1,720,978 research outputs found
Tune It or Don't Use It: Benchmarking Data-Efficient Image Classification
Full text linkData-efficient image classification using deep neural networks in settings, where only small amounts of labeled data are available, has been an active research area in the recent past. However, an objective comparison between published methods is difficult, since existing works use different datasets for evaluation and often compare against un-tuned baselines with default hyper-parameters. We design a benchmark for data-efficient image classification consisting of six diverse datasets spanning various domains (e.g., natural images, medical imagery, satellite data) and data types (RGB, grayscale, multispectral). Using this benchmark, we re-evaluate the standard cross-entropy baseline and eight methods for data-efficient deep learning published between 2017 and 2021 at renowned venues. For a fair and realistic comparison, we carefully tune the hyper-parameters of all methods on each dataset. Surprisingly, we find that tuning learning rate, weight decay, and batch size on a separate validation split results in a highly competitive baseline, which outperforms all but one specialized method and performs competitively to the remaining one
Image Classification With Small Datasets: Overview and Benchmark
Full text linkImage classification with small datasets has been an active research area in the recent past. However, as research in this scope is still in its infancy, two key ingredients are missing for ensuring reliable and truthful progress: a systematic and extensive overview of the state of the art, and a common benchmark to allow for objective comparisons between published methods. This article addresses both issues. First, we systematically organize and connect past studies to consolidate a community that is currently fragmented and scattered. Second, we propose a common benchmark that allows for an objective comparison of approaches. It consists of five datasets spanning various domains (e.g., natural images, medical imagery, satellite data) and data types (RGB, grayscale, multispectral). We use this benchmark to re-evaluate the standard cross-entropy baseline and ten existing methods published between 2017 and 2021 at renowned venues. Surprisingly, we find that thorough hyper-parameter tuning on held-out validation data results in a highly competitive baseline and highlights a stunted growth of performance over the years. Indeed, only a single specialized method dating back to 2019 clearly wins our benchmark and outperforms the baseline classifier
[Premature membrane rupture. Comparison of diagnostic tests].
No full textBACKGROUND:
Objective of this study was to evaluate the accuracy of the vaginal pH-test, the Fern-test, the research of foetal cells and of foetal fibronectin in vaginal discharge, which are used to diagnose premature rupture of membranes.
METHODS:
To this aim 40 pregnant patients between 24th and 37th weeks gestation have been examined, considered at risk for sub-clinic loss of aminiotic fluid: 23 were affected by preterm labour and 17 by suspected rupture of membranes.
RESULTS:
Subsequently amniotic sac was confirmed to be ripped in 10 cases (25%): 2 (8.7%) in the 23 patients with preterm labour, and 8 (47%) in the 17 patients with suspected PROM. Sensibility, specificity and accuracy were respectively: 70, 97 and 90% for pH-test; 70, 100 and 93% for Fern-test; 50, 93 and 82% for foetal cells; 100, 90 and 93% for fibronectin test.
CONCLUSIONS:
In personal experience fibronectin test appeared to be the most sensible and accurate marker. Fern-test was the most specific, while the research of foetal cells appeared to be the least reliable
Glucose metabolic alterations in isolated and perfused rat hepatocytes induced by pancreatic cancer conditioned medium: a low molecular weight factor possibly involved.
No full textA serious insulin resistance characterizes pancreatic cancer-associated diabetes mellitus. Elsewhere, we demonstrated that MIA PaCa2 cultured cells secrete a soluble factor responsible for reduced glucose tolerance induced in SCID mice. The intracellular mechanism of insulin resistance was investigated in isolated and perfused rat hepatocytes incubated with MIA PaCa2 conditioned medium. Lactate production was reduced compared to hepatocytes incubated with control medium while 1,2-DAG was increased and PKC was activated in the hepatocytes incubated with MIA PaCa2 conditioned medium. This behavior was not reproduced treating the hepatocytes with the growth factors EGF, interleukin Ibeta, interleukin-6, and TGF-beta1. In an attempt to make a biochemical identification of the hypothesized tumor associated-diabetogenic factors we observed a low molecular weight protein in the conditioned medium, absent in the nonconditioned one, that may be responsible for the described behaviors
Glucose metabolic alterations in isolated and perfused rat hepatocytes induced by pancreatic cancer conditioned medium: A low molecular weight factor possibly involved
No full textGoing Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Pancreatic cancer associated diebetogenic factor:a low molecular weight peptide possibly involved.
No full textThe pancreatic cancer cell line MIA PaCa2 produces one or more factors able to induce hyperglycemia in SCID mice
No full textA reduced glucose tolerance or frank diabetes mellitus is a frequent finding in patients with pancreatic cancer. The aim of this study was to verify whether the pancreatic cancer cell line MIA PaCa2 was able to produce any factor which could induce hyperglycemia in SCID (severe complete immunodeficient) mice. MIA PaCa2 cells were cultured in Dulbecco's modified Eagle's medium (DMEM) for 7 days. Twenty-five female SCID mice were used. They were daily i.p. injected with 300 ul of cell culture supernatants (Group T, n = 13) or with 300 ul of DMEM (Group C, n = 12) and followed up for 82 days. Blood glucose levels were significantly higher in Group T than in Group C on days 10 and 25. Intravenous glucose tolerance test, success-fully performed in 9 animals (4 controls and 5 treated), demonstrated a significantly reduced glucose tolerance in Group T compared to Group C mice. At sacrifice, plasma and pancreatic insulin and glucagon levels did not vary between groups. The ratio between pancreatic and plasma insulin was significantly lower in Group T than in Group C. We conclude that: 1. The pancreatic cancer cell line MIA PaCa2 produces one or more soluble factors able to cause hyperglycemia in vivo; 2. this effect is not immunologically mediated, and 3. this/these factor/s could both interfere with the pancreatic beta cells and/or with insulin peripheral action
- …
