Papers9月11日 05:23

电子健康记录中罕见疾病挖掘的代理驱动方法

RDMA: cost effective agent-driven rare disease mining from electronic health records

npj Digital Medicine@Adam Cross

Blog/Media

Read the original

Why we picked it

朋友，有个新方法叫RDMA，专门从电子健康记录里找罕见病，比传统方法更省钱还更准，值得看看。

这项研究提出了一种名为RDMA的新方法，用于从电子健康记录中挖掘罕见疾病信息。该方法通过代理驱动的方式，能够以较低成本高效地处理数据。

Body

在MIMIC-III数据集上，RDMA在准确性和召回率上均优于传统方法。该研究为利用AI进行医疗数据分析提供了新的思路。

Original · npj Digital Medicine

RDMA: cost effective agent-driven rare disease mining from electronic health records

npj Digital Medicine, Published online: 10 September 2026; doi:10.1038/s41746-026-03070-x RDMA: cost effective agent-driven rare disease mining from electronic health records

Industry Coordinates & Source Verification

This report is categorized under AITOP's "Papers" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Key entities and technical topics under continuous tracking include RDMA, 电子健康记录, 罕见疾病, 医疗AI.

Source attribution: curated from npj Digital Medicine (@Adam Cross).

#RDMA #电子健康记录 #罕见疾病 #医疗AI

Keep reading

Recent stories with the same tag or section

Papers Builders10:09

Chinese contentARK：为分离式 LLM 推理中的 KV Cache 传输解决路由冲突

KV Cache 传输是网络里的巨流，ECMP 哈希一撞就堵。ARK 只调源端口就能避开，FCT 降了 27.3%，做推理基础设施的可以看看。

#KV Cache #LLM 推理 #RDMA #ECMP

arXiv cs.AI@Hung-Chun Lin et al. (4)Source

Papers BuildersSelected09:50

TIDE Speeds FSDP2 Training with Zero-Copy Collectives on 64 GPUs

The hidden cost in FSDP training is often layout conversion and collective interference rather than compute, and TIDE shows both can be largely eliminated with standard NCCL APIs rather than custom hardware. Teams running large-model training on multi-node clusters can adopt it to reclaim step time and GPU memory without changing FSDP2's per-parameter kernels. The mixed-precision bf16-to-fp32 reduce-scatter is a capability stock NCCL lacks, making the implementation details directly reusable.

#FSDP2 #NCCL #TIDE #distributed training

arXiv cs.LG@Zejian Wang et al. (14)Source

How-to BuildersOfficial01:15

Chinese content用 MLX 与 RDMA 在多台 Mac Studio 上跑分布式大模型推理

Apple 教你用几台 Mac Studio 加 RDMA 组成本地推理集群，跑单机放不下的大模型，教程可以直接照着做。

#MLX #Mac Studio #RDMA #分布式推理

Primary sourceApple ML ResearchSource

Products BuildersSelected04:18

AMD team to show MoRI plus vLLM for cross-pod MoE serving

Serving MoE models across pods usually bottlenecks on expert dispatch and KV-cache movement, and this poster shows an RDMA-based path built on vLLM rather than a proprietary stack. Infrastructure and inference engineers evaluating AMD Instinct for large-scale serving can use the benchmarks to compare it against NVIDIA-based disaggregation setups. Registration for the 2026 conference is open via the linked page.

#AMD #vLLM #MoRI #Mixture-of-Experts