| 1. |
- Ge, Chenjie, 1991, et al.
(author)
-
Enlarged Training Dataset by Pairwise GANs for Molecular-Based Brain Tumor Classification
- 2020
-
In: IEEE Access. - 2169-3536 .- 2169-3536. ; 8:1, s. 22560-22570
-
Journal article (peer-reviewed)abstract
- This paper addresses issues of brain tumor subtype classification using Magnetic Resonance Images (MRIs) from different scanner modalities like T1 weighted, T1 weighted with contrast-enhanced, T2 weighted and FLAIR images. Currently most available glioma datasets are relatively moderate in size, and often accompanied with incomplete MRIs in different modalities. To tackle the commonly encountered problems of insufficiently large brain tumor datasets and incomplete modality of image for deep learning, we propose to add augmented brain MR images to enlarge the training dataset by employing a pairwise Generative Adversarial Network (GAN) model. The pairwise GAN is able to generate synthetic MRIs across different modalities. To achieve the patient-level diagnostic result, we propose a post-processing strategy to combine the slice-level glioma subtype classification results by majority voting. A two-stage course-to-fine training strategy is proposed to learn the glioma feature using GAN-augmented MRIs followed by real MRIs. To evaluate the effectiveness of the proposed scheme, experiments have been conducted on a brain tumor dataset for classifying glioma molecular subtypes: isocitrate dehydrogenase 1 (IDH1) mutation and IDH1 wild-type. Our results on the dataset have shown good performance (with test accuracy 88.82%). Comparisons with several state-of-the-art methods are also included.
|
|
| 2. |
|
|
| 3. |
- Ali, Muhaddisa Barat, 1986, et al.
(author)
-
A novel federated deep learning scheme for glioma and its subtype classification
- 2023
-
In: Frontiers in Neuroscience. - 1662-4548 .- 1662-453X. ; 17
-
Journal article (peer-reviewed)abstract
- Background: Deep learning (DL) has shown promising results in molecular-based classification of glioma subtypes from MR images. DL requires a large number of training data for achieving good generalization performance. Since brain tumor datasets are usually small in size, combination of such datasets from different hospitals are needed. Data privacy issue from hospitals often poses a constraint on such a practice. Federated learning (FL) has gained much attention lately as it trains a central DL model without requiring data sharing from different hospitals. Method: We propose a novel 3D FL scheme for glioma and its molecular subtype classification. In the scheme, a slice-based DL classifier, EtFedDyn, is exploited which is an extension of FedDyn, with the key differences on using focal loss cost function to tackle severe class imbalances in the datasets, and on multi-stream network to exploit MRIs in different modalities. By combining EtFedDyn with domain mapping as the pre-processing and 3D scan-based post-processing, the proposed scheme makes 3D brain scan-based classification on datasets from different dataset owners. To examine whether the FL scheme could replace the central learning (CL) one, we then compare the classification performance between the proposed FL and the corresponding CL schemes. Furthermore, detailed empirical-based analysis were also conducted to exam the effect of using domain mapping, 3D scan-based post-processing, different cost functions and different FL schemes. Results: Experiments were done on two case studies: classification of glioma subtypes (IDH mutation and wild-type on TCGA and US datasets in case A) and glioma grades (high/low grade glioma HGG and LGG on MICCAI dataset in case B). The proposed FL scheme has obtained good performance on the test sets (85.46%, 75.56%) for IDH subtypes and (89.28%, 90.72%) for glioma LGG/HGG all averaged on five runs. Comparing with the corresponding CL scheme, the drop in test accuracy from the proposed FL scheme is small (−1.17%, −0.83%), indicating its good potential to replace the CL scheme. Furthermore, the empirically tests have shown that an increased classification test accuracy by applying: domain mapping (0.4%, 1.85%) in case A; focal loss function (1.66%, 3.25%) in case A and (1.19%, 1.85%) in case B; 3D post-processing (2.11%, 2.23%) in case A and (1.81%, 2.39%) in case B and EtFedDyn over FedAvg classifier (1.05%, 1.55%) in case A and (1.23%, 1.81%) in case B with fast convergence, which all contributed to the improvement of overall performance in the proposed FL scheme. Conclusion: The proposed FL scheme is shown to be effective in predicting glioma and its subtypes by using MR images from test sets, with great potential of replacing the conventional CL approaches for training deep networks. This could help hospitals to maintain their data privacy, while using a federated trained classifier with nearly similar performance as that from a centrally trained one. Further detailed experiments have shown that different parts in the proposed 3D FL scheme, such as domain mapping (make datasets more uniform) and post-processing (scan-based classification), are essential.
|
|
| 4. |
- Khoshnood, Ardavan
(author)
-
Prehospital Diagnosis and Oxygen Treatment in ST Elevation Myocardial Infarction
- 2017
-
Doctoral thesis (other academic/artistic)abstract
- IntroductionPaper I: An Artificial Neural Network (ANN) was constructed to identify ST Elevation Myocardial Infarction (STEMI) and predict the need for Percutaneous Coronary Intervention (PCI). Paper II, III and IV: Studies suggest that O2 therapy may be harmful in STEMI patients. We therefore conducted the SOCCER study to evaluate the effects of O2 therapy in STEMI patients.MethodsPaper I: 560 ambulance ECGs sent to the Cardiac Care Unit (CCU), was together with the CCU physicians interpretation and decision of conducting an acute PCI or not collected, and compared with the interpretation and PCI decision of the ANN. Paper II, III, IV: Normoxic (≥94%) STEMI patients accepted for acute PCI were in the ambulance randomized to standard care with 10 L/min O2 or room air. A subset of the patients underwent echocardiography for determination of the Left Ventricular Ejection Fraction (LVEF) and the Wall Motion Score Index (WMSI). All patients had a Cardiac Magnetic Resonance Imaging (CMRI) to evaluate Myocardial area at Risk (MaR), Infarct Size (IS) and Myocardial Salvage Index (MSI).ResultsPaper I: The area under the ANN’s receiver operating characteristics curve for STEMI detection as well as predicting the need of acute PCI were very good.Paper II, III, IV: No significant differences could be shown in discussing MaR, MSI or IS between the O2 group (n=46) and the air group (n=49). Neither could any differences be shown for LVEF and WMSI at the index visit as well after six months between the O2 group (n=46) and the air group (n=41)ConclusionsPaper I: The results indicate that the number of ECGs sent to the CCU could be reduced with 2/3 as the ANN would safely identify ECGs not being STEMI.Paper II, III, IV: The results suggest that it is safe to withhold O2 therapy in normoxic, stable STEMI patients.
|
|
| 5. |
- Ahlander, Britt-Marie, 1954-, et al.
(author)
-
An echo-planar imaging sequence is superior to a steady-state free precession sequence for visual as well as quantitative assessment of cardiac magnetic resonance stress perfusion
- 2017
-
In: Clinical Physiology and Functional Imaging. - : Blackwell Publishing. - 1475-0961 .- 1475-097X. ; 37:1, s. 52-61
-
Journal article (peer-reviewed)abstract
- BACKGROUND: To assess myocardial perfusion, steady-state free precession cardiac magnetic resonance (SSFP, CMR) was compared with gradient-echo-echo-planar imaging (GRE-EPI) using myocardial perfusion scintigraphy (MPS) as reference.METHODS: Cardiac magnetic resonance perfusion was recorded in 30 patients with SSFP and in another 30 patients with GRE-EPI. Timing and extent of inflow delay to the myocardium was visually assessed. Signal-to-noise (SNR) and contrast-to-noise (CNR) ratios were calculated. Myocardial scar was visualized with a phase-sensitive inversion recovery sequence (PSIR). All scar positive segments were considered pathologic. In MPS, stress and rest images were used as in clinical reporting. The CMR contrast wash-in slope was calculated and compared with the stress score from the MPS examination. CMR scar, CMR perfusion and MPS were assessed separately by one expert for each method who was blinded to other aspects of the study.RESULTS: Visual assessment of CMR had a sensitivity for the detection of an abnormal MPS at 78% (SSFP) versus 91% (GRE-EPI) and a specificity of 58% (SSFP) versus 84% (GRE-EPI). Kappa statistics for SSFP and MPS was 0·29, for GRE-EPI and MPS 0·72. The ANOVA of CMR perfusion slopes for all segments versus MPS score (four levels based on MPS) had correlation r = 0·64 (SSFP) and r = 0·96 (GRE-EPI). SNR was for normal segments 35·63 ± 11·80 (SSFP) and 17·98 ± 8·31 (GRE-EPI), while CNR was 28·79 ± 10·43 (SSFP) and 13·06 ± 7·61 (GRE-EPI).CONCLUSION: GRE-EPI displayed higher agreement with the MPS results than SSFP despite significantly lower signal intensity, SNR and CNR.
|
|
| 6. |
- Petersson, Jesper, 1974
(author)
-
Medicine At A Distance In Sweden: Spatiotemporal Matters In Accomplishing Working Telemedicine
- 2011
-
In: Science Studies. - 0786-3012. ; 24:2, s. 43-62
-
Journal article (peer-reviewed)abstract
- This paper examines the accomplishment of making technology work, using the discourse around telemedicine in Swedish healthcare during 1994-2003. The paper will compare four projects launched in the mid-1990s and policymakers’ visions of healthcare through telemedicine. I will employ a sociotechnical approach developed within Actor-Network Theory that understands functioning technology not as something intrinsic but as an outcome of an ongoing process of negotiations. In the paper, I will extend the sociotechnical approach of what constitutes working technology to include spatiotemporal matters. I will also approach the closely related issue of space that has become a concern of Actor-Network Theory scholars interested in the accomplishment and continued workings of technology as it travels. In this discussion, an emphasis on fixed relations (network space) has been challenged by investigations into changing relations (fluid space). This paper suggests that in order to travel well, technology must be both fixed and fluid.⁰
|
|
| 7. |
|
|
| 8. |
- Björkman, Kristoffer, et al.
(author)
-
Genotype-phenotype correlations in patients with complex I deficiency due to mutations in NDUFS1 and NDUFV1
- 2014
-
In: Euromit 2014, 15-19 juni, Tampere, Finland.
-
Conference paper (other academic/artistic)abstract
- Objectives: To study genotype-phenotype correlations in genes encoding complex I electron input module subunits. Materials and methods: We studied five patients with isolated complex I deficiency, three with NDUFS1 mutations and two with NDUFV1 mutations. A literature review of all reported cases of mutations in the affected genes was performed. Results: The literature review revealed pathological mutations in NDUFS1 for 18 patients in 17 families and correspondingly in NDUFV1 for 26 patients in 19 families. Unpublished clinical data for our five patients were added. Our study showed quite variable clinical courses; death before two years of age was seen in 41% of patients while 18% were alive at seven years. There was a significant difference between the NDUFS1 and NDUFV1 groups for clinical onset and life-span. Mutations in NDUFS1 were linked to a worse clinical course with earlier onset and earlier death. Conclusions: Genotype-phenotype correlations in patients with mutations affecting the genes that encode the electron input module of complex I vary, but patients with NDUFS1 mutation tend to have a worse clinical course than patients with NDUFV1 mutation. Identifying the mutations is of importance for accurate prognostic information and genetic counseling.
|
|
| 9. |
- Latini, Francesco, M.D. 1982-
(author)
-
Significance of white matter anatomy in interpreting features and behaviour of low-grade gliomas and implications for surgical treatment
- 2021
-
Doctoral thesis (other academic/artistic)abstract
- Diffuse gliomas are extremely heterogeneous tumours characterized by slow growth but extensive infiltration. Their kinetic features reflect the complex interaction over time with the surrounding brain, influencing treatment planning and outcome. Indeed, resection of diffuse gliomas present a surgical challenge due to their invasiveness and the preferential location in eloquent regions. White matter bundles are the main eloquent limit to surgical resection, but this anatomical-functional information cannot be predicted preoperatively on the individual level. The incomplete description of the human brain connectome, the complex application of pathological/lesion model to the brain connectomic organization, and the underestimated role of white matter anatomy in radiological classification systems are among the major limitations for the comprehension of the glioma/white matter interaction. The overall aim of this thesis was to explore a new approach and new techniques to study the glioma/white matter interaction. A combination of white matter dissection and diffusion tensor tractography (DTT) was used to describe the connectomic organization of two major temporo-occipital connections, the inferior and the middle longitudinal fasciculus. This information was applied to patients with diffuse gliomas, demonstrating how white matter analysis was important to decode patient specific cognitive and language impairment. A new classification system for diffuse gliomas, the Brain-Grid, was created, merging local radiological anatomy with a DTT atlas for infiltration analysis. This standardized radiological tool provided information on subcortical extension (tumour invasiveness), speed, and preferential direction of glioma progression. Applied to a larger cohort of patients, differences were detected between diffuse gliomas subtypes. Tumour invasiveness and the preferential location, type, and extent of white matter involvement differed, impacting overall survival. Regional differences in white matter infiltration were detected among five major white matter bundles, and possible favourable morphological and diffusion features were investigated with transmission electron microscopy and DTT. Fibre diameter, myelin thickness, and the organization of the white matter fibres were different in regions with high infiltration frequency, providing a possible link to the preferential location of diffuse gliomas. Finally, the white matter connectivity, tumour-induced neuroplasticity, clinical and demographic information, preoperative assessment (neuropsychological and language evaluation) were compared with intraoperative findings during awake surgery. Neuropsychological impairment was associated with more invasive tumours and a higher risk of the intraoperative finding of eloquent tumour. The pattern of early cortical neuroplasticity seemed exhausted at the time of diagnosis, with age as a factor predicting the neuroplasticity potential. The combined use of these new techniques revealed new insights into the glioma/white matter interaction. The results provided in this thesis, describe a new way to structure the multidisciplinary perioperative management of these patients. This new information may improve the functional outcome at the individual level, resulting in prolonged survival for adults with diffuse gliomas.
|
|
| 10. |
- Ahlander, Britt-Marie, 1954-
(author)
-
Magnetic Resonance Imaging of the Heart : Image quality, measurement accuracy and patient experience
- 2016
-
Doctoral thesis (other academic/artistic)abstract
- Background: Non-invasive diagnostic imaging of atherosclerotic coronary artery disease (CAD) is frequently carried out with cardiovascular magnetic resonance imaging (CMR) or myocardial perfusion single photon emission computed tomography (MPS). CMR is the gold standard for the evaluation of scar after myocardial infarction and MPS the clinical gold standard for ischemia. Magnetic Resonance Imaging (MRI) is at times difficult for patients and may induce anxiety while patient experience of MPS is largely unknown.Aims: To evaluate image quality in CMR with respect to the sequences employed, the influence of atrial fibrillation, myocardial perfusion and the impact of patient information. Further, to study patient experience in relation to MRI with the goal of improving the care of these patients.Method: Four study designs have been used. In paper I, experimental cross-over, paper (II) experimental controlled clinical trial, paper (III) psychometric crosssectional study and paper (IV) prospective intervention study. A total of 475 patients ≥ 18 years with primarily cardiac problems (I-IV) except for those referred for MRI of the spine (III) were included in the four studies.Result: In patients (n=20) with atrial fibrillation, a single shot steady state free precession (SS-SSFP) sequence showed significantly better image quality than the standard segmented inversion recovery fast gradient echo (IR-FGRE) sequence (I). In first-pass perfusion imaging the gradient echo-echo planar imaging sequence (GREEPI) (n=30) had lower signal-to-noise and contrast–to-noise ratios than the steady state free precession sequence (SSFP) (n=30) but displayed a higher correlation with the MPS results, evaluated both qualitatively and quantitatively (II). The MRIAnxiety Questionnaire (MRI-AQ) was validated on patients, referred for MRI of either the spine (n=193) or the heart (n=54). The final instrument had 15 items divided in two factors regarding Anxiety and Relaxation. The instrument was found to have satisfactory psychometric properties (III). Patients who prior CMR viewed an information video scored significantly (lower) better in the factor Relaxation, than those who received standard information. Patients who underwent MPS scored lower on both factors, Anxiety and Relaxation. The extra video information had no effect on CMR image quality (IV).Conclusion: Single shot imaging in atrial fibrillation produced images with less artefact than a segmented sequence. In first-pass perfusion imaging, the sequence GRE-EPI was superior to SSFP. A questionnaire depicting anxiety during MRI showed that video information prior to imaging helped patients relax but did not result in an improvement in image quality.
|
|
