Comparative Evaluation of Large Language and Multimodal Models in Detecting Spinal Stabilization Systems on X-Ray Images.

Similarly, for the detection of non-growing rod systems, GPT-4o exhibited near-perfect sensitivity (M = 0.98, SE = 0.00), significantly surpassing BiomedClip (M = 0.23, SE = 0.02), as reflected in a highly significant statistical difference (z = 37.5, p < 0.001).

also in 132,142 other papers

For MCGR systems, GPT-4o (M = 0.18, SE = 0.10) outperformed BiomedClip (M = 0.04, SE = 0.02), although the difference did not reach statistical significance (z = 1.40, p = 0.160), suggesting a trend toward an improved performance.

also in 111,027 other papers