Fully-automated patient-level malaria assessment on field-prepared thin blood film microscopy images, including Supplementary Information
Authors:
Charles B. Delahunt,
Mayoore S. Jaiswal,
Matthew P. Horning,
Samantha Janko,
Clay M. Thompson,
Sourabh Kulhare,
Liming Hu,
Travis Ostbye,
Grace Yun,
Roman Gebrehiwot,
Benjamin K. Wilson,
Earl Long,
Stephane Proux,
Dionicia Gamboa,
Peter Chiodini,
Jane Carter,
Mehul Dhorda,
David Isaboke,
Bernhards Ogutu,
Wellington Oyibo,
Elizabeth Villasis,
Kyaw Myo Tun,
Christine Bachman,
David Bell,
Courosh Mehanian
Abstract:
Malaria is a life-threatening disease affecting millions. Microscopy-based assessment of thin blood films is a standard method to (i) determine malaria species and (ii) quantitate high-parasitemia infections. Full automation of malaria microscopy by machine learning (ML) is a challenging task because field-prepared slides vary widely in quality and presentation, and artifacts often heavily outnumb…
▽ More
Malaria is a life-threatening disease affecting millions. Microscopy-based assessment of thin blood films is a standard method to (i) determine malaria species and (ii) quantitate high-parasitemia infections. Full automation of malaria microscopy by machine learning (ML) is a challenging task because field-prepared slides vary widely in quality and presentation, and artifacts often heavily outnumber relatively rare parasites. In this work, we describe a complete, fully-automated framework for thin film malaria analysis that applies ML methods, including convolutional neural nets (CNNs), trained on a large and diverse dataset of field-prepared thin blood films. Quantitation and species identification results are close to sufficiently accurate for the concrete needs of drug resistance monitoring and clinical use-cases on field-prepared samples. We focus our methods and our performance metrics on the field use-case requirements. We discuss key issues and important metrics for the application of ML methods to malaria microscopy.
△ Less
Submitted 11 September, 2022; v1 submitted 5 August, 2019;
originally announced August 2019.
Photonic Crystal Optical Tweezers
Authors:
Benjamin K. Wilson,
Tim Mentele,
Stephanie Bachar,
Emily Knouf,
Ausra Bendoraite,
Muneesh Tewari,
Suzie H. Pun,
Lih Y. Lin
Abstract:
Non-invasive optical manipulation of particles has emerged as a powerful and versatile tool for biological study and nanotechnology. In particular, trap** and rotation of cells, cell nuclei and sub-micron particles enables unique functionality for various applications such as tissue engineering, cancer research and nanofabrication. We propose and demonstrate a purely optical approach to rotate…
▽ More
Non-invasive optical manipulation of particles has emerged as a powerful and versatile tool for biological study and nanotechnology. In particular, trap** and rotation of cells, cell nuclei and sub-micron particles enables unique functionality for various applications such as tissue engineering, cancer research and nanofabrication. We propose and demonstrate a purely optical approach to rotate and align particles using the interaction of polarized light with photonic crystal nanostructures to generate enhanced trap** force. With a weakly focused laser beam we observed efficient trap** and transportation of polystyrene beads with sizes ranging from 10 um down to 190 nm as well as cancer cell nuclei. In addition, we demonstrated alignment of non-spherical particles using a 1-D photonic crystal structure. Bacterial cells were trapped, rotated and aligned with optical intensity as low as 17 uW/um^2. Finite-difference time domain (FDTD) simulations of the optical near-field and far-field above the photonic crystal nanostructure reveal the origins for the observed results. This approach can be extended to using 2-D photonic crystal nanostructures for full rotation control.
△ Less
Submitted 25 December, 2009;
originally announced December 2009.
