| 1. |
- Ali, Muhaddisa Barat, 1986, et al.
(author)
-
Domain Mapping and Deep Learning from Multiple MRI Clinical Datasets for Prediction of Molecular Subtypes in Low Grade Gliomas
- 2020
-
In: Brain Sciences. - : MDPI AG. - 2076-3425. ; 10:7, s. 1-20
-
Journal article (peer-reviewed)abstract
- Brain tumors, such as low grade gliomas (LGG), are molecularly classified which require the surgical collection of tissue samples. The pre-surgical or non-operative identification of LGG molecular type could improve patient counseling and treatment decisions. However, radiographic approaches to LGG molecular classification are currently lacking, as clinicians are unable to reliably predict LGG molecular type using magnetic resonance imaging (MRI) studies. Machine learning approaches may improve the prediction of LGG molecular classification through MRI, however, the development of these techniques requires large annotated data sets. Merging clinical data from different hospitals to increase case numbers is needed, but the use of different scanners and settings can affect the results and simply combining them into a large dataset often have a significant negative impact on performance. This calls for efficient domain adaption methods. Despite some previous studies on domain adaptations, mapping MR images from different datasets to a common domain without affecting subtitle molecular-biomarker information has not been reported yet. In this paper, we propose an effective domain adaptation method based on Cycle Generative Adversarial Network (CycleGAN). The dataset is further enlarged by augmenting more MRIs using another GAN approach. Further, to tackle the issue of brain tumor segmentation that requires time and anatomical expertise to put exact boundary around the tumor, we have used a tight bounding box as a strategy. Finally, an efficient deep feature learning method, multi-stream convolutional autoencoder (CAE) and feature fusion, is proposed for the prediction of molecular subtypes (1p/19q-codeletion and IDH mutation). The experiments were conducted on a total of 161 patients consisting of FLAIR and T1 weighted with contrast enhanced (T1ce) MRIs from two different institutions in the USA and France. The proposed scheme is shown to achieve the test accuracy of 74.81% on 1p/19q codeletion and 81.19% on IDH mutation, with marked improvement over the results obtained without domain mapping. This approach is also shown to have comparable performance to several state-of-the-art methods.
|
|
| 2. |
- Ali, Muhaddisa Barat, 1986, et al.
(author)
-
Multi-stream Convolutional Autoencoder and 2D Generative Adversarial Network for Glioma Classification
- 2019
-
In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). - Cham : Springer International Publishing. - 1611-3349 .- 0302-9743. ; 11678 LNCS, s. 234-245
-
Conference paper (peer-reviewed)abstract
- Diagnosis and timely treatment play an important role in preventing brain tumor growth. Deep learning methods have gained much attention lately. Obtaining a large amount of annotated medical data remains a challenging issue. Furthermore, high dimensional features of brain images could lead to over-fitting. In this paper, we address the above issues. Firstly, we propose an architecture for Generative Adversarial Networks to generate good quality synthetic 2D MRIs from multi-modality MRIs (T1 contrast-enhanced, T2, FLAIR). Secondly, we propose a deep learning scheme based on 3-streams of Convolutional Autoencoders (CAEs) followed by sensor information fusion. The rational behind using CAEs is that it may improve glioma classification performance (as comparing with conventional CNNs), since CAEs offer noise robustness and also efficient feature reduction hence possibly reduce the over-fitting. A two-round training strategy is also applied by pre-training on GAN augmented synthetic MRIs followed by refined-training on original MRIs. Experiments on BraTS 2017 dataset have demonstrated the effectiveness of the proposed scheme (test accuracy 92.04%). Comparison with several exiting schemes has provided further support to the proposed scheme.
|
|
| 3. |
- Ali, Muhammad, 1990-, et al.
(author)
-
Integrated analysis of Shank1 PDZ interactions with C-terminal and internal binding motifs
- 2021
-
In: Current Research in Structural Biology. - : Elsevier. - 2665-928X. ; 3, s. 41-50
-
Journal article (peer-reviewed)abstract
- PDZ domains constitute a large family of modular domains that are well-known for binding C-terminal motifs of target proteins. Some of them also bind to internal PDZ binding motifs (PDZbms), but this aspect of the PDZ interactome is poorly studied. Here we explored internal PDZbm-mediated interactions using the PDZ domain of Shank1 as a model. We identified a series of human Shank1 ligands with C-terminal or internal PDZbms using proteomic peptide-phage display, and established that while the consensus sequence of C-terminal ligands is x-T-x-(L/F)-COOH, the consensus of internal PDZbm is exclusively x-T-x-F-x, where x is any amino acid. We found that the affinities of PDZbm interactions are in the low micromolar range. The crystal structure of the complex between Shank1 PDZ and an internal PDZbm revealed that the binding mode of internal PDZbms was similar to that of C-terminal ligands. Pull-down experiments confirmed that both C-terminal and internal PDZbm interactions can occur in the context of full-length proteins. Our study expands the interactome of Shank1 and hints at a largely unexplored interaction space of PDZ domains.
|
|
| 4. |
|
|
| 5. |
- Andersson, Linus, et al.
(author)
-
The influence of health-risk perception and distress on reactions to low-level chemical exposure
- 2013
-
In: Frontiers in Psychology. - : Frontiers Media SA. - 1664-1078. ; 4, s. 816-
-
Journal article (peer-reviewed)abstract
- The general aim of the current study was to investigate how perceived health risk of a chemical exposure and self-reported distress are related to perceived odor intensity and odor valence, symptoms, cognitive performance over time as well as reactions to blank exposure. Based on ratings of general distress, 20 participants constituted a relatively low distress group, and 20 other participants a relatively high distress group. Health risk perception was manipulated by providing positively and negatively biased information regarding n-butanol. Participants made repeated ratings of intensity, valence and symptoms and performed cognitive tasks while exposed to 4.7 ppm n-butanol for 60 min (first 10 min were blank exposure) inside an exposure chamber. Ratings by the positive and negative bias groups suggest that the manipulation influenced perceived health risk of the exposure. The high distress group did not habituate to the exposure in terms of intensity when receiving negative information, but did so when receiving positive information. The high distress group, compared with the low distress group, rated the exposure as significantly more unpleasant, reported greater symptoms and performed worse on a cognitively demanding task over time. The positive bias group and high distress group rated blank exposure as more intense. The main findings suggest that relatively distressed individuals are negatively affected by exposures to a greater degree than non-distressed.
|
|
| 6. |
- Arbiser, Jack L., et al.
(author)
-
Functional tyrosine kinase inhibitor profiling : a generally applicable method points to a novel role of platelet-derived growth factor receptor-beta in tuberous sclerosis
- 2002
-
In: American Journal of Pathology. - 0002-9440 .- 1525-2191. ; 161:3, s. 781-786
-
Journal article (peer-reviewed)abstract
- Tumors often exhibit activation of specific tyrosine kinases, which may allow targeting of therapy through inhibition of tyrosine kinase signaling. This strategy has been used successfully in the development of STI571 (gleevec), an inhibitor of bcr-abl tyrosine kinase that has been used successfully in the treatment of chronic myelogenous leukemia. STI571 also shows activity against c-kit and platelet-derived growth factor receptor-beta (PDGFRbeta) tyrosine kinase signaling, thus potentially expanding the number of tumors that may respond to it. We describe a simple and rapid method to assess functional activity of tyrosine kinase signaling that is broadly applicable to tumor types. As proof of principle, we have applied it to cells that serve as models of the autosomal-dominant tumor syndrome tuberous sclerosis (TS). We found that TS model cells derived from tuberin heterozygous mice and from a human renal angiomyolipoma are highly sensitive to PDGFR antagonists and that these cells express PDGFRbeta. Given that PDGFRbeta signaling is inhibited by STI571, we found that SV7tert human angiomyolipoma cells are sensitive to STI571. Thus, we describe a novel but simple method of determining the functional tyrosine kinase profile of a neoplastic cell and our results suggest that STI571 might be useful in the treatment of neoplasms commonly seen in patients with TS.
|
|
| 7. |
- Asplund, Kajsa
(author)
-
When profession trumps potential: The moderating role of professional identification in employees’ reactions to talent management
- 2020
-
In: International Journal of Human Resource Management. - : Taylor & Francis (Routledge): SSH Titles. - 1466-4399 .- 0958-5192. ; 31:4, s. 539-561
-
Journal article (peer-reviewed)abstract
- This study aimed to investigate the role that a professionalized context plays in shaping employee reactions to talent management decisions. We examined the mediating role of felt obligation in the relationship between talent ratings and organizational citizenship behavior. Further, the study tested whether professional identification moderates the relationship between talent ratings and felt obligation towards the organization. Five hundred and ninety-eight teachers that had recently received ratings of their talent status responded to a survey questionnaire. Felt obligation mediated the relationship between talent ratings and organizational citizenship behavior. Furthermore, professional identification moderated the relationship betweenratings of potential and felt obligation in such a way that the relationship was strongest for the teachers expressing the lowest professional identification. At high levels of professionalidentification, the relationship was not significant. These results indicate that conventional talent managementmight be less effective for increasing favorable attitudes and behaviors among employees in highly professionalized contexts, such as the education sector.
|
|
| 8. |
|
|
| 9. |
|
|
| 10. |
- Boberg, Elin, 1980-
(author)
-
Evolution of Spur Length in a Moth-pollinated Orchid
- 2010
-
Doctoral thesis (other academic/artistic)abstract
- There is considerable evidence that pollinator shifts can explain many differences in flower morphology between closely related plant species, but the extent to which pollinator shifts can explain the maintenance of among-population variation in floral traits within species is poorly known. In this thesis, I combined comparative and experimental approaches to examine the evolution of floral traits in the moth-pollinated orchid Platanthera bifolia. More specifically, I investigated (1) the relationship between flower and pollinator morphology, (2) phenotypic selection on morphology and phenology in populations in contrasting environments, (3) components of prezygotic reproductive isolation among divergent populations, and (4) the adaptive and functional significance of two correlated floral traits. A study of Scandinavian of P. bifolia populations revealed that spur length was positively correlated with proboscis length of local pollinators, which suggests that variation in spur length reflects adaptive evolution in response to geographically variable pollinator-mediated selection. A phenotypic selection study on Öland, SE Sweden, suggested that disruptive selection on spur length contributes to the maintenance of a bimodal distribution of spur length in mixed habitats, but provided very limited evidence of divergent selection on plant morphology and flowering phenology in grassland and woodland habitats. Field experiments revealed strong reproductive isolation between divergent populations on Öland, due to differences in spatial distribution, flowering phenology, and pollinators, and among-population incompatibility. The results suggest that prezygotic reproductive isolation contributes to the maintenance of population differentiation in floral traits in P. bifolia. A field manipulation experiment demonstrated that spur length but not perianth size affects pollination success and seed production. This suggests that among-population differentiation in perianth size may be the result of a genetic correlation with spur length. Taken together, the results of this thesis suggest that pollinator-mediated selection can shape the evolution of intraspecific floral variation.
|
|
