This paper presents the design and implementation of a clinical workstation software for analyzing voice disorders. The software is developed by using Java technology and MySQL database system. A variety of vocal cues, e.g. jitter and shimmer, that describe irregularities of speech cycles in sustained vowels can be automatically derived by the system. For assessing voice disorders in connected speech, a vocal cue called signal-to-dysperiodicity ratio is evaluated by carrying out a generalized variogram analysis. In the development, special attention has been paid to software engineering conventions and the principles of architectural design of software structures to achieve good quality attributes such as developmental simplicity and modifiability. Preliminary tests have shown that the system provides satisfactory usability and performance for clinical applications.

Index Terms. Pathological voice assessment, disordered voice analysis, software engineering, Java application