Search

Objective. The high-precision segmentation of retinal vessels in fundus images is important for the early diagnosis of ophthalmic diseases. However, the extraction for microvessels is challenging due to their characteristics of low contrast and high structural complexity. Although some works have been developed to improve the segmentation ability in thin vessels, they have only been successful in recognizing small vessels with relatively high contrast. Approach. Therefore, we develop a deep learning (DL) framework with a multi-stage and dual-channel network model (MSDC_NET) to further improve the thin-vessel segmentation with low contrast. Specifically, an adaptive image enhancement strategy combining multiple preprocessing and the DL method is firstly proposed to elevate the contrast of thin vessels; then, a two-channel model with multi-scale perception is developed to implement whole- and thin-vessel segmentation; and finally, a series of post-processing operations are designed to extract more small vessels in the predicted maps from thin-vessel channels. Main results. Experiments on DRIVE, STARE and CHASE_DB1 demonstrate the superiorities of the proposed MSDC_NET in extracting more thin vessels in fundus images, and quantitative evaluations on several parameters based on the advanced ground truth further verify the advantages of our proposed DL model. Compared with the previous multi-branch method, the specificity and F1 score are improved by about 2.18%, 0.68%, 1.73% and 2.91%, 0.24%, 8.38% on the three datasets, respectively. Significance. This work may provide richer information to ophthalmologists for the diagnosis and treatment of vascular-related ophthalmic diseases. [ABSTRACT FROM AUTHOR]

Endoscopic optical coherence tomography (OCT) possesses the capability to non-invasively image internal lumens; however, it is susceptible to saturation artifacts arising from robust reflective structures. In this study, we introduce an innovative deep learning network, ATN-Res2Unet, designed to mitigate saturation artifacts in endoscopic OCT images. This is achieved through the integration of multi-scale perception, multi-attention mechanisms, and frequency domain filters. To address the challenge of obtaining ground truth in endoscopic OCT, we propose a method for constructing training data pairs. Experimental in vivo data substantiates the effectiveness of ATN-Res2Unet in reducing diverse artifacts while preserving structural information. Comparative analysis with prior studies reveals a notable enhancement, with average quantitative indicators increasing by 45.4-83.8%. Significantly, this study marks the inaugural exploration of leveraging deep learning to eradicate artifacts from endoscopic OCT images, presenting considerable potential for clinical applications.

The identification and preservation of parathyroid glands (PGs) during thyroid surgery can be challenging. Many techniques have been developed to help surgeons find PGs. We have developed a novel mitoxantrone hydrochloride injection that can be used for lymphatic targeting. After local application during surgery, mitoxantrone hydrochloride injection for tracing (MHI) helps surgeons better identify and preserve PGs and helps pathologists find more lymph nodes. We conducted an open-label, multicenter, randomized clinical trial (CTR20171137) in six centers in China from 08/2017 to 12/2018. Patients with thyroid carcinoma were randomized to the MHI group or the control group. All patients received total thyroidectomy and bilateral central compartment lymph node dissection. The primary outcomes were the PG resection rate and lymph node staining rate. The full analysis set (FAS) included 461 patients, of which 228 were assigned to the MHI group, and 233 were assigned to the control group. The PG resection rates of the MHI group and the control group were 6.6% (15/228) and 26.6% (62/233), respectively, with a significant difference (P < 0.001). No PGs were stained blue with MHI. The central lymph nodes were stained blue with MHI, and the staining rate was 90.5%±12.0%. More lymph nodes were detected in the MHI group than in the control group (13.0±7.3 vs. 10.1±6.4 nodes/patient, P < 0.001). No adverse events related to MHI were observed. MHI is a safe and effective tracer that may help to preserve PGs and identify more central lymph nodes in patients with thyroid cancer., Competing Interests: None., (AJCR Copyright © 2022.)