arXiv subjects

Yuang Yao

Publications and source records attributed to Yuang Yao.

5 recordsLinked to original sources

arXiv2026

READY or Not: Reliable Enterprise Agent Deployment

An AI agent can perform well on benchmarks and still be unsuitable for deployment. Existing AI-agent benchmarks measure whether an agent can complete realistic professional work, whereas enterprise deployment asks a different question: whether an agent can meet a required reliability level, under acceptable human oversight, and at tolerable cost. We introduce Reliable Enterprise Agent Deployment (READY), a framework for qualifying AI agents for deployment on enterprise workflows. READY preserves each workflow's own definition of successful execution while applying a common qualification procedure. Given an agent, a workflow, and a class of candidate oversight policies, READY measures the reliability and operating cost of the human-AI system, selects the minimum-cost policy that satisfies a specified reliability target, and statistically qualifies it on held-out cases. The resulting deployment profile characterizes the supported operating point: reliability, human-oversight burden, and cost. READY is implemented as an open testbed that decouples workflow specification, execution, evaluation, and qualification, and runs on existing agent-evaluation infrastructure. In an end-to-end clinical-audit case study spanning 16 agent systems and 750 cases, READY reveals differences hidden by autonomous performance: two systems separated by only 0.3 points in autonomous accuracy (72.8% vs. 72.5%) require 39.2% versus 29.6% human review, respectively, to qualify at the same 76% reliability target under the evaluated oversight policy. READY thus shifts enterprise agent evaluation from how well can the agent perform the work? to under what conditions, and at what cost, can it be reliably deployed? By making those conditions explicit and statistically testable, READY provides a basis for comparing agent systems, setting oversight requirements, and making evidence-based deployment decisions.

cs.AI↗

arXiv2025

Bridging the Gap in Ophthalmic AI: MM-Retinal-Reason Dataset and OphthaReason Model toward Dynamic Multimodal Reasoning

Multimodal large language models (MLLMs) have recently demonstrated remarkable reasoning abilities with reinforcement learning paradigm. Although several multimodal reasoning models have been explored in the medical domain, most of them focus exclusively on basic reasoning, which refers to shallow inference based on visual feature matching. However, real-world clinical diagnosis extends beyond basic reasoning, demanding reasoning processes that integrate heterogeneous clinical information (such as chief complaints and medical history) with multimodal medical imaging data. To bridge this gap, we introduce MM-Retinal-Reason, the first ophthalmic multimodal dataset with the full spectrum of perception and reasoning. It encompasses both basic reasoning tasks and complex reasoning tasks, aiming to enhance visual-centric fundamental reasoning capabilities and emulate realistic clinical thinking patterns. Building upon MM-Retinal-Reason, we propose OphthaReason, the first ophthalmology-specific multimodal reasoning model with step-by-step reasoning traces. To enable flexible adaptation to both basic and complex reasoning tasks, we specifically design a novel method called Uncertainty-Aware Dynamic Thinking (UADT), which estimates sample-level uncertainty via entropy and dynamically modulates the model's exploration depth using a shaped advantage mechanism. Comprehensive experiments demonstrate that our model achieves state-of-the-art performance on both basic and complex reasoning tasks, outperforming general-purpose MLLMs, medical MLLMs, RL-based medical MLLMs, and ophthalmic MLLMs by at least 24.92\%, 15.00\%, 21.20\%, and 17.66\%. Project Page: \href{https://github.com/lxirich/OphthaReason}{link}.

cs.AI↗

arXiv2025

Continual Retinal Vision-Language Pre-training upon Incremental Imaging Modalities

Traditional fundus image analysis models focus on single-modal tasks, ignoring fundus modality complementarity, which limits their versatility. Recently, retinal foundation models have emerged, but most still remain modality-specific. Integrating multiple fundus imaging modalities into a single foundation model is valuable. However, in dynamic environments, data from different modalities often arrive incrementally, necessitating continual pre-training. To address this, we propose RetCoP, the first continual vision-language pre-training framework in the fundus domain, which incrementally integrates image and text features from different imaging modalities into a single unified foundation model. To mitigate catastrophic forgetting in continual pre-training, we introduce a rehearsal strategy utilizing representative image-text pairs and an off-diagonal information distillation approach. The former allows the model to revisit knowledge from previous stages, while the latter explicitly preserves the alignment between image and text representations. Experiments show that RetCoP outperforms all the compared methods, achieving the best generalization and lowest forgetting rate. The code can be found at https://github.com/Yuang-Yao/RetCoP.

cs.CV↗

arXiv2017

A Data Colocation Grid Framework for Big Data Medical Image Processing - Backend Design

When processing large medical imaging studies, adopting high performance grid computing resources rapidly becomes important. We recently presented a "medical image processing-as-a-service" grid framework that offers promise in utilizing the Apache Hadoop ecosystem and HBase for data colocation by moving computation close to medical image storage. However, the framework has not yet proven to be easy to use in a heterogeneous hardware environment. Furthermore, the system has not yet validated when considering variety of multi-level analysis in medical imaging. Our target criteria are (1) improving the framework's performance in a heterogeneous cluster, (2) performing population based summary statistics on large datasets, and (3) introducing a table design scheme for rapid NoSQL query. In this paper, we present a backend interface application program interface design for Hadoop & HBase for Medical Image Processing. The API includes: Upload, Retrieve, Remove, Load balancer and MapReduce templates. A dataset summary statistic model is discussed and implemented by MapReduce paradigm. We introduce a HBase table scheme for fast data query to better utilize the MapReduce model. Briefly, 5153 T1 images were retrieved from a university secure database and used to empirically access an in-house grid with 224 heterogeneous CPU cores. Three empirical experiments results are presented and discussed: (1) load balancer wall-time improvement of 1.5-fold compared with a framework with built-in data allocation strategy, (2) a summary statistic model is empirically verified on grid framework and is compared with the cluster when deployed with a standard Sun Grid Engine, which reduces 8-fold of wall clock time and 14-fold of resource time, and (3) the proposed HBase table scheme improves MapReduce computation with 7 fold reduction of wall time compare with a naïve scheme when datasets are relative small.

cs.DC↗

arXiv2017

Splenomegaly Segmentation using Global Convolutional Kernels and Conditional Generative Adversarial Networks

Spleen volume estimation using automated image segmentation technique may be used to detect splenomegaly (abnormally enlarged spleen) on Magnetic Resonance Imaging (MRI) scans. In recent years, Deep Convolutional Neural Networks (DCNN) segmentation methods have demonstrated advantages for abdominal organ segmentation. However, variations in both size and shape of the spleen on MRI images may result in large false positive and false negative labeling when deploying DCNN based methods. In this paper, we propose the Splenomegaly Segmentation Network (SSNet) to address spatial variations when segmenting extraordinarily large spleens. SSNet was designed based on the framework of image-to-image conditional generative adversarial networks (cGAN). Specifically, the Global Convolutional Network (GCN) was used as the generator to reduce false negatives, while the Markovian discriminator (PatchGAN) was used to alleviate false positives. A cohort of clinically acquired 3D MRI scans (both T1 weighted and T2 weighted) from patients with splenomegaly were used to train and test the networks. The experimental results demonstrated that a mean Dice coefficient of 0.9260 and a median Dice coefficient of 0.9262 using SSNet on independently tested MRI volumes of patients with splenomegaly.

cs.CV↗