Abstract
For oral examination records using verbal input, correct recognition of the letters “A” to “E” of the deciduous dentition is important. Among the letters from “A” to “E” of the deciduous dentition, the speech recognition of “D” and “E” was difficult. Therefore, letters “A”-“E” were input as “1-(hyphen)”-“5-”(Method H). However, since “1-” was pronounced as “ichi-hyphen”, the input time was long. Therefore, to reduce the input time, “1-”-“5-” of the deciduous dentition were input as “1ban (ichiban)”-“5ban (goban)” (Method B). In addition, terms in Method B were displayed in one cell of Excel (new Method B), and the input time was compared between the two methods.
Twenty cases of various oral states were produced. In each case, input was performed using Methods H and B five times each, and the input time was compared between the two methods. The input recognition error rate was also evaluated. In addition, in each case, input was performed five times using new Method B, and the input time and input recognition errors were evaluated.
As a result, the mean input time was 16 seconds shorter using Method B (112 seconds) than using Method H (128 seconds). The input recognition error rate was similar between Methods H (0.05%) and B (0.03%). The input time using new Method B (82 seconds) was 30 seconds shorter than that using Method B. The input recognition error rate using new Method B (0.05%) was similar to that using Method B.
These results suggest the usefulness of new Method B for oral examination records of deciduous and mixed dentitions using verbal input.
