A Cloud-Edge System for Multimodal Clinical Screening in Resource-Constrained Rural Settings

2026-08-12

AI Summary

This study addresses the challenge of deploying high-accuracy multimodal medical AI in rural areas, where limited bandwidth and constrained computational resources hinder conventional approaches. To overcome this, the authors propose a cloud-edge协同 architecture wherein lightweight, domain-specific models at the edge convert raw medical data into structured outputs, while a large language model (LLM) in the cloud fuses these multimodal results to generate clinical summaries. An LLM-driven scheduler dynamically selects diagnostic tools based on patient context, enabling bandwidth-adaptive and computationally efficient multimodal integration with factual enhancement. Under simulated rural network conditions, the system achieves diagnostic recall of 98–99% and precision of 92–96%, matching or exceeding the clinical accuracy of purely cloud-based solutions, with stable end-to-end latency of 25–35 seconds and a 4–15× reduction in token cost.

Abstract

Medical AI has demonstrated specialist-level diagnostic accuracy, yet these capabilities remain largely inaccessible in resource-constrained rural settings where bandwidth is scarce, compute is limited, and clinical decision-making requires integrating heterogeneous modalities. We introduce a cloud--edge collaborative architecture that addresses these constraints: lightweight, domain-specific models on the edge transform raw medical data into compact structured outputs, while a cloud LLM synthesizes these outputs into clinical summaries. An LLM-based orchestrator dynamically selects diagnostic tools based on patient context, promoting comprehensive modality coverage without processing irrelevant inputs. We evaluate on 20 multimodal clinical cases spanning cardiac, obstetric, trauma, and screening scenarios under three simulated network profiles (500,kbps--5,Mbps). The hybrid system achieves 98--99% diagnostic tool recall with 92--96% precision, matches or exceeds cloud-only baselines on clinical accuracy, and maintains bandwidth-invariant latency (25--35,s) at 4--15x lower token cost. These results highlight the role of architectural design in enabling efficient multimodal integration and improving factual grounding compared to cloud-only approaches under deployment constraints.

Problem

• multimodal clinical screening
• resource-constrained settings
• cloud-edge system
• medical AI deployment
• heterogeneous modalities

Innovation

• cloud-edge collaboration
• multimodal clinical screening
• lightweight edge models
• LLM-based orchestrator
• resource-constrained deployment