arXiv subjects

Yun Jin

Publications and source records attributed to Yun Jin.

5 recordsLinked to original sources

arXiv2026

AMGFormer: Adaptive Multi-Granular Transformer for Brain Tumor Segmentation with Missing Modalities

Multimodal MRI is essential for brain tumor segmentation, yet missing modalities in clinical practice cause existing methods to exhibit >40% performance variance across modality combinations, rendering them clinically unreliable. We propose AMGFormer, achieving significantly improved stability through three synergistic modules: (1) QuadIntegrator Bridge (QIB) enabling spatially adaptive fusion maintaining consistent predictions regardless of available modalities, (2) Multi-Granular Attention Orchestrator (MGAO) focusing on pathological regions to reduce background sensitivity, and (3) Modality Quality-Aware Enhancement (MQAE) preventing error propagation from corrupted sequences. On BraTS 2018, our method achieves 89.33% WT, 82.70% TC, 67.23% ET Dice scores with <0.5% variance across 15 modality combinations, solving the stability crisis. Single-modality ET segmentation shows 40-81% relative improvements over state-of-the-art methods. The method generalizes to BraTS 2020/2021, achieving up to 92.44% WT, 89.91% TC, 84.57% ET. The model demonstrates potential for clinical deployment with 1.2s inference. Code: https://github.com/guochengxiangives/AMGFormer.

eess.IV↗

arXiv2023

Cell nucleus elastography with the adjoint-based inverse solver

Background and Objectives: The mechanics of the nucleus depends on cellular structures and architecture, and impact a number of diseases. Nuclear mechanics is yet rather complex due to heterogeneous distribution of dense heterochromatin and loose euchromatin domains, giving rise to spatially variable stiffness properties. Methods: In this study, we propose to use the adjoint-based inverse solver to identify for the first time the nonhomogeneous elastic property distribution of the nucleus. Inputs of the inverse solver are deformation fields measured with microscopic imaging in contracting cardiomyocytes. Results: The feasibility of the proposed method is first demonstrated using simulated data. Results indicate accurate identification of the assumed heterochromatin region, with a maximum relative error of less than 5%. We also investigate the influence of unknown Poisson's ratio on the reconstruction and find that variations of the Poisson's ratio in the range [0.3-0.5] result in uncertainties of less than 15% in the identified stiffness. Finally, we apply the inverse solver on actual deformation fields acquired within the nuclei of two cardiomyocytes. The obtained results are in good agreement with the density maps obtained from microscopy images. Conclusions: Overall, the proposed approach shows great potential for nuclear elastography, with promising value for emerging fields of mechanobiology and mechanogenetics.

physics.med-ph↗

arXiv2020

Study of $Ξ^-$ hypernuclei in the Skyrme-Hartree-Fock approach

The properties of $Ξ^-$ hypernuclei are studied systematically using a two-dimensional Skyrme-Hartree-Fock approach combined with three different $ΞN$ Skyrme forces fitted to reproduce the existing data. We explore the impurity effect of a single $Ξ^-$ hyperon on the radii, deformations, and density distributions of the nuclear core and point out qualitative differences between the different forces. We find that the $Ξ^-$ removal energy of $^{\hskip0.10em13}_{Ξp}$B [$^{12}$C(g.s.)+ $Ξ^-$(1p)] calculated by the SLX3 force is 0.7 MeV, which is in good agreement with a possible value of $0.82\pm0.17\;$MeV from the KEK E176 experiment. The theoretical prediction for this weakly bound state depends strongly on the deformation of the nuclear core, which is analyzed in detail.

nucl-th↗

arXiv2019

Deformed $K^-$ nuclei in the Skyrme-Hartree-Fock approach

The properties of kaonic nuclei are studied using a two-dimensional Skyrme-Hartree-Fock model with a $KN$ Skyrme force. We focus in particular on the instability of the solutions for a too strong $KN$ interaction, which determines a maximum value of the kaon binding in this approach. We then analyze the change of the deformation properties of several core-deformed nuclei caused by the added kaon, and find a shrinking of the core and in some cases a complete loss of deformation.

nucl-th↗

arXiv2015

Analyzing TCP Throughput Stability and Predictability with Implications for Adaptive Video Streaming

Recent work suggests that TCP throughput stability and predictability within a video viewing session can inform the design of better video bitrate adaptation algorithms. Despite a rich tradition of Internet measurement, however, our understanding of throughput stability and predictability is quite limited. To bridge this gap, we present a measurement study of throughput stability using a large-scale dataset from a video service provider. Drawing on this analysis, we propose a simple-but-effective prediction mechanism based on a hidden Markov model and demonstrate that it outperforms other approaches. We also show the practical implications in improving the user experience of adaptive video streaming.

cs.NI↗