"From Cancer Diagnosis to Treatment Design in Just One Day"... LG AI Research Institute Unveils 'Cancer Agentic AI'

Research Achievements Unveiled at the American Association for Cancer Research
Joint Research with Vanderbilt University Medical Center
AI Agent Collaboration Based on EXAONE Path
Analyzing Cancer Tissues to Validate Drugs and Design Treatment Strategies

LG AI Research Institute has successfully developed "Cancer Agentic AI," an artificial intelligence system in which multiple AI agents collaborate to design a comprehensive cancer diagnostic, analysis, and treatment plan—all within a single day. Notably, the system establishes a foundation for reducing unnecessary tests for patients and enables the early identification of patient groups eligible for drug therapies.

LG AI Research Institute and Vanderbilt University Medical Center in the United States have unveiled their joint research achievements on "Cancer Agentic AI" at the American Association for Cancer Research (AACR) 2026, held in San Diego, U.S., from April 17 to 22 (local time).

This technology is characterized by its ability to complete the entire process—from tissue analysis to treatment strategy design—within a single day. The starting point is a single pathology image, analyzed by the "EXAONE Path" pathology AI, which predicts cancer gene activity within the tissue in less than one minute. LG AI Research Institute has enhanced the predictive accuracy of this technology, thereby reducing unnecessary testing and facilitating the early identification of patients suitable for targeted drug therapies. In July of last year, LG AI Research Institute announced plans to collaborate with Professor Hwang Tae-hyun’s research team to develop a multimodal medical AI platform that enables personalized precision medicine by advancing therapeutic effect prediction technologies.

LG's EXAONE-Based Cancer Agentic AI: A Real Patient Case Study. LG Corporation.

Cancer Agentic AI operates as a collaborative structure of multiple AI agents, built upon LG’s ExaONE and cancer pathology-specialized AI technologies. Each AI agent is responsible for one of the following stages: analyzing cancer tissue images; identifying the location and activation information of cancer genes within tissue; cross-verifying and validating AI prediction results against actual measurements; verifying and evaluating candidate drug responses; designing treatment strategies; and supporting final clinical decision-making for cancer treatment.

The system is structured as a cycle of cognition, reasoning, planning, and execution, with each result passed sequentially to the next agent. To establish a safe and reliable personalized treatment system, the research team incorporated a safety mechanism that enables sharing and validation of opinions between medical professionals and AI agents at each clinical decision-making step. Medical professionals make critical decisions in stages such as reviewing patient history, comparing predictions with actual data, validating drug responses, and finalizing treatment plans, collaborating closely with AI. At the same time, the AI checks for safety, guideline compliance, discrepancies with real-world data, and correlations in drug responses, summarizing areas of uncertainty and providing explanations to clinicians.

Jongseong Jang, Head of Bio Intelligence Lab at LG AI Research Institute, stated, “LG is developing a ‘brain’ that enables AI agents to collaborate with specialist clinicians, transforming personalized anticancer therapy—reducing the period from diagnosis to treatment decision, which previously took an average of over four weeks, to just one day, thereby helping cancer patients receive treatment within the golden time.”

LG AI Research Institute and Professor Hwang Tae-hyun’s research team plan to expand the application of agentic AI from gastric cancer to a broader range of cancers, including colorectal and lung cancer. Additionally, both parties intend to introduce the ExaONE-based cancer research methodology and AI agent clinical applications to global pharmaceutical companies and university hospitals, fostering further collaboration discussions.

This content was produced with the assistance of AI translation services.