Skip to main content
. 2022 Jun 21;62(2):153–157. doi: 10.2169/internalmedicine.9592-22

Figure.

Figure.

Speech recognition (SR) system workflow. All words spoken by the endoscopist were displayed in the “All detected words” section at the bottom of the screen. The recognized findings were first displayed in the “Pre-registration findings” section at the upper part of the screen (a). When the proper findings were displayed, the endoscopist used the command “Register” to finalize the findings in the center of the screen as “Registration findings” (b).