arXiv subjects

Jonathan Feng

Publications and source records attributed to Jonathan Feng.

3 recordsLinked to original sources

arXiv2026

A multitask framework for automated interpretation of multi-frame right upper quadrant ultrasound in clinical decision support

Ultrasound is a cornerstone of emergency and hepatobiliary imaging, yet its interpretation remains highly operator-dependent and time-sensitive. Here, we present a multitask vision-language agent (VLM) developed to assist with comprehensive right upper quadrant (RUQ) ultrasound interpretation across the full diagnostic workflow. The system was trained on a large, multi-center dataset comprising a primary cohort from Johns Hopkins Medical Institutions (9,189 cases, 594,099 images) and externally validated on cohorts from Stanford University (108 cases, 3,240 images) and a major Chinese medical center (257 cases, 3,178 images). Built on the Qwen2.5-VL-7B architecture, the agent integrates frame-level visual understanding with report-grounded language reasoning to perform three tasks: (i) classification of 18 hepatobiliary and gallbladder conditions, (ii) generation of clinically coherent diagnostic reports, and (iii) surgical decision support based on ultrasound findings and clinical data. The model achieved high diagnostic accuracy across all tasks, generated reports that were indistinguishable from expert-written versions in blinded evaluations, and demonstrated superior factual accuracy and information density on content-based metrics. The agent further identified patients requiring cholecystectomy with high precision, supporting real-time decision-making. These results highlight the potential of generalist vision-language models to improve diagnostic consistency, reporting efficiency, and surgical triage in real-world ultrasound practice.

eess.IV↗

arXiv2013

Dark Matter Detection in Focus Point Supersymmetry

We determine the prospects for direct and indirect detection of thermal relic neutralinos in supersymmetric theories with multi-TeV squarks and sleptons. We consider the concrete example of the focus point region of minimal supergravity, but our results are generically valid for all models with decoupled scalars and mixed Bino-Higgsino or Higgsino-like dark matter. We determine the parameter space consistent with a 125 GeV Higgs boson including 3-loop corrections in the calculation of the Higgs mass. These corrections increase m_h by 1-3 GeV, lowering the preferred scalar mass scale and decreasing the fine-tuning measure in these scenarios. We then systematically examine prospects for dark matter direct and indirect detection. Direct detection constraints do not exclude these models, especially for μ< 0. At the same time, the scenario generically predicts spin-independent signals just beyond current bounds. We also consider indirect detection with neutrinos, gamma rays, anti-protons, and anti-deuterons. Current IceCube neutrino constraints are competitive with direct detection, implying bright prospects for complementary searches with both direct and indirect detection.

hep-ph↗

arXiv1993

Squark Mass Determination at the Next Generation of Linear $e^+e^-$ Colliders

Current mass limits allow the possibility that squarks may be produced in large numbers at the next generation of linear $e^+e^-$ colliders. In this paper we investigate the prospects for precision studies of squark masses at such colliders. We assume that squarks are lighter than gluinos, and discuss both direct and cascade decay scenarios. By exploiting the clean environment and polarizable beams of linear $e^+e^-$ colliders, we find that squark mass determinations at the level of a few GeV are possible in a large part of the parameter space.

hep-ph↗