scholarly journals Optimisation of 2D U-Net Model Components for Automatic Prostate Segmentation on MRI

2020 ◽  
Vol 10 (7) ◽  
pp. 2601 ◽  
Author(s):  
Indriani P. Astono ◽  
James S. Welsh ◽  
Stephan Chalup ◽  
Peter Greer

In this paper, we develop an optimised state-of-the-art 2D U-Net model by studying the effects of the individual deep learning model components in performing prostate segmentation. We found that for upsampling, the combination of interpolation and convolution is better than the use of transposed convolution. For combining feature maps in each convolution block, it is only beneficial if a skip connection with concatenation is used. With respect to pooling, average pooling is better than strided-convolution, max, RMS or L2 pooling. Introducing a batch normalisation layer before the activation layer gives further performance improvement. The optimisation is based on a private dataset as it has a fixed 2D resolution and voxel size for every image which mitigates the need of a resizing operation in the data preparation process. Non-enhancing data preprocessing was applied and five-fold cross-validation was used to evaluate the fully automatic segmentation approach. We show it outperforms the traditional methods that were previously applied on the private dataset, as well as outperforming other comparable state-of-the-art 2D models on the public dataset PROMISE12.

PEDIATRICS ◽  
1979 ◽  
Vol 63 (3) ◽  
pp. 360-360
Author(s):  
M. L. Larson

The particular relation that professionalism bears to individualism and to the subjective illusion deserves to be noted. Their special competence empowers professionals and experts to act in situations where laymen feel incompetent or baffled. In fact, the assumption by the public that the expert is competent creates a sort of pragmatic compulsion for the expert: to certify his worth in the eyes of the laity, he must act. Deferentially requested to intervene by his clients, the expert practitioner is compelled to do something; from this point of view, anything is better than nothing. As Freidson remarks: "Indeed, so impressed is he by the perplexity of his clients and by his apparent capacity to deal with those perplexities, that the practitioner comes to consider himself an expert not only in the problems he is trained to deal with but in all human problems." Most particularly in the personal professions, the behavior of the expert asserts, ideologically, that a variety of ills—and, in particular, those that can most affect the person—have individual remedies. This reinforces the optimistic illusion of ideological individualism: personal problems of all kinds are purely private and admit, as such, individual and ad hoc solutions. In the predominant ideological way of addressing social issues and social relations experienced by individuals, therefore, structural causes, as well as collective action upon those causes, are relegated to a vaguely utopian realm. At the same time, the practitioner's "compulsion to act" reiterates to the layman that education confers superior powers upon the individual and superior mastery over physical and social environments.


Author(s):  
Bin Wang ◽  
Guojun Qi ◽  
Sheng Tang ◽  
Tianzhu Zhang ◽  
Yunchao Wei ◽  
...  

Semantic segmentation suffers from the fact that densely annotated masks are expensive to obtain. To tackle this problem, we aim at learning to segment by only leveraging scribbles that are much easier to collect for supervision. To fully explore the limited pixel-level annotations from scribbles, we present a novel Boundary Perception Guidance (BPG) approach, which consists of two basic components, i.e., prediction refinement and boundary regression. Specifically, the prediction refinement progressively makes a better segmentation by adopting an iterative upsampling and a semantic feature  enhancement strategy. In the boundary regression, we employ class-agnostic edge maps for supervision to effectively guide the segmentation network in localizing the boundaries between different semantic regions, leading to producing finer-grained representation of feature maps for semantic segmentation. The experiment results on the PASCAL VOC 2012 demonstrate the proposed BPG achieves mIoU of 73.2% without fully connected Conditional Random Field (CRF) and 76.0% with CRF, setting up the new state-of-the-art in literature.


2020 ◽  
Vol 34 (05) ◽  
pp. 8472-8479
Author(s):  
Saurav Manchanda ◽  
George Karypis

Credit attribution is the task of associating individual parts in a document with their most appropriate class labels. It is an important task with applications to information retrieval and text summarization. When labeled training data is available, traditional approaches for sequence tagging can be used for credit attribution. However, generating such labeled datasets is expensive and time-consuming. In this paper, we present Credit Attribution With Attention (CAWA), a neural-network-based approach, that instead of using sentence-level labeled data, uses the set of class labels that are associated with an entire document as a source of distant-supervision. CAWA combines an attention mechanism with a multilabel classifier into an end-to-end learning framework to perform credit attribution. CAWA labels the individual sentences from the input document using the resultant attention-weights. CAWA improves upon the state-of-the-art credit attribution approach by not constraining a sentence to belong to just one class, but modeling each sentence as a distribution over all classes, leading to better modeling of semantically-similar classes. Experiments on the credit attribution task on a variety of datasets show that the sentence class labels generated by CAWA outperform the competing approaches. Additionally, on the multilabel text classification task, CAWA performs better than the competing credit attribution approaches1.


2020 ◽  
Vol 10 (16) ◽  
pp. 5701
Author(s):  
Zhijun Gao ◽  
Xingle Wang ◽  
Yi Li

The number and volume of retinal macular edemas are important indicators for screening and diagnosing retinopathy. Aiming at the problem that the segmentation method of macular edemas in a retinal optical coherence tomography (OCT) image is not ideal in segmentation of diverse edemas, this paper proposes a new method of automatic segmentation of macular edema regions in retinal OCT images using the improved U-Net++. The proposed method makes full use of the U-Net++ re-designed skip pathways and dense convolution block; reduces the semantic gap of the feature maps in the encoder/decoder sub-network; and adds the improved Resnet network as the backbone, which make the extraction of features in the edema regions more accurate and improves the segmentation effect. The proposed method was trained and validated on the public dataset of Duke University, and the experiments demonstrated the proposed method can not only improve the overall segmentation effect, but also can significantly improve the segmented precision for diverse edema in multi-regions, as well as reducing the error of the number of edema regions.


Sensors ◽  
2021 ◽  
Vol 21 (15) ◽  
pp. 4986
Author(s):  
Bai Zhao ◽  
Xiaolin Gong ◽  
Jian Wang ◽  
Lingchao Zhao

Due to the non-uniform illumination conditions, images captured by sensors often suffer from uneven brightness, low contrast and noise. In order to improve the quality of the image, in this paper, a multi-path interaction network is proposed to enhance the R, G, B channels, and then the three channels are combined into the color image and further adjusted in detail. In the multi-path interaction network, the feature maps in several encoding–decoding subnetworks are used to exchange information across paths, while a high-resolution path is retained to enrich the feature representation. Meanwhile, in order to avoid the possible unnatural results caused by the separation of the R, G, B channels, the output of the multi-path interaction network is corrected in detail to obtain the final enhancement results. Experimental results show that the proposed method can effectively improve the visual quality of low-light images, and the performance is better than the state-of-the-art methods.


2019 ◽  
Vol 2 (2) ◽  
pp. 1-7
Author(s):  
Andi Samsu Rijal ◽  
Andi Mega Januarti Putri

The essence of language is human activity. Communication with language is carried out through two basic human activities; speaking and listening during the interaction in a group of people. Immigrants in Makassar city communicate with immigrant communities and Makassar people. They used English and Indonesia to communicate with others. The aims of this article were to find out determinant factors of English as language choice among Unaccompanied Migrant Children (UMC) in Makassar and why they used English as their language choice to communicate with other people out of them. The data were taken from UMC in the shelter under the auspices of Makassar’s Social Office and in the public area of Makassar. This research was a qualitative approach; it was from a sociolinguistic perspective and focuses its analysis with the language choice among UMC. This research showed that most immigrants chose English as their language choice since they were in Makassar because they have acquired better than other international language and it has been mastered naturally by doing social interaction among themselves and people outside their community. UMC had more difficulties to socialize with Indonesian than the adult of Immigrants. Other than their lack of language mastery, they also have the anxiety to adapt to other immigrants and Makassar people. English was used by UMC to show their status as a foreigner who lived in a multicultural situation. Language becomes a power for a human being and it becomes a social identity for language user in one community. During the interaction of UMC in Makassar city, the role of English as an International language is shown.


Author(s):  
Maxim B. Demchenko ◽  

The sphere of the unknown, supernatural and miraculous is one of the most popular subjects for everyday discussions in Ayodhya – the last of the provinces of the Mughal Empire, which entered the British Raj in 1859, and in the distant past – the space of many legendary and mythological events. Mostly they concern encounters with inhabitants of the “other world” – spirits, ghosts, jinns as well as miraculous healings following magic rituals or meetings with the so-called saints of different religions (Hindu sadhus, Sufi dervishes),with incomprehensible and frightening natural phenomena. According to the author’s observations ideas of the unknown in Avadh are codified and structured in Avadh better than in other parts of India. Local people can clearly define if they witness a bhut or a jinn and whether the disease is caused by some witchcraft or other reasons. Perhaps that is due to the presence in the holy town of a persistent tradition of katha, the public presentation of plots from the Ramayana epic in both the narrative and poetic as well as performative forms. But are the events and phenomena in question a miracle for the Avadhvasis, residents of Ayodhya and its environs, or are they so commonplace that they do not surprise or fascinate? That exactly is the subject of the essay, written on the basis of materials collected by the author in Ayodhya during the period of 2010 – 2019. The author would like to express his appreciation to Mr. Alok Sharma (Faizabad) for his advice and cooperation.


Public Voices ◽  
2016 ◽  
Vol 12 (1) ◽  
pp. 67 ◽  
Author(s):  
Sharon Mastracci

In this paper, the author examines public service as depicted in the television series Buffy the Vampire Slayer (BtVS). First, she shows how slaying meets the economist’s definition of a public good, using the BtVS episode “Flooded” (6.04). Second, she discusses public service motivation (PSM) to determine whether or not Buffy, a public servant, operates from a public service ethic. Relying on established measures and evidence from shooting scripts and episode transcripts, the author concludes Buffy is a public servant motivated by a public service ethic. In this way, BtVS informs scholarship on public service by broadening the concept of PSM beyond the public sector; prompting one to wonder whether it is located in a sector, an occupation, or in the individual. These conclusions allow the author to situate Buffy alongside other idealized public servants in American popular culture.


Author(s):  
Andrew M. Yuengert

Although most economists are skeptical of or puzzled by the Catholic concept of the common good, a rejection of the economic approach as inimical to the common good would be hasty and counterproductive. Economic analysis can enrich the common good tradition in four ways. First, economics embodies a deep respect for economic agency and for the effects of policy and institutions on individual agents. Second, economics offers a rich literature on the nature of unplanned order and how it might be shaped by policy. Third, economics offers insight into the public and private provision of various kinds of goods (private, public, common pool resources). Fourth, recent work on the development and logic of institutions and norms emphasizes sustainability rooted in the good of the individual.


Sign in / Sign up

Export Citation Format

Share Document