Search

(1) Background: The difficulty of pelvic operation is greatly affected by anatomical constraints. Defining this difficulty and assessing it based on conventional methods has some limitations. Artificial intelligence (AI) has enabled rapid advances in surgery, but its role in assessing the difficulty of laparoscopic rectal surgery is unclear. This study aimed to establish a difficulty grading system to assess the difficulty of laparoscopic rectal surgery, as well as utilize this system to evaluate the reliability of pelvis-induced difficulties described by MRI-based AI. (2) Methods: Patients who underwent laparoscopic rectal surgery from March 2019 to October 2022 were included, and were divided into a non-difficult group and difficult group. This study was divided into two stages. In the first stage, a difficulty grading system was developed and proposed to assess the surgical difficulty caused by the pelvis. In the second stage, AI was used to build a model, and the ability of the model to stratify the difficulty of surgery was evaluated at this stage, based on the results of the first stage; (3) Results: Among the 108 enrolled patients, 53 patients (49.1%) were in the difficult group. Compared to the non-difficult group, there were longer operation times, more blood loss, higher rates of anastomotic leaks, and poorer specimen quality in the difficult group. In the second stage, after training and testing, the average accuracy of the four-fold cross validation models on the test set was 0.830, and the accuracy of the merged AI model was 0.800, the precision was 0.786, the specificity was 0.750, the recall was 0.846, the F1-score was 0.815, the area under the receiver operating curve was 0.78 and the average precision was 0.69; (4) Conclusions: This study successfully proposed a feasible grading system for surgery difficulty and developed a predictive model with reasonable accuracy using AI, which can assist surgeons in determining surgical difficulty and in choosing the optimal surgical approach for rectal cancer patients with a structurally difficult pelvis.

In this study, a polymer dispersed cholesteric liquid crystal (PDChLC) film obtained via a one-step fabrication technique based on photopolymerization of a thiol-acrylate reaction system was prepared and characterized for the first time. The effects of the chiral dopant, the influence of thiol monomer functionality and content on the morphology and subsequent performance of the PDChLC films were systematically investigated. It was demonstrated that the addition of a small amount of chiral dopant slightly increased the driving voltage, but decreased the off-state transmittance significantly. Furthermore, scanning electron micrographs (SEM) shown that the liquid crystal (LC) droplet size decreased at first and then increased with the increasing amount of thiol monomer functionality, while increasing the thiol content increased the LC droplet size. Correspondingly, the electro-optical switching behavior was directly dependent on LC droplet size. By tuning the raw material composition, PDChLC film with optimized electro-optical performance was prepared.

Polymer-dispersed liquid crystal (PDLC) films were prepared by the ultraviolet-light-induced polymerization of photopolymerizable monomers in nematic liquid crystal/chiral dopant/thiol-acrylate reaction monomer composites. The effects of the chiral dopant and crosslinking agents on the electro-optical properties of the PDLC films were systematically investigate. While added the chiral dopant S811 into the PDLC films, the initial off-state transmittance of the films was decreased. It was found that the weight ratio among acrylate monomers, thiol monomer PETMP and the polymercaptan Capcure 3-800 showed great influence on the properties of the fabricated PDLC films because of the existence of competition between thiol-acrylate reaction and acrylate monomer polymerization reaction. While adding polymercaptans curing agent Capcure 3-800 with appropriate concentration into the PDLC system, lower driven voltage and higher contrast ratio were achieved. This made the polymer network and electro-optical properties of the PDLC films easily tunable by the introduction of the thiol monomers.

Perylene diimide derivatives were used as probes to investigate the effect of the molecular structures on the fluorescence quenching mechanism in a perylene diimide/graphene oxide system. The electrons transferred from the excited state of dyes to the conductive band of graphene oxide with different concentrations were determined by fluorescence spectra. The results indicated that the quenching efficiency of perylene diimides by graphene oxide was not only dependent on the difference between the lowest unoccupied molecular orbital level of dyes and the conduction band of the graphene oxide, but also mainly on the difference in the molecular structures.

Holographic communication is a three-dimensional (3D) video communication technology based on computer-generated holograms (CGHs) which has the potential to give users a more realistic visual perception. As this is an emerging field, the encrypted encoding and decoding methods in holographic communication have not been widely studied. In this work, a watermarking and encryption method for holographic communication is proposed. A watermark is inserted into the original image using the discrete cosine transform before the calculation of the CGH, while a secret key is employed to produce the encrypted CGH during the holographic calculation. Through the proposed watermarking and encryption method, the signal of holographic communication is difficult to decrypt. Even if the signal is decrypted, the source of the leak is easy to trace due to the existence of the watermark. The watermarking and encryption method can provide a practical solution for the privacy protection and copyright protection of 3D video communication.

With the explosive developments of deep learning, learning–based computer–generated holography (CGH) has become an effective way to achieve real–time and high–quality holographic displays. Plentiful learning–based methods with various deep neural networks (DNNs) have been proposed. In this paper, we focus on the rapid progress of learning–based CGH in recent years. The generation principles and algorithms of CGH are introduced. The DNN structures frequently used in CGH are compared, including U–Net, ResNet, and GAN. We review the developments and discuss the outlook of the learning–based CGH.