Digital video otoscope is an indispensable tool in otology that allows inspection of the external auditory canal and tympanic membrane. However, existing solutions have limitations in the diagnosis of various ear diseases and portability. Here, we propose a mobile, deep learning-assisted otoscope for low-resource settings. Our deep learning architecture was trained on clinical data to identify and classify various ear diseases. To evaluate our platform, we compared its performance with the device used in the hospital practice. Our preliminary results demonstrated high diagnostic accuracy indicating a strong potential to become a viable screening solution in low-resource, non-specialist settings.
Zebrafish is a useful biological model for analyzing genetic modification and large-scale screening. Its morphological evaluation, carrying meaningful information about genotype-phenotype relationship, is equally important. However, analysis of large amounts across development stages is a labor-intensive task. Here, we suggest a high-throughput monitoring technique using office scanner. Moreover, we developed deep learning models for extraction and analysis of massive statistical information. CNN-based architecture, forming the core of segmentation, serves as a basis for quantitative analysis and an early signal for embryo’s abnormal growth. Finally, compared to conventional microscope imaging, our scanning technique offers high-throughput, accurate, and fast quantitative phenotype analysis.
Xenopus laevis are emerging models to study human diseases and to investigate pharmaceutical effects in vivo due to
smaller size and faster developmental rates. It is also an effective organism to observe drug effects on phenotypic
characteristics because it can provide many biological systems in a short time and remain optically accessible at the early
stages of development. Although morphological evaluation of massive Xenopus data is an essential procedure, it requires
labor-intensive and manual inspection under an optical microscope. In this study, we propose a high-throughput, widefield,
and time-lapse phenotype screening system modifying the office scanner. We also fabricated the customized
PDMS well plate for efficient and stress-free imaging of living Xenopus laevis samples in normal and drug environments.
With our manipulated device, we were successfully able to monitor the morphological changes of Xenopus laevis
embryos acquired from more than 180 wells throughout 72 hours post fertilization stage. Our home-built software
combines best practices of image processing and deep learning for automated accurate segmentation of large Xenopus
data. Importantly, phenotypic features are quantitatively extracted to monitor the early-stage morphological
abnormalities. In addition, the convolutional neural network (CNN) based algorithm enable to classify phenotype
precisely. In conclusion, compared to conventional microscope screening, our platform offers high-throughput, accurate,
and fast quantitative phenotype analysis. The suggested platform could become a promising tool in massive and dynamic
observation such as developmental studies, drug testing, and phenotype-genotype assays, where statistical knowledge is
critical.
Optical coherence tomography (OCT) has been used for visualization of morphological change of tissues over time. Although current OCT technology allows the volumetric and high throughput information of tissues, its quantification and analysis still uses time inefficient and tedious process. In order to fully utilize benefits of OCT, it is desired to integrate the intelligent software platform. As deep learning technology is advanced, it has been emerged as the alternative way for quantitative and automated image processing in bio-imaging field including optical imaging. Deep leaning technique is based on the sufficient training data which could overcome the drawback of traditional handcrafted optical image processing algorithms.
In this study, we introduce a novel and intelligent OCT software platform for accurate skin analysis and classification using deep learning module. Our platform is equipped with automated calculations of morphological skin parameters, such as surface roughness, wrinkle depth, volume, and epidermal thickness. To date, most promising tool for quantitative skin analysis is to use a software package of PRIMOS device which relies on three-dimensional camera systems. In order to evaluate our software platform, we compared OCT skin parameters based on deep learning technique and conventional PRIMOS data. Our preliminary study shows that proposed software platform for 3D OCT is a promising tool for accurate, efficient, and quantitative analysis of volumetric skin. It could be also a better alternative than existing PRIMOS solutions to both cosmeceutical and dermatological field.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.