arXiv subjects

Jae Sung Lee

Publications and source records attributed to Jae Sung Lee.

At least 19 recordsLinked to original sources

arXiv2026

A Specialized Large Multimodal Model for Interpreting PET/CT in Head and Neck Cancer

Background: Diagnosing head and neck cancer using PET/CT is clinically challenging and time-consuming due to the anatomical complexity of the region, motivating computer-aided diagnosis (CAD). Generalist Large Multimodal Models (LMMs) remain limited in medical contexts by insufficient domain-specific knowledge, privacy and security concerns, and verbosity, motivating specialized standalone LMMs. Purpose: We evaluated the feasibility of a specialized LMM for automated PET/CT interpretation in head and neck cancer using a large-scale multi-institutional PET/CT dataset, a tailored training curriculum, and autoregressive training. Methods: LLaVA-NeXT was fine-tuned using a two-level curriculum with image-conversation pairs curated by two radiologists from public data. The dataset included clinically important annotations such as primary tumor presence and metastatic lymph node location. Level 1 used 28,000 image-conversation pairs to learn basic information, including modality type and hypermetabolism. Level 2 used 12,975 pairs to learn primary tumor presence and the existence and anatomical location of cervical lymph node metastases. External validation included four institutions with diverse imaging devices. Results: The specialized LMM substantially outperformed ChatGPT and LLaVA-NeXT. In Level-2 external validation, ROUGE-L, ROUGE-S, Cosine Similarity, Precision, Recall, and F1 were 0.8751, 0.8794, 0.8324, 0.8794, 0.8711, and 0.8751, while generalist models consistently scored below 0.1. Primary tumor classification accuracy was 83.14 +/- 1.15% internally and 69.03 +/- 0.81% externally. For lymph node localization, the corresponding scores were 0.6389, 0.6257, 0.5287, 0.5782, 0.6371, and 0.6648. Conclusion: Specialized LMMs show promising results for fast, accurate PET/CT-based diagnostic support and medical education, highlighting their potential for clinical translation.

cs.CV↗

arXiv2026

Nonequilibrium Fluctuation-Response Theory in the Frequency Domain

We establish a unified fluctuation-response theory in the frequency domain for nonequilibrium steady states governed by overdamped Langevin dynamics and Markov jump processes. The central identity is an exact fluctuation-response relation that reconstructs the power spectrum of general observables from local responses measured at the same frequency. This relation applies to state-dependent observables, current-like observables, and their combinations, and reveals how local dynamics shape global fluctuations. From this identity, we derive a hierarchy of relations in the frequency domain, including response uncertainty relations, kinetic and thermodynamic uncertainty relations, the equilibrium fluctuation-dissipation theorem, and Harada-Sasa relations. This establishes an overarching framework for fluctuation-response theory in the frequency domain. Applications to stochastic networks and driven diffusive systems demonstrate how the theory decomposes fluctuation spectra into local contributions and reveals frequency-dependent tradeoffs between fluctuations, response, and dissipation.

cond-mat.stat-mech↗

arXiv2026

Multi-Level Analyzation of Imbalance to Resolve Non-IID-Ness in Federated Learning

Class imbalance is a common problem in deep learning that severely degrades performance. In federated learning (FL), it is a critical factor contributing to non-identically distributed data (non-IID). Building on several previous attempts, we define and analyze imbalance issues in FL at three levels: inter-case, inter-class, and inter-client. Inter-case imbalance addresses the imbalance in every single class; inter-class imbalance compares the number of data between different classes. Inter-client imbalance represents different skewness of local data between clients. Based on these concepts, we propose FedBB, which consists of two main components: (1) Positive Negative Balanced (PNB) loss function addresses the inter-case and inter-class imbalances in local training, enhancing generalization on highly skewed local client datasets. It optimizes both multi-label and multi-class classifications by assigning higher weights to minority cases or classes. (2) Client Balanced Reweighting (CBR) reweights clients based on inter-client imbalance during model aggregation, giving greater weight to models trained on less skewed datasets. Various experiments on X-ray and natural image datasets demonstrate that FedBB outperforms other algorithms in both performance and efficiency. Additionally, it requires limited statistical information, which is beneficial for privacy protection. Through ablation studies, we proved that PNB loss and CBR independently contribute to performance. As FedBB aims to build a global model that accurately classifies all classes, it can serve as a baseline for the generic and personalized FL.

cs.LG↗

arXiv2026

Towards Brain MRI Foundation Models for the Clinic: Findings from the FOMO25 Challenge

Clinical deployment of automated brain MRI analysis faces a fundamental challenge: clinical data is heterogeneous and noisy, and high-quality labels are prohibitively costly to obtain. Self-supervised learning (SSL) can address this by leveraging the vast amounts of unlabeled data produced in clinical workflows to train robust \textit{foundation models} that adapt out-of-domain with minimal supervision. However, the development of foundation models for brain MRI has been limited by small pretraining datasets and in-domain benchmarking focused on high-quality, research-grade data. To address this gap, we organized the FOMO25 challenge as a satellite event at MICCAI 2025. FOMO25 provided participants with a large pretraining dataset, FOMO60K, and evaluated models on data sourced directly from clinical workflows in few-shot and out-of-domain settings. Tasks covered infarct classification, meningioma segmentation, and brain age regression, and considered both models trained on FOMO60K (method track) and any data (open track). Nineteen foundation models from sixteen teams were evaluated using a standardized containerized pipeline. Results show that (a) self-supervised pretraining improves generalization on clinical data under domain shift, with the strongest models trained \textit{out-of-domain} surpassing supervised baselines trained \textit{in-domain}. (b) No single pretraining objective benefits all tasks: MAE favors segmentation, hybrid reconstruction-contrastive objectives favor classification, and (c) strong performance was achieved by small pretrained models, and improvements from scaling model size and training duration did not yield reliable benefits.

cs.CV↗

arXiv2026

Fluctuation-Response Theory for Nonequilibrium Langevin Dynamics

We establish a unified fluctuation-response relation for Langevin dynamics. By exploiting the common mathematical structures underlying fluctuations and responses of empirical density and current, we derive a unified identity that generalizes the fluctuation-dissipation theorem from equilibrium to nonequilibrium settings. This relation connects global fluctuations of observables with their local responses to perturbations in force, mobility, and temperature. We further derive finite-time fluctuation-response inequalities, leading to response uncertainty relations that complement the identity by providing more practical bounds. These derivations establish a unified theoretical framework linking the fluctuation-dissipation theorem and thermodynamic uncertainty relations. Using the $F_1$-ATPase molecular motor model, we illustrate how these response-based bounds constrain the long-time diffusion coefficient.

cond-mat.stat-mech↗

arXiv2025

FedPOD: the deployable units of training for federated learning

This paper proposes FedPOD, which ranked first in the 2024 Federated Tumor Segmentation (FeTS) Challenge, for optimizing learning efficiency and communication cost in federated learning among multiple clients. Inspired by FedPIDAvg, we define a round-wise task for FedPOD to enhance training efficiency. FedPIDAvg achieved performance improvement by incorporating the training loss reduction for prediction entropy as weights using differential terms. Furthermore, by modeling data distribution with a Poisson distribution and using a PID controller, it reduced communication costs even in skewed data distribution. However, excluding participants classified as outliers based on the Poisson distribution can limit data utilization. Additionally, PID controller requires the same participants to be maintained throughout the federated learning process as it uses previous rounds' learning information in the current round. In our approach, FedPOD addresses these issues by including participants excluded as outliers, eliminating dependency on previous rounds' learning information, and applying a method for calculating validation loss at each round. In this challenge, FedPOD presents comparable performance to FedPIDAvg in metrics of Dice score, 0.78, 0.71 and 0.72 for WT, ET and TC in average, and projected convergence score, 0.74 in average. Furthermore, the concept of FedPOD draws inspiration from Kubernetes' smallest computing unit, POD, designed to be compatible with Kubernetes auto-scaling. Extending round-wise tasks of FedPOD to POD units allows flexible design by applying scale-out similar to Kubernetes' auto-scaling. This work demonstrated the potentials of FedPOD to enhance federated learning by improving efficiency, flexibility, and performance in metrics.

cs.CV↗

arXiv2025

Patlak Parametric Image Estimation from Dynamic PET Using Diffusion Model Prior

Dynamic PET enables the quantitative estimation of physiology-related parameters and is widely utilized in research and increasingly adopted in clinical settings. Parametric imaging in dynamic PET requires kinetic modeling to estimate voxel-wise physiological parameters based on specific kinetic models. However, parametric images estimated through kinetic model fitting often suffer from low image quality due to the inherently ill-posed nature of the fitting process and the limited counts resulting from non-continuous data acquisition across multiple bed positions in whole-body PET. In this work, we proposed a diffusion model-based kinetic modeling framework for parametric image estimation, using the Patlak model as an example. The score function of the diffusion model was pre-trained on static total-body PET images and served as a prior for both Patlak slope and intercept images by leveraging their patch-wise similarity. During inference, the kinetic model was incorporated as a data-consistency constraint to guide the parametric image estimation. The proposed framework was evaluated on total-body dynamic PET datasets with different dose levels, demonstrating the feasibility and promising performance of the proposed framework in improving parametric image quality.

eess.IV↗

arXiv2025

Memory-aware feedback enhances power in active information engines

We study an information engine operating in an active bath, where a Brownian particle confined in a harmonic trap undergoes feedback-driven displacement cycles. Unlike thermal environments, active baths exhibit temporally correlated fluctuations, introducing memory effects that challenge conventional feedback strategies. Extending the framework of stochastic thermodynamics to account for such memory, we analyze a feedback protocol that periodically shifts the potential minimum based on noisy measurements of the particle's position. We show that conventional feedback schemes, optimized for memoryless thermal baths, can degrade performance in active media due to the disruption of bath-particle memory by abrupt resetting. To overcome this degradation, we introduce a class of memory-preserving feedback protocols that partially retain the covariance between the particle's displacement and active noise, thereby exploiting the temporal persistence of active fluctuations. Through asymptotic analysis, we show how the feedback gain -- which quantifies the strength of positional shifts -- nontrivially shapes the engine's work and power profiles. In particular, we demonstrate that in active media, intermediate gains outperform full-shift resetting. Our results reveal the critical interplay between bath memory, measurement noise, and feedback gain, offering guiding principles for designing high-performance information engines in nonequilibrium environments.

cond-mat.stat-mech↗

arXiv2025

Investigating the Timing Behavior of Compton Scattering in BGO for Time-of-Flight PET

Bismuth germanate (BGO) is gaining renewed attention as a viable material for hybrid Cherenkov/scintillation time-of-flight positron emission tomography (TOF-PET) detectors. While single-crystal studies have demonstrated excellent timing resolution by leveraging prompt Cherenkov photons, practical detector modules based on pixelated arrays introduce a high prevalence of inter-crystal scattering (InterCS) events, complicating timing accuracy. In this study, we experimentally investigated the impact of InterCS on BGO Cherenkov timing using a dual-pixel detector coupled to a segmented SiPM readout. Events were classified into full-energy deposition (FED; primary crystal 511 keV absorption), InterCS, and penetration types via energy-weighted positioning and validated using GATE simulations, which also revealed that over 25% of the experimentally identified full-energy events involved intra-crystal scatter (IntraCS). For InterCS events, the optimal timestamp selection was achieved by choosing the earlier of the two timestamps, yielding a coincidence timing resolution of 221 ps FWHM (831 ps FWTM)-markedly worse than the 184 ps (603 ps FWTM) obtained for FED events. Furthermore, prompt photon yield was found to decrease measurably due to energy splitting: InterCS events averaged 4.73 detected photons in the first 1 ns, compared to 5.76 for FED events. These results emphasize the importance of incorporating time-aware, per-pixel timestamping strategies in pixelated BGO TOF-PET systems to maintain optimal timing performance in the presence of scatter.

physics.ins-det↗

arXiv2025

GPDM: Generation-Prior Diffusion Model for Accelerated Direct Attenuation and Scatter Correction of Whole-body 18F-FDG PET

Accurate attenuation and scatter corrections are crucial in positron emission tomography (PET) imaging for accurate visual interpretation and quantitative analysis. Traditional methods relying on computed tomography (CT) or magnetic resonance imaging (MRI) have limitations in accuracy, radiation exposure, and applicability. Deep neural networks provide potential approaches to estimating attenuation and scatter-corrected (ASC) PET from non-attenuation and non-scatter-corrected (NASC) PET images based on VAE or CycleGAN. However, the limitations inherent to conventional GAN-based methods, such as unstable training and mode collapse, need further advancements. To address these limitations and achieve more accurate attenuation and scatter corrections, we propose a novel framework for generating high-quality ASC PET images from NASC PET images: Generation-Prior Diffusion Model (GPDM). Our GPDM framework is based on the Denoising Diffusion Probabilistic Model (DDPM), but instead of starting sampling from an entirely different image distribution, it begins from a distribution similar to the target images we aim to generate. This similar distribution is referred to as the Generation-Prior. By leveraging this Generation-Prior, the GPDM framework effectively reduces the number of sampling steps and generates more refined ASC PET images. Our experimental results demonstrate that GPDM outperforms existing methods in generating ASC PET images, achieving superior accuracy while significantly reducing sampling time. These findings highlight the potential of GPDM to address the limitations of conventional methods and establish a new standard for efficient and accurate attenuation and scatter correction in PET imaging.

physics.med-ph↗

arXiv2025

Coherence enhanced by detrained oscillators: Breaking $π$-reflection symmetry

We study a generalized Kuramoto model in which each oscillator carries two coupled phase variables, representing a minimal swarmalator system. Assuming perfect correlation between the intrinsic frequencies associated with each phase variable, we identify a novel dynamic mode characterized by bounded oscillatory motion that breaks the $π$-reflection symmetry. This symmetry breaking enhances global coherence and gives rise to a non-trivial mixed state, marked by distinct degrees of ordering in each variable. Numerical simulations confirm our analytic predictions for the full phase diagram, including the nature of transition. Our results reveal a fundamental mechanism through which detrained (dynamic) oscillators can promote global synchronization, offering broad insights into coupled dynamical systems beyond the classical Kuramoto paradigm.

cond-mat.stat-mech↗

arXiv2025

Violation of kinetic uncertainty relation in maser heat engines: Role of spontaneous emission

We investigate the kinetic uncertainty relation (KUR)-a fundamental trade-off between dynamical activity and current fluctuations-in two configurations of a maser heat engine. We find that KUR violations arise only in one model. This asymmetry originates from spontaneous emission, which breaks the structural symmetry between the configurations and modifies their coherence dynamics. While we analyze several contributing factors-including statistical signatures such as the Fano factor and the ratio of dynamical activity to current-our results show that the decisive mechanism is the slower decoherence in one configuration, which enables quantum violations of the classical steady-state KUR bound. By contrast, the faster coherence decay in the other configuration suppresses such violations, driving it closer to classical behavior. These findings highlight the critical role of decoherence mechanisms in determining fundamental thermodynamic bounds and provide insights for the design of quantum heat engines in which the control of decoherence is central to suppressing fluctuations and enhancing reliable performance.

quant-ph↗

arXiv2025

Discrete-time thermodynamic speed limit

As a fundamental thermodynamic principle, speed limits reveal the lower bound of entropy production (EP) required for a system to transition from a given initial state to a final state. While various speed limits have been developed for continuous-time Markov processes, their application to discrete-time Markov chains remains unexplored. In this study, we investigate the speed limits in discrete-time Markov chains, focusing on two types of EP commonly used to measure the irreversibility of a discrete-time process: time-reversed EP and time-backward EP. We find that time-reversed EP satisfies the speed limit for the continuous-time Markov processes, whereas time-backward EP does not. Additionally, for time-reversed EP, we derive practical speed limits applicable to systems driven by cyclic protocols or with unidirectional transitions, where conventional speed limits become meaningless or invalid. We show that these relations also hold for continuous-time Markov processes by taking the time-continuum limit of our results. Finally, we validate our findings through several examples.

cond-mat.stat-mech↗

arXiv2025

Thermodynamic uncertainty relation for systems with active Ornstein-Uhlenbeck particles

Thermodynamic uncertainty relations (TURs) delineate tradeoff relations between the thermodynamic cost and the magnitude of an observable's fluctuation. While TURs have been established for various nonequilibrium systems, their applicability to systems influenced by active noise remains largely unexplored. Here, we present an explicit expression of TUR for systems with active Ornstein-Uhlenbeck particles (AOUPs). Our findings reveal that active noise introduces modifications to the terms associated with the thermodynamic cost in the TUR expression. The altered thermodynamic cost encompasses not only the conventional entropy production but also the energy consumption induced by the active noise. We examine the capability of this TUR as an accurate estimator of the extent of anomalous diffusion in systems with active noise driven by a constant force in free space. By introducing the concept of a contracted probability density function, we derive a steady-state TUR tailored to this system. Moreover, through the adoption of a new scaling parameter, we enhance and optimize the TUR bound further. Our results demonstrate that active noise tends to hinder accurate estimation of the anomalous diffusion extent. Our study offers a systematic approach for exploring the fluctuation nature of biological systems operating in active environments.

cond-mat.stat-mech↗

arXiv2025

Thermodynamic anomalies in overdamped systems with time-dependent temperature

One of the key objectives in investigating small stochastic systems is the development of micrometer-sized engines and the understanding of their thermodynamics. However, the primary mathematical tool used for this purpose, the overdamped approximation, has a critical limitation: it fails to fully capture the thermodynamics when the temperature varies over time, as the velocity is not considered in the approximation. Specifically, we show that heat dissipation and entropy production calculated under the overdamped approximation deviate from their true values. These discrepancies are termed thermodynamic anomalies. To overcome this limitation, we analytically derive expressions for these anomalies in the presence of a general time-varying temperature. One notable feature of the result is that high viscosity and small mass, though both leading to the same overdamped dynamic equations, result in different thermodynamic anomaly relations. Our results have significant implications, particularly for accurately calculating the efficiency of heat engines operating in overdamped environments with time-varying temperatures, without requiring velocity measurements. Additionally, our findings offer a simple method for estimating the kinetic energy of an overdamped system.

cond-mat.stat-mech↗

arXiv2024

A unified framework for classical and quantum uncertainty relations using stochastic representations

Thermodynamic uncertainty relations (TURs) and kinetic uncertainty relations (KURs) provide tradeoff relations between measurement precision and thermodynamic cost such as entropy production and activity. Conventionally, these relations are derived using the Cramér-Rao inequality, which involves an auxiliary perturbation in deterministic differential equations governing the time evolution of the system's probability distribution. In this study, without relying on the previous formulation based on deterministic evolving equation, we demonstrate that all previously discovered uncertainty relations can be derived solely through the stochastic representation of the same dynamics. For this purpose, we propose a unified method based on stochastic representations for general Markovian dynamics. Extending beyond classical systems, we apply this method to Markovian open quantum systems by unraveling their dynamics, deriving quantum uncertainty relations that are physically more accessible and tighter in regimes where quantum effects play a significant role. This fully establishes uncertainty relations for both classical and quantum systems as intrinsic properties of their stochastic nature.

cond-mat.stat-mech↗

arXiv2024

Fluctuation-response inequalities for kinetic and entropic perturbations

We derive fluctuation-response inequalities for Markov jump processes that link the fluctuations of general observables to the response to perturbations in the transition rates within a unified framework. These inequalities are derived using the Cramér-Rao bound, enabling broader applicability compared to existing fluctuation-response relations formulated for static responses of current-like observables. The fluctuation-response inequalities are valid for a wider class of observables and are applicable to finite observation times through dynamic responses. Furthermore, we extend these inequalities to open quantum systems governed by the Lindblad quantum master equation and find the quantum fluctuation-response inequality, where dynamical activity plays a central role.

cond-mat.stat-mech↗

arXiv2024

Segmented readout for Cherenkov time-of-flight positron emission tomography detectors based on bismuth germanate

Positron emission tomography (PET) is the most sensitive biomedical imaging modality for non-invasively detecting and visualizing positron-emitting radiopharmaceuticals within a subject. In PET, measuring the time-of-flight (TOF) information for each pair of 511-keV annihilation photons improves effective sensitivity but requires high timing resolution. Hybrid materials that emit both scintillation and Cherenkov photons, such as bismuth germanate (BGO), recently offer the potential for more precise timing information from Cherenkov photons while maintaining adequate energy resolution from scintillation photons. However, a significant challenge in using such hybrid materials for TOF PET applications lies in the event-dependent timing spread caused by the mixed detection of Cherenkov and scintillation photons due to relatively lower production of Cherenkov photons. This study introduces an innovative approach by segmenting silicon photomultiplier (SiPM) pixels coupled to a single crystal, rather than using traditional SiPMs that are as large as or larger than the crystals they read. We demonstrated that multiple time stamps and photon counts obtained from the segmented SiPM can classify events by providing temporal photon density, effectively addressing this challenge. The approach and findings would lead to new opportunities in applications that require precise timing and photon counting, spanning the fields of medical imaging, high-energy physics, and optical physics.

physics.ins-det↗