arXiv subjects

Andrei Zhukov

Publications and source records attributed to Andrei Zhukov.

8 recordsLinked to original sources

arXiv2026

A robust and modular cesium magneto-optical trap using diverging laser beams

Magneto-optical traps (MOTs) are a workhorse technology for neutral-atom quantum sensing, simulation, and computing. Here, we demonstrate a MOT optimized for inertial quantum sensing and precision metrology applications. Our MOT traps $4\times10^8$ cesium (Cs) atoms in a robust, modular system that achieves stable operation through two design features: (1) we use diverging cooling laser beams to reduce unwanted reflections and (2) optical elements are mounted in a fiber-coupled, compact, and modular cage rigidly attached to the vacuum chamber. Following polarization-gradient cooling (PGC), the system produces atom samples with temperatures below 10 $μ$K, similar to those achieved in conventional MOTs that use collimated laser beams. In addition, we observe trapping of $2\times10^7$ Cs atoms in a non-conventional MOT geometry, where the cooling laser beams are diagonal to the principal axis of the quadrupole magnetic field.

physics.atom-ph↗

arXiv2026

Routine laboratory trajectories encode the onset of organ-level complications in cancer

Routine laboratory panels drawn during cancer treatment constitute longitudinal physiological recordings of organ function, yet their temporal structure is discarded by single-timepoint prognostic tools. A transformer trained on 2,777,595 laboratory measurements from 3,905 patients with multiple myeloma or ovarian cancer predicted the two-year onset of 162 treatment-associated complications, including therapy-related myelodysplastic syndromes, spanning eight clinical categories, achieving 1.5- to 6.1-fold enrichment above prevalence at the group level. It matched or outperformed non-sequential baselines across grouped endpoints (AUROC gains up to +0.11), demonstrating that longitudinal laboratory trajectories capture evolving complication-specific physiology inaccessible from isolated measurements. Predictions generalised across both cancers, divergence concentrating in disease-specific complications, and biomarker masking recovered signatures consistent with established pathophysiology. External validation on MIMIC-IV and MMRF CoMMpass confirmed transferability across independent healthcare systems (AUROC up to 0.85). Routine oncological laboratory data encode organ deterioration weeks to months before clinical onset, enabling complication-specific surveillance without additional testing infrastructure.

cs.LG↗

arXiv2026

Agentic clinical reasoning over longitudinal myeloma records: a retrospective evaluation against expert consensus

Multiple myeloma is managed through sequential lines of therapy over years to decades, with each decision depending on cumulative disease history distributed across dozens to hundreds of heterogeneous clinical documents. Whether LLM-based systems can synthesise this evidence at a level approaching expert agreement has not been established. A retrospective evaluation was conducted on longitudinal clinical records of 811 myeloma patients treated at a tertiary centre (2001-2026), covering 44,962 documents and 1,334,677 laboratory values, with external validation on MIMIC-IV. An agentic reasoning system was compared against single-pass retrieval-augmented generation (RAG), iterative RAG, and full-context input on 469 patient-question pairs from 48 templates at three complexity levels. Reference labels came from double annotation by four oncologists with senior haematologist adjudication. Iterative RAG and full-context input converged on a shared ceiling (75.4% vs 75.8%, p = 1.00). The agentic system reached 79.6% concordance (95% CI 76.4-82.8), exceeding both baselines (+3.8 and +4.2 pp; p = 0.006 and 0.007). Gains rose with question complexity, reaching +9.4 pp on criteria-based synthesis (p = 0.032), and with record length, reaching +13.5 pp in the top decile (n = 10). The system error rate (12.2%) was comparable to expert disagreement (13.6%), but severity was inverted: 57.8% of system errors were clinically significant versus 18.8% of expert disagreements. Agentic reasoning was the only approach to exceed the shared ceiling, with gains concentrated on the most complex questions and longest records. The greater clinical consequence of residual system errors indicates that prospective evaluation in routine care is required before these findings translate into patient benefit.

cs.AI↗

arXiv2025

Three-dimensional Reconstruction and Propagation of an Asymmetric Flux-rope Coronal Mass Ejection

We report on the characterization of a coronal mass ejection (CME) observed on 22 October 2003 by the LASCO-C2 and C3 coronagraphs over a time interval of 6 hours. This CME clearly appears as an asymmetric flux-rope in self-similar expansion and in spite of having a single vantage point, this relatively simple morphology and the geometry of the observations allow us to reconstruct its shape and its trajectory. The images are first processed in order to remove most of the background or foreground coronal structures (essentially streamers) which are superimposed on the CME. The analysis is based on forward modeling of an asymmetric structure where the plasma is homogeneously distributed in a thin shell and synthetic images are calculated from Thomson scattering by the electrons. They are best fitted to the images to determine the exact shape of the flux rope, and to track its evolution characterized by a radial propagation in self-similar expansion. The analysis supports the forward propagation (over the backward one) characterized by constant accelerations of 76 and 46 m/sec^2 for the front and rear parts of the CME, respectively. The respective velocities at 20 Rsun reach 2000 and 1100 km/s, and its mass unbiased by projection effects is estimated at 1.7x10^16 g. Altogether, these are quite exceptionally large values among CMEs and imply a very energetic event. No erupting event could be identified at or near the calculated initial location of the CME. It lies 25° west of the active region NOAA 10484, the site of contemporary, violent, and recurrent solar eruptions (the ``Halloween event'') and a putative connection would require a strong deflection over a very short path. Therefore, the origin of this CME remains unclear.

astro-ph.SR↗

arXiv2025

Buildup, Explosion, and Untwisting of a Solar Active Region Jet Observed with Solar Orbiter, IRIS, and SDO

We present detailed analysis of an active region coronal jet accompanying a minifilament eruption that is fully captured and well-resolved in high spatial resolution 174A coronal images from Solar Orbiters Extreme Ultraviolet Imager (EUI). The active region jet is simultaneously observed by the Interface Region Imaging Spectrograph (IRIS) and the Solar Dynamics Observatory (SDO). An erupting minifilament is rooted at the edge of an active region where mixed-polarity magnetic flux is present. Minority-polarity positive flux merges and cancels with the active regions dominant negative flux at an average rate of 1019 Mx/hr, building a minifilament-holding flux rope and triggering its eruption. The eruption shows a slow rise followed by a fast rise, akin to large-scale filament eruptions. EUI images and Mg II k spectra, displaying simultaneously blueshifts and redshifts at the opposite edges of the spire, indicate counterclockwise untwisting of the jet spire. This jet is the clearest, most comprehensively observed active-region jet with this instrument set, displaying striking similarities with quiet Sun and coronal hole jets. Its magnetic, thermal, and kinetic energies suggest a significant contribution to local coronal heating. We conclude that magnetic flux cancelation builds a minifilament-carrying twisted flux rope and also eventually triggers the flux ropes eruption that makes the coronal jet, in line with our recent results on the buildup and explosion of solar coronal jets in quiet Sun and coronal holes. That is, this active region jet clearly works the same way as the vast majority of quiet Sun and coronal hole jets.

astro-ph.SR↗

arXiv2025

Heatpipe-cooled in-vacuum electromagnet for quantum science experiment

Quantum inertial sensors test general relativity, measure fundamental constants, and probe dark matter and dark energy in the laboratory with outstanding accuracy. Their precision relies heavily on carefully choreographed quantum control of the atomic states with a collection of lasers, microwaves, electric and magnetic fields. Making this technology available outside of the laboratory would unlock many applications, such as geophysics, geodesy and inertial navigation. However, this requires an apparatus of reduced size, weight, power use and increased robustness, modularity and ease-of-use. Here, we describe the design and implementation of an in-vacuum electromagnet able to create the magnetic fields necessary for various quantum control operations, such as magneto-optical trapping or magnetic levitation to assist evaporative cooling. Placing the electromagnet inside the vacuum chamber has significant advantages, such as fast switching times that are not limited by induced current inside the vacuum chamber metal, reduced size, weight and power usage. However, dissipating the heat produced typically requires complex designs that include bulky metal heatsinks or cooling using water or cryogens. Our design implements heatpipes in a compact, low-vibration and robust apparatus, which use a phase transition in the working fluid to achieve thermal conductivity that is more than one hundred times larger than that of typical bulk metal. We show that the setup can conduct more than 50 W of thermal power in a configuration that provides ample optical access and is compatible with the ultra-high vacuum requirements of atomic and molecular experiments.

physics.atom-ph↗

arXiv2024

Incorporating Anatomical Awareness for Enhanced Generalizability and Progression Prediction in Deep Learning-Based Radiographic Sacroiliitis Detection

Purpose: To examine whether incorporating anatomical awareness into a deep learning model can improve generalizability and enable prediction of disease progression. Methods: This retrospective multicenter study included conventional pelvic radiographs of 4 different patient cohorts focusing on axial spondyloarthritis (axSpA) collected at university and community hospitals. The first cohort, which consisted of 1483 radiographs, was split into training (n=1261) and validation (n=222) sets. The other cohorts comprising 436, 340, and 163 patients, respectively, were used as independent test datasets. For the second cohort, follow-up data of 311 patients was used to examine progression prediction capabilities. Two neural networks were trained, one on images cropped to the bounding box of the sacroiliac joints (anatomy-aware) and the other one on full radiographs. The performance of the models was compared using the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity. Results: On the three test datasets, the standard model achieved AUC scores of 0.853, 0.817, 0.947, with an accuracy of 0.770, 0.724, 0.850. Whereas the anatomy-aware model achieved AUC scores of 0.899, 0.846, 0.957, with an accuracy of 0.821, 0.744, 0.906, respectively. The patients who were identified as high risk by the anatomy aware model had an odds ratio of 2.16 (95% CI: 1.19, 3.86) for having progression of radiographic sacroiliitis within 2 years. Conclusion: Anatomical awareness can improve the generalizability of a deep learning model in detecting radiographic sacroiliitis. The model is published as fully open source alongside this study.

cs.CV↗

arXiv2021

The Magnetic Origin of Solar Campfires

Solar campfires are fine-scale heating events, recently observed by Extreme Ultraviolet Imager (EUI), onboard Solar Orbiter. Here we use EUI 174Å images, together with EUV images from SDO/AIA, and line-of-sight magnetograms from SDO/HMI to investigate the magnetic origin of 52 randomly selected campfires in the quiet solar corona. We find that (i) the campfires are rooted at the edges of photospheric magnetic network lanes; (ii) most of the campfires reside above the neutral line between majority-polarity magnetic flux patch and a merging minority-polarity flux patch, with a flux cancelation rate of $\sim$10$^{18}$Mx hr$^{-1}$; (iii) some of the campfires occur repeatedly from the same neutral line; (iv) in the large majority of instances, campfires are preceded by a cool-plasma structure, analogous to minifilaments in coronal jets; and (v) although many campfires have `complex' structure, most campfires resemble small-scale jets, dots, or loops. Thus, `campfire' is a general term that includes different types of small-scale solar dynamic features. They contain sufficient magnetic energy ($\sim$10$^{26}$-10$^{27}$ erg) to heat the solar atmosphere locally to 0.5--2.5MK. Their lifetimes range from about a minute to over an hour, with most of the campfires having a lifetime of $<$10 minutes. The average lengths and widths of the campfires are 5400$\pm$2500km and 1600$\pm$640km, respectively. Our observations suggest that (a) the presence of magnetic flux ropes may be ubiquitous in the solar atmosphere and not limited to coronal jets and larger-scale eruptions that make CMEs, and (b) magnetic flux cancelation is the fundamental process for the formation and triggering of most campfires.

astro-ph.SR↗