Searching by parts: Towards fine-grained image retrieval respecting species correlation

Most of the existing works on fine-grained image categorization and retrieval focus on finding similar images from the same species and often give little importance to inter-species similarities. However, these similarities may carry species correlations such as the same ancestors or similar habits, which are helpful in taxonomy and understanding biological traits. In this paper, we devise a new fine-grained retrieval task that searches for similar instances from different species based on body parts. To this end, we propose a two-step strategy. In the first step, we search for visually similar parts to a query image using a deep convolutional neural network (CNN). To improve the quality of the retrieved candidates, structural cues are introduced into the CNN using a novel part-pooling layer, in which the receptive field of each part is adjusted automatically. In the second step, we re-rank the retrieved candidates to improve the species diversity. We achieve this by formulating a novel ranking function that balances between the similarity of the candidates to the queried parts, while decreasing the similarity to the query species. We provide experiments on the benchmark CUB200 dataset and Columbia Dogs dataset, and demonstrate clear benefits of our schemes.     

Development of an optimal antigen retrieval protocol for immunohistochemistry of retinoblastoma protein (pRB) in formalin fixed,paraffin sections based on comparison of different methods

A novel protocol for antigen retrieval (AR) for immunohistochemistry (IHC) of retinoblastoma protein (pRB) in formalin fixed, paraffin embedded (FFPE) tissue sections was developed using 0.05% citraconic anhydride as the AR solution for heat treatment based on comparison of different methods. This new protocol has advantages including superior morphological preservation, greater reproducibility, and more intense staining after retrieval. Our study demonstrates the importance of comparing various AR protocols to obtain maximal IHC for standardization and for quantitative IHC.     

Immunohistochemistry in the evaluation of neovascularization in tumor xenografts

Angiogenesis, or neovascularization, is known to play an important role in the neoplastic progression leading to metastasis. CD31 or Factor VIII-related antigen (F VIII RAg) immunohistochemistry is widely used in experimental studies for quantifying tumor neovascularization in immunocompromised animal models implanted with transformed human cell lines. Quantification, however, can be affected by variations in the methodology used to measure vascularization including antibody selection, antigen retrieval (AR) pretreatment, and evaluation techniques. To examine this further, we investigated the microvessel density (MVD) and the intensity of microvascular staining among five different human tumor xenografts and a mouse syngeneic tumor using anti-CD31 and F VIII RAg immunohistochemical staining. Different AR methods also were evaluated. Maximal retrieval of CD31 was achieved using 0.5 M Tris (pH 10) buffer, while maximum retrieval of F VIII RAg was achieved using 0.05% pepsin treatment of tissue sections. For each optimized retrieval condition, anti-CD31 highlighted small vessels better than F VIII RAg. Furthermore, the MVD of CD31 was significantly greater than that of F VIII RAg decorated vessels (p<0.001). The choice of antibody and AR method has a significant affect on immunohistochemical findings when studying angiogenesis. One also must use caution when comparing studies in the literature that use different techniques and reagents.     

Molecular mechanisms of antigen retrieval: antigen retrieval reverses steric interference caused by formalin-induced cross-links

Abstract

The overwhelming majority of antibodies useful for formalin fixed, paraffin embedded (FFPE) tissues require antigen retrieval to reverse the effect of formalin fixation and re-establish immunoreactivity. How this reversal happens is poorly understood. We developed a new experimental model for studying the mechanism of formalin fixation and antigen retrieval. Epitope mapping studies on nine antibodies useful for FFPE tissues revealed that each consisted of a contiguous stretch of amino acids in the native protein (linear epitope). Small peptides representing the epitopes of antibodies to human epidermal growth factor receptor type (HER2), estrogen, and progesterone receptors were attached covalently to glass microscope slides in a peptide array. Most peptides retained immunoreactivity after formalin fixation. Immunoreactivity was completely abrogated for all peptides, however, if an irrelevant large protein was present during formalin-induced cross-linking. We hypothesize that cross-linking the irrelevant protein to the peptide epitopes sterically blocked antibodies from binding. Antigen retrieval dissociates irrelevant proteins and restores immunoreactivity. Because the epitopes for clinical antibodies require only primary protein structure, the fact that antigen retrieval probably denatures the secondary and tertiary structure of the protein is irrelevant. The same mechanism may occur in tissue samples subjected to formalin fixation and antigen retrieval.     

Combined effects of formalin fixation and tissue processing on immunorecognition

Abstract

It is accepted that aldehyde-based fixation of cells can affect immunodetection of antigens; however, the effects of tissue processing on immunodetection have not been analyzed systematically. We investigated the effects of aldehyde-based fixation and the various cumulative steps of tissue processing on immunohistochemical detection of specific antigens. DU145 (prostate) and SKOV3 (ovarian) cancer cell lines were cultured as monolayers on microscope slides. Immunohistochemical detection of Ki67/MIB-1 and proliferating cell nuclear antigen (PCNA) was evaluated after various fixation times in 10% neutral buffered formalin and after each of the several cumulative steps of tissue processing. The effect of antigen retrieval (AR) was evaluated concomitantly as an additional variable. Our results indicate that in addition to fixation, each of the tissue processing steps has effects on immunorecognition of the epitopes recognized by these antibodies. Extensive dehydration through ethanols to absolute ethanol had only modest effects, except for the detection of Ki67/MIB-1 in SKOV-3 cells where the effect was stronger. In general, however, establishment of a hydrophobic environment by xylene resulted in the greatest decrease in immunorecognition. AR compensated for most, but not all, of the losses in staining following fixation and exposure to xylene; however, AR gave consistent results for most steps of tissue processing, which suggests that AR also should be used for staining PCNA. The cellular variations that were observed indicate that the effects of fixation and other steps of tissue processing may depend on how antigens are packaged by specific cells.     

News in Brief

Archival formalin-fixed, paraffin-embedded (FFPE) tissue and their associated diagnostic records represent an invaluable source of retrospective proteomic information on diseases for which the clinical outcome and response to treatment are known. However, analysis of archival FFPE tissues by high-throughput proteomic methods has been hindered by the adverse effects of formaldehyde fixation and subsequent tissue histology. This review examines recent methodological advances for extracting proteins from FFPE tissue suitable for proteomic analysis. These methods, based largely upon heat-induced antigen retrieval techniques borrowed from immunohistochemistry, allow at least a qualitative analysis of the proteome of FFPE archival tissues. The authors also discuss recent advances in the proteomic analysis of FFPE tissue; including liquid-chromatography tandem mass spectrometry, reverse phase protein microarrays and imaging mass spectrometry.     

Rule extraction in gene–disease relationship discovery

Background

Biomedical data available to researchers and clinicians have increased dramatically over the past years because of the exponential growth of knowledge in medical biology. It is difficult for curators to go through all of the unstructured documents so as to curate the information to the database. Associating genes with diseases is important because it is a fundamental challenge in human health with applications to understanding disease properties and developing new techniques for prevention, diagnosis and therapy.

Methods

Our study uses the automatic rule-learning approach to gene–disease relationship extraction. We first prepare the experimental corpus from MEDLINE and OMIM. A parser is applied to produce some grammatical information. We then learn all possible rules that discriminate relevant from irrelevant sentences. After that, we compute the scores of the learned rules in order to select rules of interest. As a result, a set of rules is generated.

Results

We produce the learned rules automatically from the 1000 positive and 1000 negative sentences. The test set includes 400 sentences composed of 200 positives and 200 negatives. Precision, recall and F-score served as our evaluation metrics. The results reveal that the maximal precision rate is 77.8% and the maximal recall rate is 63.5%. The maximal F-score is 66.9% where the precision rate is 70.6% and the recall rate is 63.5%.

Conclusions

We employ the rule-learning approach to extract gene–disease relationships. Our main contributions are to build rules automatically and to support a more complete set of rules than a manually generated one. The experiments show exhilarating results and some improving efforts will be made in the future.     

3D HUMAN MOTION RETRIEVAL BASED ON HUMAN HIERARCHICAL INDEX STRUCTURE

With the development and wide application of motion capture technology, the captured motion data sets are becoming larger and larger. For this reason, an efficient retrieval method for the motion database is very important. The retrieval method needs an appropriate indexing scheme and an effective similarity measure that can organize the existing motion data well. In this paper, we represent a human motion hierarchical index structure and adopt a nonlinear method to segment motion sequences. Based on this, we extract motion patterns and then we employ a fast similarity measure algorithm for motion pattern similarity computation to efficiently retrieve motion sequences. The experiment results show that the approach proposed in our paper is effective and efficient.     

丙泊酚复合地佐辛用于无痛取卵手术的临床观察

目的:探讨丙泊酚复合地佐辛在生殖科取卵手术中的安全性和有效性。方法:选取择期行取卵手术的的患者100例,随机分为丙泊酚复合芬太尼组(F组)和丙泊酚复合地佐辛组(D组),观察和比较两组患者麻醉前(T0)、手术开始时(T1)、10min(T2)和手术结束时刻(T3)的呼吸、血压、心率、SPO_2的变化情况,意识丧失和唤醒时间,丙泊酚用量和不良反应的发生情况。对记录的数据进行组间比较,计量资料采用x±s,采用t检验,计数资料采用x2检验。结果:两组患者给药前后不同时点的心率、血压变化和意识丧失时间比较均无统计学差异,但D组丙泊酚用量低于F组,唤醒时间短于F组,F组疼痛和恶心的发生率高于D组,差异具有统计学意义(P0.05)。结论:丙泊酚复合地佐辛在生殖科取卵手术中可以提供相对较好的麻醉效果。