Search

Numerous studies have employed prompt learning structures to enhance dense prediction tasks by integrating additional semantic or geometric information. While the inclusion of extra information has shown improvements in performance, it also poses challenges for applications that cannot provide extra input. To address this issue, this study evaluates the performance of different prompts and introduces an additional-input-free method, called self-prompting perceptual edge learning (SPPEL), which extracts edge-embedded semantic prompts directly from the image feature itself using trainable handcrafted edge operators within a plug-and-play module. To obtain the edge features, our approach incorporates an adversarial structure that compares the similarity between two edge features generated by the Hog and Kirsch operators, where the edge features are measured using multiplication, finetuned through a trainable all-one embedding, and enhanced with channel-to-channel attention. We conduct extensive evaluations of SPPEL on 7 tasks, utilizing 7 different backbones and applying 5 distinct methods. Our experimental results demonstrate that SPPEL achieves strong competitiveness in various settings with an average improvement of 1.7% across all 7 tasks, including ADE20K, COCO (Instance Segmentation), COCO (Object Detection), Pascal VOC2012, STARE, CHASE DB1, and HRF, while incurring a parameter increase of less than 3% (the detailed computation analysis of parameters and Gflops are shown in different experimental tables). Code will be released at: https://github.com/chenhao-zju/sppel

ABSTRACTIt is well established that 5-methylcytosine (5mC) in genomic DNA of mammalian cells can be oxidized into 5-hydroxymethylcytosine (5hmC) and other derivates by DNA dioxygenase TETs. While conversion of 5mC to 5hmC plays an important role in active DNA demethylation through further oxidation steps, a certain proportion of 5hmCs remain in the genome. Although 5hmCs contribute to the flexibility of chromatin and protect bivalent promoters from hypermethylation, the direct effect of 5hmCs on gene transcription is unknown. In this present study, we have engineered a zinc-finger protein-based P16-specific DNA dioxygenase (P16-TET) to induce P16hydroxymethylation and demethylation in cancer cells. Our results demonstrate, for the first time, that although the hydroxymethylated P16alleles retain transcriptionally inactive, hydroxymethylation could increase the susceptibility of reactivation of methylated P16alleles.

Audio fingerprinting is the process to obtain a compact content-based signature that summarizes the essence of an audio clip. In general, existing audio fingerprinting schemes based on wavelet transforms are not robust against large linear speed changes. The authors present a novel framework for content-based audio retrieval based on the audio fingerprinting scheme that is robust against large linear speed changes. In the proposed scheme, 8 levels Daubechies wavelet decomposition is adopted for extracting time-frequency features and two fingerprint extraction algorithms are designed. The experimental results from this study are discussed further into the article.

This paper presents a novel image watermarking scheme utilizing Block Truncation Coding (BTC). This scheme uses BTC to guide the watermark embedding and extraction processes. During the embedding process, the original cover image is first partitioned into non-overlapping 4×4 blocks. Then, BTC is performed on each block to obtain its BTC bitplane, and the number of ‘1’s in the bitplane is counted. If the watermark bit to be embedded is ‘1’ and the number of ‘1’s is odd, or the watermark bit to be embedded is ‘0’ and the number of ‘1’s is even, then no change is made. Otherwise, by changing at most three pixels in the original image block, the number of ‘1’s (or ‘0’s) in the renewed bitplane are forced to be odd for the watermark bit ‘1’ or to be even for the watermark bit ‘0’. During the extraction process, BTC is first performed on each block to obtain its bitplane. If the number of ‘1’s in the bitplane is odd, then the embedded watermark bit is ‘1’. Otherwise, the embedded watermark bit is ‘0’. The experimental results show that the proposed watermarking method is semi-fragile except for the changes in brightness and contrast; therefore, the proposed method can be used for image authentication.

Abstract: Objective: To elucidate protein markers that differentiate stage I non-small cell lung cancer (NSCLC) that subsequently develop metastatic disease to those that do not develop metastasis by protein expression profiles. Methods: Matrix assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) and two dimensional difference gel electrophoresis (2D-DIGE) platforms were used to separate proteins in whole tumor specimens. Quantitating mRNA expression was used for validation. Results: Twelve proteins were identified as expressed differentially between two groups from protein expression platforms. But from gene expression platform no marker could distinguish patients with recurrent vs. nonrecurrent disease. Conclusion: Analysis of multiple protein markers may be more informative to predict prognosis of early stage lung cancer.

Abstract: This paper proposes a self-embedding watermarking scheme for digital images. It exploits a digital halftoning technique to transform the host image to a halftone image, in which the content features of the host image are well preserved. This halftone image is utilized as an approximate or compressed version of the host image and further adopted as a watermark. Unlike most available self-embedding watermarking techniques, in which watermark embedding, tamper detection and image recovery are block-wise-based, our scheme is a pixel-wise and block-wise composite scheme. That is, the watermark is pixel-wise permuted and embedded in the least-significant-bit plane of the host image, while in tamper detection and image recovery, a block-wise strategy is employed. This pixel-wise and block-wise complementary mechanism can overcome a common disadvantage of previous block-wise-based self-embedding watermarking techniques. Our scheme can be used for tamper detection and alteration location. Moreover, the tampered area can be approximately recovered using the extracted watermark. Experimental results demonstrate the effectiveness of the proposed scheme.

Abstract: Objective: To look for the further evidence for HPV L1 HPV16 E6, HPV 18 E6 and EBV as carcinogenic factors in laryngeal carcinoma. Method: we examined representative numbers of specimens from laryngeal cancer with highly sensitive PCR technique for the presence of HPV L1 and high-risk types HPV16 E6, HPV18 E6 and EBV LMP1. Results: Using PCR detection, 7.3% samples were HPV L1 positive, 52.03% were HPV16 E6 positive, 30.89% were HPV18 E6 positive and 9.13% were EBV LMP1 positive. The low incidence of HPV L1 and high incidence of HPV-16 E6 and HPV18 E6 genes suggest that HPV might be integrated into tumor cells. Our results support a role of HPV-16 and HPV-18 infection in the pathogenesis of laryngeal carcinoma in China. Conclusion: Integration of E6 into host genome and stable expression of these genes may be associated with the carcinogenesis of laryngeal carcinoma. HPV-16 and HPV-18 may synergistically function on the pathogenesis of laryngeal carcinoma. Our results suggest an association of laryngeal carcinogenesis and infection with the high-risk HPV types 16, HPV 18 and EBV.

Human papillomavirus (HPV) plays an important role in the development of carcinomas at various body sites. It was found previously that the p53 codon 72 polymorphism (C/G) is a high-risk factor for the development HPV-associated cervical carcinoma. However, it still was considered controversial in several studies of cervical and esophageal carcinoma. In the current study, the authors used an allele specific polymerase chain reaction (PCR) method to analyze correlation between the p53 codon 72 (C/G) polymorphism and HPV-associated, noncancerous esophageal epithelium as well as esophageal, ovarian, and breast carcinoma in the Chinese population. Esophageal balloon cytology examination samples were obtained from high-incidence and low-incidence populations for esophageal carcinoma in Anyang (Henan Province). Thirty-six of 48 esophageal balloon samples from the high-incidence population were HPV positive, and 13 of 33 esophageal balloon samples from the low-incidence population were HPV positive. Thirty-nine of 62 esophageal carcinoma samples from Anyang Tumor Hospital were HPV positive. Twenty-six of 39 ovarian carcinoma samples from the Second Affiliated Hospital of Inner Mongolia Medical College were HPV positive. Nineteen of 82 breast carcinoma samples from Beijing Cancer Hospital were HPV positive. It is noteworthy that the distribution of the p53 codon 72 Arg homozygous genotype in HPV positive samples of esophageal epithelium, ovarian carcinoma, and breast carcinoma was significantly higher compared with HPV negative tumor samples. (P < 0.05). The current results suggest that the p53 codon 72 Arg homozygous genotype is one of the high-risk genetic factors for HPV-associated malignancies among the Chinese population. Cancer 2002;95:2571–6. © 2002 American Cancer Society. DOI 10.1002/cncr.11008

To investigate the potential role of human papillomavirus (HPV) infection in the pathogenesis of esophageal carcinomas in the Anyang area of China, we have evaluated specimens collected by balloon cytology examination from volunteers in two regions with significantly different incidences of esophageal carcinoma. 138 donors were from a village in a county with an esophageal carcinoma (EC) age-adjusted mortality rate of 132×105, the remaining 68 were resident in a second village from another county with an EC mortality rate of 52×105. Specimens were evaluated using both polymerase chain reaction (PCR) amplification and in situ hybridization (ISH) protocols. PCR results showed that the prevalence of the human papillomavirus type 16 (HPV-16) E6 gene in the high incidence area was 1.9-fold higher than that of the low incidence area (72 and 37%, respectively, P < 0.01). Moreover, the positive rate corresponded with pathology grade. Similar results were obtained with the HPV-16 E7 gene. As the cells undergoing cytopathological progress, the HPV-16 E6 positive rate was increased, in both villages. In contrast to HPV-16 E6 and E7, detection of the HPV L1 gene was consistently lower, and its prevalence decreased with increasing dysplasia grades (P < 0.05). By ISH analyses, the expression rate of HPV-16 E6 in the specimens collected from the high incidence area was 2.2-fold higher than those from the low incidence area (49 versus 22%, respectively; P < 0.05), and transcription of the E6 gene paralleled cytopathology. HPV-18 was also detected in 17 and 15% of the specimens from the high and low incidence areas, respectively, but most of these samples were also simultaneously HVP-16 positive. These results suggest that HVP-16 plays a causative role in the high incidence of esophageal cancer in the Anyang region of China.

A multiresolution watermarking method with grey-level watermarking is presented. Both the watermark and the original image are decomposed into three-level multiresolution structures using different mechanisms. The decomposed watermark information is further coded by error correction coding. Each level of watermark information is embedded into the corresponding level of the transformed image. Experimental results show that the proposed algorithm is robust against common image manipulations.

A new fast codeword search algorithm for image vector quantisation (VQ) is introduced. This algorithm performs a fast codeword search in the Hadamard transform (HT) domain using the partial distance search (PDS) technique. Experimental results show that the algorithm needs only 2–3% of the distortion calculations of the exhaustive search method.

A vector-quantisation (VQ)-based watermarking method is presented which utilises the codebook expansion technique. This method is efficient, provides enhanced security and the watermarked image is robust against the effects of VQ compression. Moreover, the watermark extraction can be performed without the original image. Experimental results are presented which demonstrate the effectiveness of this algorithm.