Medical image translation using an edge-guided generative adversarial network with global-to-local feature fusion

In this paper,we propose a framework based deep learning for medical image translation using paired and unpaired training data.Initially,a deep neural network with an encoder-decoder structure is proposed for image-to-image translation using paired training data.A multi-scale context aggregation approach is then used to extract various features from different levels of encoding,which are used during the corresponding network decoding stage.At this point,we further propose an edge-guided generative adversarial network for image-to-image translation based on unpaired training data.An edge constraint loss function is used to improve network performance in tissue boundaries.To analyze framework performance,we conducted five different medical image translation tasks.The assessment demonstrates that the proposed deep learning framework brings significant improvement beyond state-of-the-arts.

Nigel Wiseman:Modern Texts,Classics,Terminology,and Language-learning Material

In this article,I trace my career from an early interest in language that eventually led me to engage in Chinese medical translation.Born in the United Kingdom in 1954,I gained a Bachelor of Arts degree in German and Spanish from the University of Heriot-Watt,Edinburgh.Subsequent study of Chinese took me to Taiwan,China,where I gained an interest in Chinese medicine and developed a coherent linguistic approach to Chinese medical term translation,which I and various colleagues have systematically applied in the translation and compilation of numerous works spanning modern texts,classics,dictionaries,and language-learning materials.Our translation approach,though conforming to principles universally recognized by translation theorists,has proven to be highly controversial,with opponents in both Western and Chinese circles.

Learning by Living: Empathy Learned through an Extended Medical Education Immersion Project

Introduction: Medical student training and experiences in nursing homes are often viewed as negative;however, long-term care services represent a growing aspect of our medical system that receives little attention in medical education. The University of New England College of Osteopathic Medicine Learning by Living Nursing Home Immersion Project accelerates students learning about older adult care and the importance of empathy. Methods: Learning by Living applies qualitative ethnographic/autobiographic research methods to answer the question: “What is it like for me to live the life of an older adult nursing home resident?” Two first year medical students (female, 25 y/o & male, 27 y/o) were “admitted” into two different nursing homes for 11 days each. The students were wheelchair reliant from their assigned diagnoses of dominant-side paralysis secondary to stroke and aspiration pneumonia;requiring standard procedures of care experienced by older adults residents with similar conditions. Data were collected in the form of journal notes for pre-fieldwork, fieldwork, and post-fieldwork and included subjective and objective reporting of observations, experiences, and resident encounters. Analyses included journal review and thematic categorization and coding through content analysis. Results: Four themes of significance intersected for the two medical students: 1) control;2) quality of life;3) communication;and 4) isolation. Being immersed in nursing home life provided the students with firsthand experiences of loss and challenges faced by nursing home residents. This aided students’ learning about empathy and communication as key components of providing health care. Conclusion: This experience imparted knowledge to the medical students’ about aging, dignity, and the importance of making meaningful connections in order to thrive. These insights provided a gateway to establish a model of patient-centered care that is comprehensive and empathic for the older adult population.

A Novel Unsupervised MRI Synthetic CT Image Generation Framework with Registration Network

In recent years,radiotherapy based only on Magnetic Resonance(MR)images has become a hot spot for radiotherapy planning research in the current medical field.However,functional computed tomography(CT)is still needed for dose calculation in the clinic.Recent deep-learning approaches to synthesized CT images from MR images have raised much research interest,making radiotherapy based only on MR images possible.In this paper,we proposed a novel unsupervised image synthesis framework with registration networks.This paper aims to enforce the constraints between the reconstructed image and the input image by registering the reconstructed image with the input image and registering the cycle-consistent image with the input image.Furthermore,this paper added ConvNeXt blocks to the network and used large kernel convolutional layers to improve the network’s ability to extract features.This research used the collected head and neck data of 180 patients with nasopharyngeal carcinoma to experiment and evaluate the training model with four evaluation metrics.At the same time,this research made a quantitative comparison of several commonly used model frameworks.We evaluate the model performance in four evaluation metrics which achieve Mean Absolute Error(MAE),Root Mean Square Error(RMSE),Peak Signal-to-Noise Ratio(PSNR),and Structural Similarity(SSIM)are 18.55±1.44,86.91±4.31,33.45±0.74 and 0.960±0.005,respectively.Compared with other methods,MAE decreased by 2.17,RMSE decreased by 7.82,PSNR increased by 0.76,and SSIM increased by 0.011.The results show that the model proposed in this paper outperforms other methods in the quality of image synthesis.The work in this paper is of guiding significance to the study of MR-only radiotherapy planning.

Advances in prostate cancer research models:From transgenic mice to tumor xenografting models

The identification of the origin and molecular characteristics of prostate cancer(PCa)has crucial implications for personalized treatment.The development of effective treatments for PCa has been limited;however,the recent establishment of several transgenicmouse lines and/or xenografting models is better reflecting the disease in vivo.With appropriate models,valuable tools for elucidating the functions of specific genes have gone deep into prostate development and carcinogenesis.In the present review,we summarize a number of important PCa research models established in our laboratories(PSA-Cre-ERT2/PTEN transgenic mouse models,AP-OX model,tissue recombination-xenografting models and PDX models),which represent advances of translational models from transgenic mouse lines to human tumor xenografting.Better understanding of the developments of these models will offer new insights into tumor progression and may help explain the functional significance of genetic variations in PCa.Additionally,this understanding could lead to new modes for curing PCa based on their particular biological phenotypes.

Animal modeling in bone research—Should we follow the White Rabbit?

Animal models are live subjects applied to translational research.They provide insights into human diseases and enhance biomedical knowledge.Livestock development has favored the pace of human social development over millennia.Today's society is more aware of animal welfare than past generations.The general public has marked objections to animal research and many species are falling into disuse.The search for an ideal methodology to replace animal use is on,but animal modeling still holds great importance to human health.Bone research,in particular,has unmet requirements that in vitro technologies cannot fully address.Standardizing novel models remains necessary and rabbits are gaining in popularity as potential bone models.Our aim here is to provide a broad overview of animal modeling and its ethical implications,followed by a narrower focus on bone research and the role rabbits are playing in the current scenario.

Gastric cancer molecular classification based on immunohistochemistry and in situ hybridization:Analysis in western patients after curative-intent surgery

BACKGROUND Gastric cancer(GC)is a highly heterogeneous disease,and the identification of molecular subtyping of gastric adenocarcinoma emerged as a promising option to define therapeutic strategies and prognostic subgroups.However,the costs and technical complexity of molecular methodologies remains an obstacle to its adoption,and their clinical significance by other approaches needs further evidence.AIM To evaluate the clinicopathological characteristics and long-term survival of GC based on the subgroups of molecular classification by immunohistochemistry(IHC)and in situ hybridization(ISH).METHODS We retrospectively evaluated all patients who underwent D2-gastrectomy between 2009 and 2016 in a Western cohort of GC patients treated with curative intent.Microsatellite instability(MSI)status,E-cadherin,and p53 expression were analyzed by IHC,and Epstein-Barr virus(EBV)by ISH.Tissue microarrays were constructed for analysis.Clinicopathological characteristics and survival of GC were evaluated according to subtypes defined by The Cancer Genome Atlas(TCGA)Research Network Group and Asian Cancer Research Group(ACRG)classification systems.RESULTS A total of 287 GC patients were included.Based on IHC and ISH analysis,five profiles were defined as follows:E-cadherin aberrant(9.1%),MSI(20.9%),p53 aberrant(36.6%),EBV positivity(10.5%),and p53 normal(31%),which corresponded to tumors that showed no alteration in another profile.A flowchart according to the TCGA and ACRG classifications were used to define the subtypes,where clinical and pathological characteristics associated with GC subtypes were evidenced.Proximal location(P<0.001),total gastrectomy(P=0.001),and intense inflammatory infiltrate(P<0.001)were characteristics related to EBV subtype.MSI subtype was predominantly associated with advanced age(P=0.017)and the presence of comorbidities(P=0.011).While Laurén diffuse type(P<0.001)and advanced stage(P=0.029)were related to genomically stable(GS)subtype.GS tumors and microsatellite stable/epithelial to mesenchymal transition phenotype subtype had worse disease-free survival(DFS)and overall survival(OS)than other subtypes.Conversely,MSI subtype of GC had better survival in both classifications.Type of gastrectomy,pT and the TCGA subtypes were independent factors associated to DFS and OS.CONCLUSION The IHC/ISH analysis was able to distinguish immunophenotypic groups of GC with distinct characteristics and prognosis,resembling the subtypes of the molecular classifications.Accordingly,this method of classification may represent a viable option for use in a clinical setting.