University of Paris

About: University of Paris is a education organization based out in Paris, France. It is known for research contribution in the topics: Population & Transplantation. The organization has 102426 authors who have published 174180 publications receiving 5041753 citations. The organization is also known as: Sorbonne.

Benchmark Analysis of Representative Deep Neural Network Architectures

Abstract: This paper presents an in-depth analysis of the majority of the deep neural networks (DNNs) proposed in the state of the art for image recognition. For each DNN, multiple performance indices are observed, such as recognition accuracy, model complexity, computational complexity, memory usage, and inference time. The behavior of such performance indices and some combinations of them are analyzed and discussed. To measure the indices, we experiment the use of DNNs on two different computer architectures, a workstation equipped with a NVIDIA Titan X Pascal, and an embedded system based on a NVIDIA Jetson TX1 board. This experimentation allows a direct comparison between DNNs running on machines with very different computational capacities. This paper is useful for researchers to have a complete view of what solutions have been explored so far and in which research directions are worth exploring in the future, and for practitioners to select the DNN architecture(s) that better fit the resource constraints of practical deployments and applications. To complete this work, all the DNNs, as well as the software used for the analysis, are available online.

Use of prostate systematic and targeted biopsy on the basis of multiparametric MRI in biopsy-naive patients (MRI-FIRST): a prospective, multicentre, paired diagnostic study.

Abstract: Summary Background Whether multiparametric MRI improves the detection of clinically significant prostate cancer and avoids the need for systematic biopsy in biopsy-naive patients remains controversial. We aimed to investigate whether using this approach before biopsy would improve detection of clinically significant prostate cancer in biopsy-naive patients. Methods In this prospective, multicentre, paired diagnostic study, done at 16 centres in France, we enrolled patients aged 18–75 years with prostate-specific antigen concentrations of 20 ng/mL or less, and with stage T2c or lower prostate cancer. Eligible patients had been referred for prostate multiparametric MRI before a first set of prostate biopsies, with a planned interval of less than 3 months between MRI and biopsies. An operator masked to multiparametric MRI results did a systematic biopsy by obtaining 12 systematic cores and up to two cores targeting hypoechoic lesions. In the same patient, another operator targeted up to two lesions seen on MRI with a Likert score of 3 or higher (three cores per lesion) using targeted biopsy based on multiparametric MRI findings. Patients with negative multiparametric MRI (Likert score ≤2) had systematic biopsy only. The primary outcome was the detection of clinically significant prostate cancer of International Society of Urological Pathology grade group 2 or higher (csPCa-A), analysed in all patients who received both systematic and targeted biopsies and whose results from both were available for pathological central review, including patients who had protocol deviations. This study is registered with ClinicalTrials.gov, number NCT02485379, and is closed to new participants. Findings Between July 15, 2015, and Aug 11, 2016, we enrolled 275 patients. 24 (9%) were excluded from the analysis. 53 (21%) of 251 analysed patients had negative (Likert ≤2) multiparametric MRI. csPCa-A was detected in 94 (37%) of 251 patients. 13 (14%) of these 94 patients were diagnosed by systematic biopsy only, 19 (20%) by targeted biopsy only, and 62 (66%) by both techniques. Detection of csPCa-A by systematic biopsy (29·9%, 95% CI 24·3–36·0) and targeted biopsy (32·3%, 26·5–38·4) did not differ significantly (p=0·38). csPCa-A would have been missed in 5·2% (95% CI 2·8–8·7) of patients had systematic biopsy not been done, and in 7·6% (4·6–11·6) of patients had targeted biopsy not been done. Four grade 3 post-biopsy adverse events were reported (3 cases of prostatitis, and 1 case of urinary retention with haematuria). Interpretation There was no difference between systematic biopsy and targeted biopsy in the detection of ISUP grade group 2 or higher prostate cancer; however, this detection was improved by combining both techniques and both techniques showed substantial added value. Thus, obtaining a multiparametric MRI before biopsy in biopsy-naive patients can improve the detection of clinically significant prostate cancer but does not seem to avoid the need for systematic biopsy. Funding French National Cancer Institute.

CHAPTER 1 – Finite Automata

Abstract: Publisher Summary This chapter presents the field of finite automata, which is a branch of mathematics connected with the algebraic theory of semigroups and associative algebras. From another viewpoint, finite automata are a branch of algorithm design concerned with string manipulation and sequence processing. The chapter presents the basic definitions and a proof of the fundamental result of S. C. Kleene. The chapter discusses the notion of star-height, which is one of the fast developing areas of automata theory. The chapter discusses star-free sets, which is an important notion especially because of its connection with logic. The chapter presents a complete proof of P. Schutzenberger's theorem, stating the equality between star-free sets and aperiodic sets. It also presents the syntactic characterization of two important subfamilies of star-free sets, namely, locally testable and piecewise testable. The chapter describes the applications of finite automata, such as string matching or file indexing. The chapter introduces the field of automata recognizing numbers expanded at some basis. This is an aspect of finite automata regarding several fields in classical mathematics, such as number theory and ergodic theory.

Occurrence of brown adipocytes in rat white adipose tissue: molecular and morphological characterization.

Abstract: Brown adipocytes are thermogenic cells which play an important role in energy balance. Their thermogenic activity is due to the presence of a mitochondrial uncoupling protein (UCP). Until recently, it was admitted that in rodents brown adipocytes were mainly located in classical brown adipose tissue (BAT). In the present study, we have investigated the presence of UCP protein or mRNA in white adipose tissue (WAT) of rats. Using polymerase chain reaction or Northern blot hybridization, UCP mRNA was detected in mesenteric, epidydimal, retroperitoneal, inguinal and particularly in periovarian adipose depots. The uncoupling protein was detected by Western blotting in mitochondria from periovarian adipose tissue. When rats were submitted to cold or to treatment with a beta-adrenoceptor agonist, UCP expression was increased in this tissue as in typical brown fat. Moreover, the expression was decreased in obese fa/fa rats compared to lean controls. Morphological studies showed that periovarian adipose tissue of rats kept at 24 degrees C contained cells with numerous typical BAT mitochondria with or without multilocular lipid droplets. Immunocytochemistry confirmed that multilocular cells expressed mitochondrial UCP. Furthermore, the number of brown adipocytes and the density of mitochondrial cristae increased in parallel with exposure to cold. These results demonstrate that adipocytes expressing UCP are present in adipose deposits considered as white fat. They suggest the existence of a continuum in rodents between BAT and WAT, and a great plasticity between adipose tissue phenotypes. The physiological importance of brown adipocytes in WAT and the regulation of UCP expression remain open questions.