Fig. 5: Forest plots showing meta-analyses of the reviewed literature.
From: Large language model integrations in cancer decision-making: a systematic review and meta-analysis
The figure presents LLM accuracy rates found across different study categories, with the circle sizes indicating sample size. a Shows findings for “general accuracy” of LLMs across all included meta-analysis studies, with sample sizes ranging from 20 to 131. b Focuses on “general accuracy” rates for studies that specifically examined cancer diagnosis, where the sample sizes ranged from 10 to 117. c Illustrates the “general accuracy” rates for studies centered on treatment planning, where the sample sizes ranged from 3 to 109. The red dotted lines show the mean values.
