With the development of wireless capsule endoscopy (WCE) and its extensive applications in clinic, doctors need to spend more time on reviewing the WCE images for lesions diagnosis. Therefore, automatic lesion detection for WCE has gradually become a research hotspot, which aims to reduce the pressure of doctors and improves the diagnosis efficiency. Many researchers adopted the traditional machine learning method to realize polyp detections, however, these methods need to extract the features manually, which were unable to find higher features of WCE images. So, in this study, we proposed a novel method that based on convolution neural network (CNN) to automatically recognize polyp in small bowel WCE image. We utilized the Alexnet architecture, one of the classical CNN, to extract the features of WCE images and classify polyp images from normal ones. We selected 14408 images from different patients, including 408 polyp images and 14000 normal images. Since the amount of initial polyp images is small, then, we did the data augmentation, including rotation, luminance change, blurring, and noise. At last the experimental results demonstrated that the method we proposed had a promising performance in polyp detection, whose accuracy, sensitivity and specificity can reach at 99.88%, 99.40% and 99.93%, respectively. Additionally, we evaluated ROC curve and its AUC value, which further confirmed that our model has a high accuracy and reliability in polyp detection. This proposed method has great potential to be used in the clinical examination to help doctors from the tedious image reviewing work.
We presented an improved dual channel dual focus spectral domain optical coherence tomography (SD-OCT) with two illuminations at 840 nm and 1050 nm for whole eye segment imaging and biometry in vivo. The two light beams were coupled and optically optimized to scan the anterior and posterior segment of the eye simultaneously. This configuration with dichroic mirrors integrated in the sample arm enables us to acquire images from the anterior segment and retina effectively with minimum loss of sample signal. In addition, the full resolved complex (FRC) method was applied to double the imaging depth for the whole anterior segment imaging by eliminating the mirror image. The axial resolution for 1050 nm and 840 nm OCT was 14 μm and 8 μm in air, respectively. Finally, the system was successfully tested in imaging the unaccommodated and accommodated eyes. The preliminary results demonstrated the significant improvements comparing with our previous dual channel SD-OCT configuration in which the two probing beams had the same central wavelength of 840 nm.
For OCT imaging, enhancing contrast efficiency will lead to significant improvements in the detection limits in cancer. Recently, noble metal nanoparticles are considered to be better contrast agents than traditional ones, especially for gold and silver. Silver nanoparticles have more attractive optical properties than gold nanoparticles. But they are employed far less because of its poor chemical stability. In this paper, we introduced our recent progress on a new application of using gold/silver alloy nanoparticles as OCT contrast agents in the detection of ovarian cancer. The scattering properties and sensitivity of silver were investigated. By means of tuning LSPR wavelengths of the nanoparticles, they were able to match the central wavelength of light used in OCT. Before carrying out animal experiments, we evaluated the different performances of alloy nanoparticles and gold nanorods in vitro. It has been sufficiently demonstrated that the alloy nanoparticles revealed stronger OCT signals than gold nanorods because of the better scattering properties. Then in vivo study, we compared the contrast enhancement of gold/silver alloy nanoparticles and gold nanorods on the ovarian cancer model mice. This study contributes a new kind of contrast agent in OCT imaging, which has a profound effect on drug delivery and further therapeutic action.
To research retinal stretching or distortion with accommodation, accommodation-induced changes in retinal thickness (RT) in the macular area were investigated in a population of young adults (n=23) by using a dual-channel spectral domain optical coherence tomography (SD-OCT) system manufactured in-house for this study. This dual-channel SD-OCT is capable of imaging the cornea and retina simultaneously with an imaging speed of 24 kHz A-line scan rate, which can provide the anatomical dimensions of the eye, including the RT and axial length. Thus, the modification of the RT with accommodation can be calculated. A significant decrease in the RT (13.50±1.25 μm) was observed during maximum accommodation. In the 4 mm×4 mm macular area centered at the fovea, we did not find a significant quadrant-dependent difference in retinal volume change, which indicates that neither retinal stretching nor distortion was quadrant-dependent during accommodation. We speculate that the changes in RT with maximum accommodation resulted from accommodation-induced ciliary muscle contractions.