miDIAGNOSTICS, a global developer of integrated diagnostic solutions for the point-of-need, announces that its research partners The Johns Hopkins University (Baltimore, Maryland, US) and imec (Leuven, Belgium) have developed and tested novel approaches to accurately process images from imec’s holographic lens-free imaging technology. Compared to previous approaches, these advances allow fast and precise high-throughput counting of blood cells in a compact system, an application that miDIAGNOSTICS is aiming to bring to the market. The research results are now published online in the 2017 IEEE International Symposium on Biomedical Imaging (ISBI) Proceedings.
miDIAGNOSTICS’ goal is to develop breakthrough technology for rapid, reliable and low-cost diagnostic testing. The technology, with key functionalities implemented on silicon chips, will allow diagnostic testing by detecting cells, proteins, nucleic acids, and/or small molecules. A small and easily available sample, such as a drop of blood from a finger prick, will be used, allowing results to be provided within minutes. This will enable caregivers and patients to take decisions at the point-of-need with a high level of confidence. The effort is backed by the know-how and technical expertise of researchers at imec (Leuven, Belgium), the world’s leading research and innovation hub in nanoelectronics and digital technologies, and The Johns Hopkins University (Baltimore, Maryland), a world-renowned academic medicine and research institution.
One of the key features that miDIAGNOSTICS wants to integrate in its chip-based tests is the ability to identify and count blood cells, leveraging imec’s lens-free imaging (LFI) technology. LFI is an advanced microscopy technique that creates holograms by passing light through a sample. LFI is uniquely suited for this compact application because it does not require expensive lenses or manual focusing and offers wider fields of view compared to traditional microscopy.
Researchers from The Johns Hopkins University have recently developed LFI processing techniques that improve the image quality of the holograms and allow a more accurate analysis by separating the foreground, i.e. the blood cells, from the background. This method was further optimized with convolutional sparse coding and dictionary learning techniques. Preliminary results demonstrate that these methods help to reduce unwanted holographic artifacts and greatly improve cell detection and counting. “These developments are getting us closer to fast and accurate detection and counting of blood cells in lens-free images,” noted senior author René Vidal, professor of Biomedical Engineering at the Johns Hopkins School of Medicine and director of the Vision Dynamics and Learning Lab in the Center for Imaging Science.
Leander Van Neste, Vice President Scientific and Clinical Affairs at miDIAGNOSTICS, adds that “These algorithms allow us to implement the compact lens-free imaging approach and therefore represents an important step in the development of our accurate and rapid complete blood count solution for the point-of-need.”
Both publications were presented at ISBI 2017 in Melbourne, Australia, 18-21 April and are now available online at: