Publications by authors named "Jen-Huo Wang"
An Adaptable Continuous Restricted Boltzmann Machine in VLSI for Fusing the Sensory Data of an Electronic Nose.
IEEE Trans Neural Netw Learn Syst
April 2017
An embedded system capable of fusing sensory data is demanded for many portable or implantable microsystems. The continuous restricted Boltzmann machine (CRBM) is a probabilistic neural network not only capable of classifying data reliably but also amenable to very-large-scale-integration (VLSI) implementation. Although the embedded system based on the CRBM has been demonstrated with analog VLSI, the precision required by the learning algorithm is hardly achievable with analog circuits.
View Article and Find Full Text PDFA fully integrated nose-on-a-chip for rapid diagnosis of ventilator-associated pneumonia.
IEEE Trans Biomed Circuits Syst
December 2014
Article Synopsis
- A study addresses the ongoing challenge of rapid diagnosis of ventilator-associated pneumonia (VAP) by proposing a "nose-on-a-chip" device to detect pneumonia metabolites in patients early on.
- The chip is built using advanced 90-nm CMOS technology and includes multiple sensors and a low-power processing unit, enabling effective detection in a compact form.
- The experimental results demonstrate a high identification rate of 94.06% for VAP, with perfect accuracy in recognizing specific bacteria and fungi, presenting a significant advancement in VAP diagnostics.