Corrections to “Deep-Learning Approach for Tissue Classification Using Acoustic Waves During Ablation With an Er:YAG Laser”

Abstract

In the above article [1], we found major issues with the data we used. Specifically, we received data from [2] for five distinct tissues, with ten specimens per tissue. However, upon closer examination, we realized that the data for these specimens were not unique; rather, they were scaled variations derived from a single specimen. As a result, our training, testing, and validation datasets were not independent, leading to an artificially high accuracy rate of 100%.