IEEE TCSVT Special Issue 2021 : Learning with Multimodal Data for Biomedical Informatics

Fast-growing biomedical and healthcare data of multiple modalities have encompassed multiple scales ranging from molecules, individuals, to populations. Meanwhile, the heterogeneous and increasingly more diverse modalities of the data present major barriers toward their understanding, fusion, and translation into effective clinical actions. For example, electronic health records (EHRs) are representative examples of multimodal/multisource data collections; including not only traditional medical measurements, but also images, videos, audios, and free texts. Other examples include mobile health for remote patient care with typical data modalities such as patient- or caregiver-generated photos, self-reported symptoms of pain, and body temperature. The diversity of such information sources and the increasing amounts of medical data produced by healthcare institutes annually, pose significant challenges for data-driven biomedical analysis. While biomedical and healthcare research traditionally focuses on the structured measurement data, the growing availability of novel data modalities has created a compelling demand for novel machine learning, image/video/audio/text processing, and multi-modal fusion algorithms that specifically tackle the unique challenges associated with biomedical and healthcare data and allow decision-makers and stakeholders to better interpret and exploit the data. This special issue aims at catalyzing synergies among image/video processing, text/speech understanding, machine learning, multi-modal learning and other related fields with the goals to (1) develop novel data-driven models to accelerate knowledge discovery in biomedicine through the seamless integration of medical data collected from imaging systems, laboratory and wearable devices, as well as other related medical devices; (2) promote the development of new multi-modal learning systems to enhance the healthcare quality and patient safety; and (3) promote new applications in biomedical informatics that can leverage or benefits from the integration of multi-modal data and machine learning.