Search

SignificanceOral cancer is among the most common cancers globally, especially in low- and middle-income countries. Early detection is the most effective way to reduce the mortality rate. Deep learning-based cancer image classification models usually need to be hosted on a computing server. However, internet connection is unreliable for screening in low-resource settings.AimTo develop a mobile-based dual-mode image classification method and customized Android application for point-of-care oral cancer detection.ApproachThe dataset used in our study was captured among 5025 patients with our customized dual-modality mobile oral screening devices. We trained an efficient network MobileNet with focal loss and converted the model into TensorFlow Lite format. The finalized lite format model is ∼16.3  MB and ideal for smartphone platform operation. We have developed an Android smartphone application in an easy-to-use format that implements the mobile-based dual-modality image classification approach to distinguish oral potentially malignant and malignant images from normal/benign images.ResultsWe investigated the accuracy and running speed on a cost-effective smartphone computing platform. It takes ∼300  ms to process one image pair with the Moto G5 Android smartphone. We tested the proposed method on a standalone dataset and achieved 81% accuracy for distinguishing normal/benign lesions from clinically suspicious lesions, using a gold standard of clinical impression based on the review of images by oral specialists.ConclusionsOur study demonstrates the effectiveness of a mobile-based approach for oral cancer screening in low-resource settings.

Gastrointestinal cancers continue to account for a disproportionately large percentage of annual cancer deaths in the United States. Advancements in miniature imaging technology combined with a need for precise and thorough tumor detection in gastrointestinal cancer screenings fuel the demand for new, small-scale, and low-cost methods of localization and margin identification with improved accuracy. Here, we report the development of a miniaturized, chip-on-tip, multispectral, fluorescence imaging probe designed for compatibility with a gastroscope working channel with the aim of detecting cancerous lesions in point-of-care endoscopy of the gastrointestinal lumen. Preclinical testing has confirmed fluorescence sensitivity and supports that this miniature probe can locate structures of interest via detection of fluorescence emission from exogenous contrast agents. This work demonstrates the design and preliminary performance evaluation of a miniaturized, single-use, chip-on-tip fluorescence imaging system, capable of detecting multiplexed fluorophores, and devised for deployment via the accessory channel of a standard gastroscope. [ABSTRACT FROM AUTHOR]

Multispectral fluorescence imaging (MSFI) is a technique that measures endogenous and exogenous tissue fluorescence to reveal crucial insights into underlying biological mechanisms. Developing well-characterized high-performance fluorescence imaging equipment is crucial to any clinical or research application. Considering the diverse range of clinical scenarios and ongoing research areas in biomedicine leveraging MSFI instrumentation, there is an evident and pressing need for comprehensive resources that detail the development of such tools. This study provides a template for developing an MSFI instrument by highlighting the development and verification of our instrumentation. We present the design roadmap alongside the development of a large-area MSFI instrument to measure tissue fluorescence and diffuse reflectance properties. We divide our design approach into four subsections, highlighting the important physical milestones: illumination, imaging, detection, and computation. Each subsection includes design considerations, the methods used to validate the performance, and finally, the results and discussion of the validation process. We present the validation of the instrument across the illumination, imaging, detection, and computation subsystems. Two fluorophores are used to validate the instrument through serial dilution to establish a detection threshold and spectral capabilities limit. A mouse model expressing multi-colored fluorescent proteins is used to verify the multispectral performance. Our study lays out the groundwork for researchers to design and validate their own MSFI instrument. We present this alongside our instrument design to study fluorescence and diffuse reflectance properties of large-area tissues, such as murine or resected surgical specimens. Findings from the application of MSFI instrumentation can be translated to motivate and guide the design of fluorescence-based medical tools for the clinic or research environment. [ABSTRACT FROM AUTHOR]