Presenter: Hau-Tieng Wu NYU
Title: Toward high quality biorhythm deciphering in digital health: Time-Frequency Analysis and Statistical Inference
Abstract: With the proliferation of wearable devices and other technologies, physicians can now easily access multimodal physiological waveforms spanning extensive timeframes. Yet, the signal quality of these datasets can pose a serious challenge for healthcare providers and medical researchers, particularly when they require reliable, useful and actionable clinical information. In this talk, I will discuss recent progress in nonstationary time series analysis aimed at addressing these challenges. Specifically, I will focus on the development of time-frequency analysis as a key tool for extracting relevant information from these complex datasets, while also reconsidering the meaning of signal quality, particularly in digital health. Some clinical examples will be provided to illustrate the practical application of these methodologies.