

The load-bearing element is NV-mediated optical hyperpolarization of lattice $^{13}$C nuclei: sub-bandgap green light polarizes NV$^-$ electron spins, and microwave sweeps across the NV ESR spectrum drive Landau-Zener transitions that transfer polarization orientation-independently to the $^{13}$C bath, producing ~0.3% $^{13}$C polarization in 40 s at ~38 mT. A second mechanism carries the background-suppression claim: the NV fluorescence depends on the angle between the NV axis and an applied field (simulated by a seven-level kinetic model), so a pulsed field modulates the optical signal; and reversing the microwave sweep reverses the $^{13}$C hyperpolarization sign, providing full-contrast modulation in MRI. The acceleration argument rests on Fourier reciprocity: optics samples real space while MRI samples k-space, so a truncated k-space window gives a sinc-blurred real-space image that, after thresholding, confines the optical raster to a sparse subset of pixels.

What would settle it

Time the actual protocol on the demonstrated hardware: record total wall-clock acquisition for a sparse phantom with one MRI k-space line per ~40 s hyperpolarization cycle plus optical rastering, and compare with pure optical rastering at the same target resolution. If the measured speedup is far below $(1-s)^{-1/2}$, the equal-cost premise is the reason.

Watch

Extended reading notes

Core claim

The central claim is that a single material platform - diamond particles hosting NV centers - can be imaged simultaneously in the optical and MR domains, with each mode improving the other. Optically, the particles fluoresce brightly under 520 nm light; the same light polarizes the NV electron spins, and chirped microwave sweeps transfer that polarization to $^{13}$C nuclei, giving a hyperpolarized $^{13}$C MRI signal reported as over three orders of magnitude brighter than conventional MRI at low field (enhancement ~280 over thermal at 7 T, ~206 over 9.4 T). Since the NV fluorescence depends on applied magnetic field and the hyperpolarization sign depends on microwave sweep direction, both image modes can be modulated on demand, enabling lock-in background suppression in optics and difference imaging in MRI. The paper also proposes a third regime: a hybrid acquisition protocol that samples a few low-order k-space points by MRI, thresholds the resulting blurry image, and feeds that information forward to restrict real-space optical scanning to promising pixels; at high sparsity $s$ this yields acceleration scaling as $(1-s)^{-1/2}$, with an optimal k-space sample count scaling as $(1-s)^{1/4}$.

Load-bearing premise

The acceleration calculation assumes every extra sample costs the same time in optics and in MRI, even though one MRI k-space sample can require a fresh hyperpolarization cycle while an optical pixel is read in milliseconds; if that ratio is large, the headline speedup shrinks.

Editorial extensions

If this is right

  • Diamond particles can be tracked by optical microscopy and by MRI from the same illumination source, so scattering environments that blind optics can be cross-checked with MRI and vice versa.
  • On-demand modulation of both modes enables background-free imaging: lock-in suppression recovers diamond signals under a dye that is twice as bright, and difference imaging cancels a $^{13}$C-methanol background five times stronger.
  • In sparse-imaging settings, the hybrid k-space/real-space protocol promises more than an order-of-magnitude acquisition speedup with a matching reduction in delivered optical power.
  • Because polarization is replenished continuously at low field and is detection-field agnostic, low-field MRI becomes practical: the agent's brightness does not depend on the detection magnet.
  • MRI resolution can be pushed toward optical resolution by rastering a focused beam to hyperpolarize one pixel at a time, trading power density for pixel size.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the equal per-sample cost assumption is replaced by a realistic ratio where one MRI k-space line requires a fresh ~40 s hyperpolarization cycle, the optimal k-space sample count and the $(1-s)^{-1/2}$ acceleration both shift; the practical speedup will depend on that cost ratio and may be much smaller than plotted.
  • The same Fourier-reciprocity trick could be combined with compressed sensing: a random k-space subsample plus convex reconstruction could produce the feed-forward mask, potentially improving the blurry-image quality and pushing the protocol to lower sparsity.
  • The 0.3% polarization number was obtained in 200 micrometer particles; the paper itself notes roughly 10^-2 lower hyperpolarizability for <100 nm particles, so a key extension is to verify whether material improvements (annealing, $^{13}$C enrichment) close that gap before clinical-scale agents are practical.
  • The background-suppression scheme is generic: any agent whose fluorescence or hyperpolarization can be switched on demand could use the same lock-in/difference logic, so the method may transfer to other optically polarizable spin labels.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. The paper reports dual-mode imaging of diamond microparticles using NV-center fluorescence and optically pumped 13C hyperpolarization. The authors demonstrate about 0.3% 13C polarization in 40 s under 1 W green illumination, corresponding to roughly 280-fold enhancement over thermal 13C at 7 T, and use it for FLASH MRI of a ring phantom. They also demonstrate optical and MR background suppression by field modulation and MW-sweep reversal, and they propose a 'Regime III' hybrid protocol that samples low-order k-space with MRI and feeds the thresholded result forward to guide real-space optical rastering. Simulations of this protocol are used to claim acquisition acceleration and optical power reduction scaling as (1-s)^-1/2 at high sparsity s, with 'more than an order of magnitude' and 'as much as two orders of magnitude' speedups in sparse-imaging scenarios.

Significance. The experimental part is a useful proof of concept: the same green illumination both fluoresces and hyperpolarizes, and the demonstrated lock-in and sweep-reversal cancellation give concrete, quantified background suppression. The use of measured rate constants and experimental parameters (e.g., Ref. [20] kinetics, T1 data) in the supporting analysis is a strength. However, the paper's headline quantitative claims for Regime III are not yet supported: the acceleration and power-reduction scalings assume equal per-sample time costs for optics and MRI and omit the fixed 40 s hyperpolarization overhead. Until those claims are re-derived or substantially qualified, the paper's significance is prospective rather than established.

major comments (2)
  1. [Main text, 'Accelerated conjugate-space imaging'; SI §III.A, Eq. (8)] The central Regime III claim of order-of-magnitude time savings and up to two orders of magnitude acceleration (Fig. 4E, Abstract) rests on the assumption that 'the time cost to be accrued per sample (pixel) is identical for both optical and MR imaging dimensions.' In the demonstrated setup this is not the case: each MR acquisition requires a 40 s hyperpolarization step (main text, Results; Fig. 1B) followed by a FLASH train with TR = 6 ms, while an optical pixel can be read in milliseconds. The MR term in Eq. (8) is l^2/N^2 with no cost ratio; scaling it by eta gives l_opt proportional to eta^{-1/4}(1-s)^{1/4} and tau_opt proportional to eta^{1/2}(1-s)^{1/2}. For eta values of order 10^3-10^4, representative of the 40 s hyperpolarization overhead, the advertised savings vanish and the protocol can be slower than full rastering. The paper notes that results can be scaled by eta but never computes the regime in which the headline numbers survive; this is load-bearing for the abstract and Table I.
  2. [Table I, Regime III row; main text, 'Finally, we comment that imaging acceleration results in a lower total optical…] The claim of power reduction by the same factor as the acceleration omits the optical power spent on hyperpolarization. The experimental hyperpolarization consumes 1 W for 40 s (Fig. 1B), a fixed energy cost per MR acquisition. At high sparsity, where the raster energy is reduced by the acceleration factor, this fixed cost can dominate; the power-reduction factors of 14-25 quoted in the text and Table I are therefore not the total optical power budget. A full energy accounting including the hyperpolarization laser, or a clear statement that the quoted reduction applies only to the raster component, is needed.
minor comments (5)
  1. [Abstract and main text (Fig. 1C)] The abstract states the particles are 'over three-orders of magnitude brighter than in conventional MRI,' but the measured enhancement is 280-fold over thermal 13C at 7 T; specify the reference field or revise the wording.
  2. [Fig. 4B and surrounding text] The text quotes '16-fold' and '~14 times' acceleration, while the figure panel says 'Acceleration = 8'; clarify what is being counted (reduction in real-space samples vs total acquisition time).
  3. [SI §III.A and Fig. S2] The analytic model uses r0 = 0.75, selected by thresholding the sinc shoulder at 0.3; a sensitivity analysis or a principled selection rule for r0 would strengthen the scaling results.
  4. [Fig. 2D and definition of Delta] For full sign reversal the denominator I+I in the modulation contrast is zero; state how the 194% contrast is computed, e.g., using absolute values or a noise floor.
  5. [References] Several references are duplicated or repeated (e.g., Refs. [8] and [64], [13] and [67]); consolidate to avoid confusion.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular reduction found: experimental dual-mode results are measured against external benchmarks, and the Regime III acceleration analysis is an explicitly conditional analytic model rather than a fitted prediction.

full rationale

The paper's load-bearing experimental assertions are direct measurements: approximately 0.3% 13C polarization under 1 W illumination in 40 s (Fig. 1B-C), 10% optical and 194% MR modulation contrast (Fig. 2), and background suppression by factors of 2 and 5 (Fig. 3). These do not reduce to definitions or to fitted parameters. The NV fluorescence simulation uses the externally measured 7-level rate constants of Ref. [20], and the hyperpolarization mechanism is supported by the independently published, experimentally grounded work in Refs. [8,11]; self-citation here is enabling background, not an unverified uniqueness claim. The Regime III acceleration scalings (l_opt proportional to (1-s)^(1/4), tau proportional to (1-s)^(1/2)) are derived in SI Eq. (8) from a sinc-blur model, with r0 = 0.75 set by the stated 0.3 threshold rather than fitted to simulation outputs; the agreement with the paper's own simulations is a consistency check, not a fit renamed as prediction. The paper explicitly flags the equal per-sample time-cost assumption and says results can be scaled by an appropriate cost ratio eta. That caveat is a modeling limitation affecting quantitative applicability of the acceleration numbers, but it is not a circular reduction of the conclusion to its inputs. No self-definitional, fitted-input, or imported-uniqueness step was found.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The paper relies on the prior NV-hyperpolarization machinery (Refs [8], [11]) plus standard Fourier and spin kinetics. The genuinely new proposal, Regime III acceleration, depends on a few hand-picked constants (r0, threshold) and an equal-time assumption that are not experimentally validated.

free parameters (3)
  • r0 (sinc-shoulder threshold radius) = 0.75
    Introduced in SI Eq. (7)-(8) as the point where the truncated k-space sinc point-spread function falls to 0.3 of its maximum; it sets the effective real-space blur and therefore directly controls the derived l_opt and tau scalings in the acceleration model.
  • optical sampling threshold = 0.1 times mean pixel value
    Chosen in the Fig 4B protocol to decide which real-space pixels to scan after the coarse MRI pass; affects the acceleration and fidelity trade-off and is not optimized systematically.
  • gamma_0 (optical/MR SNR ratio at depth 0) = 15
    Measured from Fig 1E-F after power and time normalization (SI Section II.A); used to extrapolate Regime I critical depth, SNR gain, and power reduction. It is an experimental calibration, included for transparency.
assumptions (6)
  • standard math k-space truncation to an l x l window produces real-space convolution with a sinc kernel (Fourier convolution theorem)
    Used in SI Eq. (5)-(6) to derive the point-spread function and the acceleration scaling.
  • domain assumption A seven-level NV rate model with transition rates from Ref [20] describes powder-averaged fluorescence response to magnetic field
    Used for the optical modulation simulation in Fig 2A and SI Fig S11; the paper itself notes charge-state interconversion and scattering are omitted, and the model overestimates contrast (40% predicted vs 10% measured).
  • domain assumption Randomly oriented particle ensemble samples all NV axis angles uniformly
    Needed for the angular average in SI Eq. (21) that produces the fluorescence modulation curve in Fig 2A.
  • domain assumption Scattering media are homogeneous with Gaussian angular spreading, so a Wigner-function slice model gives resolution that degrades linearly with depth
    Used in SI Section IV.D and Fig S3 to compare optical and MRI resolution in tissue; this is an idealized tissue model.
  • domain assumption Per-pixel imaging time cost is identical for optical and MR sampling
    Stated in the main text before Fig 4C-F and used to derive all Regime III acceleration scalings; the paper notes it can be scaled by a cost ratio eta but does not quantify eta.
  • domain assumption 13C hyperpolarization is detection-field agnostic and T1 is long enough for a roughly one-second shuttle to 9.4T
    Underpins the low-field polarization, high-field detection workflow; SI Fig S4 provides supporting T1 data, but the transfer and field-cycling losses are not measured in this paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of High contrast dual-mode optical and 13C magnetic resonance imaging in diamond particles." pith.science (2026). https://pith.science/paper/YHE6VCYX

@misc{pith2026190908064,
  author       = {Pith},
  title        = {Pith review of: High contrast dual-mode optical and 13C magnetic resonance imaging in diamond particles},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YHE6VCYX}},
  note         = {Machine review of arXiv:1909.08064}
}
read the original abstract

Multichannel imaging -- the ability to acquire images of an object through more than one imaging mode simultaneously -- has opened interesting new perspectives in areas ranging from astronomy to medicine. Visible optics and magnetic resonance imaging (MRI) offer complementary advantages of resolution, speed and depth of penetration, and as such would be attractive in combination. In this paper, we take first steps towards marrying together optical and MR imaging in a class of biocompatible particulate materials constructed out of diamond. The particles are endowed with a high density of quantum defects (Nitrogen Vacancy centers) that under optical excitation fluoresce brightly in the visible, but also concurrently electron spin polarize. This allows the hyperpolarization of lattice 13C nuclei to make the particles over three-orders of magnitude brighter than in conventional MRI. Dual-mode optical and MR imaging permits immediate access to improvements in resolution and signal-to-noise especially in scattering environments. We highlight additional benefits in background-free imaging, demonstrating lock-in suppression by factors of 2 and 5 in optical and MR domains respectively. Ultimate limits could approach as much as two orders of magnitude in each domain. Finally, leveraging the ability of optical and MR imaging to simultaneously probe Fourier-reciprocal domains (real and k-space), we elucidate the ability to employ hybrid sub-sampling in both conjugate spaces to vastly accelerate dual-image acquisition, by as much as two orders of magnitude in practically relevant sparse-imaging scenarios. This is accompanied by a reduction in optical power by the same factor. Our work suggests interesting possibilities for the simultaneous optical and low-field MR imaging of targeted diamond nanoparticles.

Figures

Figures reproduced from arXiv: 1909.08064 by the authors.

Figure 1
Figure 1. Dual-mode optical and 13C MR imaging. (A) Experiment schematic. Diamond particles with NV￾centers are imaged with fluorescence under green (520nm) excitation by a CMOS detector, as well as under 13C MRI through polarization transferred to lattice 13C nuclei from optically polarized NV￾electrons. (B) Hyperpolarization and detection protocol. 13C DNP occurs at low-field ∼38mT under MW sweeps across the NV- ESR spectru… view at source ↗
Figure 2
Figure 2. On-demand dual-mode image modulation. (A) (i) Normalized fluorescence signal for randomly oriented diamond particle ensemble under an applied magnetic field (points: experiment, purple-line: simulation). We ascribe the discrepancy to scattering effects. (ii) Optical modulation under 40±2mT pulsed magnetic field showing a signal contrast ∼10%. (B) Optical images under 0 and ∼40mT applied field showing weak ∼10% optic… view at source ↗
Figure 3
Figure 3. Dual-mode background suppression. (A-B) Schematic of imaging phantoms. Diamonds are arranged in ring-shaped phantom, co￾situated with Alexa 647 dye and 13C-methanol that present an artificial backgrounds for optical and MR imaging respectively. (C-D) Optical and MR images with the background. Dashed lines serve as a guide to the eye for the imaging phantom. Diamond particles are indistin￾guishable from the backgroun… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Accelerated x-k conjugate-space imaging. (A) Protocol for accelerated imaging. ` samples of the image are first acquired in k-space, and the resulting image upon thresholding is fed-forward to constrain the real-space points to be scanned over. (B) Exemplary scenario w…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 30 canonical work pages

  1. [20]

    signal path

    Similarities Apart from the differences described above, CS and dual- mode accelerated protocol share some similarities in the follow- ing ways. Subsampling – Both methods exploit subsampling to avoid scanning across the entire imaging space and therefore gain time saving. CS MRI subsamples k-space while our protocol subsam- ples both Fourier conjugate sp...

  2. [1]

    Compressed sensing 14

  3. [2]

    Accelerated dual-mode protocol 14

  4. [3]

    round-trip

    Similarities 15 IV . Imaging efficiency of two modalities 15 A. Imaging efficiency of optics 15 B. Imaging efficiency of MRI 16 C. Estimation of imaging SNR in scattering media 16 D. Estimation of imaging resolution in scattering media 16 V . Materials 17 A. Diamond particles 17 B. T1 relaxation in diamond 17 C. Hyperpolarization in 13C enriched diamond 18 D...

  5. [4]

    round-trip

    Red fluorescence loss. Optical imaging incurs “round-trip” losses due to the need to also collect NV fluorescence. Similar to Step 1 above, one can estimate these attenuation and scattering losses. We take the attenuation coefficient at 650nm as∼1 cm−1, and the scattering coefficient as ∼ 12.1cm−1 (calculated based on fatty tissue data in [33]). For diamonds ...

  6. [5]

    The finite numerical aperture (NA) of the detection optics restricts the total number of photons that can be collected

    Geometric effects in light collection. The finite numerical aperture (NA) of the detection optics restricts the total number of photons that can be collected. Only fluorescence light within a solid angle Ω can be captured by the objective lens. We have: Ω = 2π(1− cos(ϑmax)), (13) whereϑmax = arcsin(NA/nd). Assuming that the fluorescence emission is initially...

  7. [6]

    round-trip

    Signal amplitude reduction by lock-in detection. Finally, we consider the amplitude loss caused by lock-in detection in practi- cal scenarios. Lock-in techniques are normally used when optical background presents, and the signal of interest can be discerned by the method. However, such benefit of background suppression 16 comes at the cost of reduced overa...

  8. [18]

    CS is the origin of a major evolution in signal processing [36, 37]

    Compressed sensing Let us first briefly review basics of CS, in order to better com- pare it with our protocol. CS is the origin of a major evolution in signal processing [36, 37]. It is widely applied in photography, medical imaging, as well as astronomy, because of its ability to recover the signal from very few linear measurements. Operating regime – CS,...

Show all 30 references
  1. [19]

    generated

    Accelerated dual-mode protocol The accelerated dual-mode protocol proposed in this work takes sub-sampled k-space data by MR as a reference to guide real-space measurement by optics. This protocol warrants acqui- sition time saving concurrently with optical power reduction. Op...

  2. [21]

    Scattering and attenuation lead to loss of photon flux

    Green photon scattering and loss at media. Scattering and attenuation lead to loss of photon flux. The attenuation loss can be evaluated by Beer’s law:φp φm = exp{ ∫d0 0 −[α(d) +µ(d)]· dd}, whereφp is the photon flux of the incident pump beam, andφm is the part that reaches diam...

  3. [22]

    In fact, during optical il- lumination, only a small portion of the incident green photons finally convert to red photons by NV centers

    Green-to-red photon conversion. In fact, during optical il- lumination, only a small portion of the incident green photons finally convert to red photons by NV centers. We now estimate this conversion rate ηo,2 based on Ref [38] . A waveguide is employed on a 0.1 ppm NV diamond...

  4. [23]

    NV centers emit in red, and a portion of the emission light can be restricted within the diamond because of total internal reflection (TIR)

    Total internal reflection loss. NV centers emit in red, and a portion of the emission light can be restricted within the diamond because of total internal reflection (TIR). The critical angle of TIR for a diamond-media interface is ϑc = arcsin (nm/nd) = 24.6◦. We calculate ηo,3 ...

  5. [27]

    Green laser loss at media. Similar to the optical imaging case, since the hyperpolarization is optically induced, we con- tinue to incur one-way optical scattering and attenuation losses through the imaging media, corresponding to a factor ηm,1 = ηo,1 = 9.6× 10−3 as above

  6. [28]

    Let us now estimate efficiency of the 13C hyperpolarization process

    Quantum efficiency of 13C hyperpolarization. Let us now estimate efficiency of the 13C hyperpolarization process. In Fig- ure 1C, the hyperpolarization signal is obtained by 1W laser irra- diation with a duration of 40s; as a result, the 10mg natural abun- dance diamond sample i...

  7. [29]

    The sample filling factor in MR imaging plays a similar role as the geometric fac- tors related to finite numerical aperture in optical imaging

    Geometric factors related to detection. The sample filling factor in MR imaging plays a similar role as the geometric fac- tors related to finite numerical aperture in optical imaging. In our experiments, the volume of the coil is approximately 2400 mm3, and the typical volume o...

  8. [30]

    object region

    Detection frequency. We note that the hyperpolarization process delivers 13C polarization that is agnostic to detection field. However, the choice of detection field does play a role in determining the final obtained SNR, scaling as ∝ (Qω0)1/2 [27], whereQ is the quality factor o...

  9. [58]

    Zhang, R

    X.-Q. Zhang, R. Lam, X. Xu, E. K. Chow, H.-J. Kim, and D. Ho, Advanced materials 23, 4770 (2011)

  10. [59]

    G. Xi, E. Robinson, B. Mania-Farnell, E. F. Vanin, K.-W. Shim, T. Takao, E. V . Allender, C. S. Mayanil, M. B. Soares, D. Ho,et al., Nanomedicine: Nanotechnology, Biology and Medicine 10, 381 (2014)

  11. [60]

    E. K. Chow, X.-Q. Zhang, M. Chen, R. Lam, E. Robinson, H. Huang, D. Schaffer, E. Osawa, A. Goga, and D. Ho, Science translational medicine 3, 73ra21 (2011)

  12. [61]

    Ho, Clinical Trials (2020)

    D. Ho, Clinical Trials (2020)

  13. [62]

    P. W. Goodwill and S. M. Conolly, IEEE transactions on medical imaging 29, 1851 (2010)

  14. [63]

    Zurbuchen, F

    U. Zurbuchen, F. Poch, O. Gemeinhardt, M. E. Kreis, S. M. Niehues, J. L. Vahldieck, and K. S. Lehmann, Acta Radiologica 58, 164 (2017)

  15. [64]

    A. Ajoy, K. Liu, R. Nazaryan, X. Lv, P. R. Zangara, B. Safvati, G. Wang, D. Arnold, G. Li, A. Lin, et al., Science Advances 4, eaar5492 (2018)

  16. [65]

    M. L. Denton, M. S. Foltz, L. E. Estlack, D. J. Stolarski, G. D. Noojin, R. J. Thomas, D. Eikum, and B. A. Rockwell, Investigative ophthalmology & visual science 47, 3065 (2006)

  17. [66]

    Takegoshi and C

    K. Takegoshi and C. McDowell, Chemical physics letters 116, 100 (1985)

  18. [67]

    M. A. Frey, M. Michaud, J. N. VanHouten, K. L. Insogna, J. A. Madri, and S. E. Barrett, Proceedings of the National Academy of Sciences 109, 5190 (2012)

  19. [68]

    E. Y . Chang, J. Du, and C. B. Chung, Journal of magnetic reso- nance imaging 41, 870 (2015)

  20. [69]

    Thorlabs, Apd410x operation manual, Tech. Rep. (2018)

  21. [70]

    Lightning-link rapid alexa fluor 647 conjugate data sheet , Tech. Rep. (2018)

  22. [71]

    Dual mode

    A. S. Merbach, L. Helm, and E. Toth, The chemistry of con- trast agents in medical magnetic resonance imaging (John Wiley & Sons, 2013). 9 Supplementary Information Background-free dual-mode optical and 13C magnetic resonance imaging in diamond particles X. Lv,1 J. H. Walton,2...

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.