We have investigated an optimal sensing platform that can be used to monitor the sleep quality and sleep abnormalities for at-home sleep disorder monitoring and diagnosis in cancer patient. As an alternative to a traditional Polysomnography system at the sleep lab, the proposed multisensory suit was embedded with a nonlinear signal processing algorithm and decision-rule-based machine learning model to provide at-home diagnostics of obstructive sleep apnea (OSA), a very common sleep breathing disorder. An ongoing extension of this system, designed as a wearable device with the added cardiorespiratory monitoring functions, has been developed for the screening and diagnosis of central sleep apnea, nocturnal asthma and cardiovascular-related diseases. Research in this direction has produced 3 so far patent applications and ongoing translational research projects within the industry

We have introduced a new method for MI localization based on capturing the complex cardiac excitation and propagation dynamics as a random walk network reconstructed from VCG signals. Various topological and dynamic quantifiers of the random walk network were found to be sensitive to the location of cardiac tissues damaged as a result of an MI and, consequently, they serve as effective features for MI localization. The results suggest that the network measures identified in the present work can be used to characterize the propagation patterns of the cardiac action potential during the depolarization (roughly contemporaneous with heart contraction and manifestation of P and QRS loops in VCG) and repolarization (contemporaneous with heart relaxation and manifestation of T loop) phases for accurate MI localization.