| 1. |
- Thambawita, V., et al.
(författare)
-
SinGAN-Seg: Synthetic training data generation for medical image segmentation
- 2022
-
Ingår i: PLoS ONE. - : Public Library of Science (PLoS). - 1932-6203. ; 17:5 May
-
Tidskriftsartikel (refereegranskat)abstract
- Analyzing medical data to find abnormalities is a time-consuming and costly task, particularly for rare abnormalities, requiring tremendous efforts from medical experts. Therefore, artificial intelligence has become a popular tool for the automatic processing of medical data, acting as a supportive tool for doctors. However, the machine learning models used to build these tools are highly dependent on the data used to train them. Large amounts of data can be difficult to obtain in medicine due to privacy reasons, expensive and time-consuming annotations, and a general lack of data samples for infrequent lesions. In this study, we present a novel synthetic data generation pipeline, called SinGAN-Seg, to produce synthetic medical images with corresponding masks using a single training image. Our method is different from the traditional generative adversarial networks (GANs) because our model needs only a single image and the corresponding ground truth to train. We also show that the synthetic data generation pipeline can be used to produce alternative artificial segmentation datasets with corresponding ground truth masks when real datasets are not allowed to share. The pipeline is evaluated using qualitative and quantitative comparisons between real data and synthetic data to show that the style transfer technique used in our pipeline significantly improves the quality of the generated data and our method is better than other state-of-the-art GANs to prepare synthetic images when the size of training datasets are limited. By training UNet++ using both real data and the synthetic data generated from the SinGAN-Seg pipeline, we show that the models trained on synthetic data have very close performances to those trained on real data when both datasets have a considerable amount of training data. In contrast, we show that synthetic data generated from the SinGAN-Seg pipeline improves the performance of segmentation models when training datasets do not have a considerable amount of data. All experiments were performed using an open dataset and the code is publicly available on GitHub.
|
|
| 2. |
- Wäneskog, Marcus, et al.
(författare)
-
Escherichia coli EC93 deploys two plasmid- encoded class I contact- dependent growth inhibition systems for antagonistic bacterial interactions
- 2021
-
Ingår i: Microbial Genomics. - : MICROBIOLOGY SOC. - 2057-5858. ; 7:3
-
Tidskriftsartikel (refereegranskat)abstract
- The phenomenon of contact- dependent growth inhibition (CDI) and the genes required for CDI (cdiBAI) were identified and isolated in 2005 from an Escherichia coli isolate (EC93) from rats. Although the cdiBAIEC93 locus has been the focus of extensive research during the past 15 years, little is known about the EC93 isolate from which it originates. Here we sequenced the EC93 genome and find two complete and functional cdiBAI loci (including the previously identified cdi locus), both carried on a large 127 kb plasmid. These cdiBAI systems are differentially expressed in laboratory media, enabling EC93 to outcompete E. coli cells lacking cognate cdiI immunity genes. The two CDI systems deliver distinct effector peptides that each dissipate the membrane potential of target cells, although the two toxins display different toxic potencies. Despite the differential expression and toxic potencies of these CDI systems, both yielded similar competitive advantages against E. coli cells lacking immunity. This can be explained by the fact that the less expressed cdiBAI system (cdiBAIEC93-2) delivers a more potent toxin than the highly expressed cdiBAIEC93-1 system. Moreover, our results indicate that unlike most sequenced CDI+ bacterial isolates, the two cdi loci of E. coli EC93 are located on a plasmid and are expressed in laboratory media.
|
|
| 3. |
- Skogeng, Lovise P., et al.
(författare)
-
Spatial distribution of Dechlorane Plus and dechlorane related compounds in European background air
- 2023
-
Ingår i: Frontiers in Environmental Science. - : Frontiers Media SA. - 2296-665X. ; 10
-
Tidskriftsartikel (refereegranskat)abstract
- The highly chlorinated chemical Dechlorane Plus (DP) was introduced as a replacement flame retardant for Mirex, which is banned through the Stockholm Convention (SC) for its toxicity (T), environmental persistence (P), potential for bioaccumulation (B) and long-range environmental transport potential (LRETP). Currently, Dechlorane Plus is under consideration for listing under the Stockholm Convention and by the European Chemical Agency as it is suspected to also have potential for P, B, T and LRET. Knowledge of atmospheric concentrations of chemicals in background regions is vital to understand their persistence and long-range atmospheric transport but such knowledge is still limited for Dechlorane Plus. Also, knowledge on environmental occurrence of the less described Dechlorane Related Compounds (DRCs), with similar properties and uses as Dechlorane Plus, is limited. Hence, the main objective of this study was to carry out a spatial mapping of atmospheric concentrations of Dechlorane Plus and Dechlorane Related Compounds at background sites in Europe. Polyurethane foam passive air samplers were deployed at 99 sites across 33 European countries for 3 months in summer 2016 and analyzed for dechloranes. The study showed that syn- and anti-DP are present across the European continent (3 and 3, respectively), including parts of the Arctic. This supports that these compounds have potential for long-range atmospheric transport to remote regions. The highest concentrations of Dechlorane Plus were observed in central continental Europe, with anti-DP fractions close to the commercial mixture of Dechlorane Plus. The only detected Dechlorane Related Compounds was Dechlorane-602, which was found in 27% of the samples (3). The measured concentrations and spatial patterns of Dechlorane Plus and Dechlorane-602 in air across Europe indicate the influence of primary sources of these compounds on background concentrations in European air. Future air monitoring efforts targeting dechloranes is needed in both background and source areas, including consistent temporal trends.
|
|
| 4. |
- Jansson, E., et al.
(författare)
-
Global, regional, and cryptic population structure in a high gene-flow transatlantic fish
- 2023
-
Ingår i: Plos One. - : Public Library of Science (PLoS). - 1932-6203. ; 18:3
-
Tidskriftsartikel (refereegranskat)abstract
- Lumpfish (Cyclopterus lumpus) is a transatlantic marine fish displaying large population sizes and a high potential for dispersal and gene-flow. These features are expected to result in weak population structure. Here, we investigated population genetic structure of lumpfish throughout its natural distribution in the North Atlantic using two approaches: I) 4,393 genome wide SNPs and 95 individuals from 10 locations, and II) 139 discriminatory SNPs and 1,669 individuals from 40 locations. Both approaches identified extensive population genetic structuring with a major split between the East and West Atlantic and a distinct Baltic Sea population, as well as further differentiation of lumpfish from the English Channel, Iceland, and Greenland. The discriminatory loci displayed similar to 2-5 times higher divergence than the genome wide approach, revealing further evidence of local population substructures. Lumpfish from Isfjorden in Svalbard were highly distinct but resembled most fish from Greenland. The Kattegat area in the Baltic transition zone, formed a previously undescribed distinct genetic group. Also, further subdivision was detected within North America, Iceland, West Greenland, Barents Sea, and Norway. Although lumpfish have considerable potential for dispersal and gene-flow, the observed high levels of population structuring throughout the Atlantic suggests that this species may have a natal homing behavior and local populations with adaptive differences. This fine-scale population structure calls for consideration when defining management units for exploitation of lumpfish stocks and in decisions related to sourcing and moving lumpfish for cleaner fish use in salmonid aquaculture.
|
|
