Current clinical diagnosis of diabetes yields high medical costs and inconvenience, particularly for the elderly. The need for alternative solutions to alleviate these factors paved the way for the development of our revolutionary work. This innovation has the capacity to transform diabetes diagnosis, leading to a notable reduction in the substantial expenses linked to conducting multiple tests on patients.
In our initial research, we discovered notable distinctions in voice characteristics between individuals with diabetes and those without the condition. Specifically, patients with diabetes exhibited higher pitch, increased shimmer, increased Cepstral Peak Prominence (CPP), increased jitter, and a lower Harmonic-to-Noise Ratio (HNR) compared to non-diabetic individuals.