| 1. |
- Ge, Chenjie, 1991, et al.
(författare)
-
Enlarged Training Dataset by Pairwise GANs for Molecular-Based Brain Tumor Classification
- 2020
-
Ingår i: IEEE Access. - 2169-3536 .- 2169-3536. ; 8:1, s. 22560-22570
-
Tidskriftsartikel (refereegranskat)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. |
- Ali, Muhaddisa Barat, 1986, et al.
(författare)
-
A novel federated deep learning scheme for glioma and its subtype classification
- 2023
-
Ingår i: Frontiers in Neuroscience. - 1662-4548 .- 1662-453X. ; 17
-
Tidskriftsartikel (refereegranskat)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.
|
|
| 3. |
- Johansson, Björn
(författare)
-
Opacification of anterior part of hydrophilic acrylic IOL or a prelenticular inflammatory membrane?
- 2012
-
Ingår i: Journal of cataract and refractive surgery. - Philadelphia : Elsevier. - 0886-3350 .- 1873-4502. ; 38:6, s. 1115-1116
-
Tidskriftsartikel (refereegranskat)abstract
- In their recent case report, Park and Chuck1 describe the bilateral appearance of an opacification at the plane of the anterior surface of the hydrophilic acrylic Akreos MI60 intraocular lens (IOL) (Bausch & Lomb). The patient's general history of diabetes mellitus, proliferative retinopathy, and iris rubeosis explains the limited pupil dilation preventing visualization of the capsulorhexis opening in their slitlamp images.
|
|
| 4. |
- Borrelli, Pablo, et al.
(författare)
-
Artificial intelligence-based detection of lymph node metastases by PET/CT predicts prostate cancer-specific survival
- 2021
-
Ingår i: Clinical Physiology and Functional Imaging. - : Wiley. - 1475-0961 .- 1475-097X. ; 41:1, s. 62-67
-
Tidskriftsartikel (refereegranskat)abstract
- Introduction Lymph node metastases are a key prognostic factor in prostate cancer (PCa), but detecting lymph node lesions from PET/CT images is a subjective process resulting in inter-reader variability. Artificial intelligence (AI)-based methods can provide an objective image analysis. We aimed at developing and validating an AI-based tool for detection of lymph node lesions. Methods A group of 399 patients with biopsy-proven PCa who had undergone(18)F-choline PET/CT for staging prior to treatment were used to train (n = 319) and test (n = 80) the AI-based tool. The tool consisted of convolutional neural networks using complete PET/CT scans as inputs. In the test set, the AI-based lymph node detections were compared to those of two independent readers. The association with PCa-specific survival was investigated. Results The AI-based tool detected more lymph node lesions than Reader B (98 vs. 87/117;p = .045) using Reader A as reference. AI-based tool and Reader A showed similar performance (90 vs. 87/111;p = .63) using Reader B as reference. The number of lymph node lesions detected by the AI-based tool, PSA, and curative treatment was significantly associated with PCa-specific survival. Conclusion This study shows the feasibility of using an AI-based tool for automated and objective interpretation of PET/CT images that can provide assessments of lymph node lesions comparable with that of experienced readers and prognostic information in PCa patients.
|
|
| 5. |
- Ahlander, Britt-Marie, 1954-, et al.
(författare)
-
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
-
Ingår i: Clinical Physiology and Functional Imaging. - : Blackwell Publishing. - 1475-0961 .- 1475-097X. ; 37:1, s. 52-61
-
Tidskriftsartikel (refereegranskat)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. |
- Ahlander, Britt-Marie, 1954-
(författare)
-
Magnetic Resonance Imaging of the Heart : Image quality, measurement accuracy and patient experience
- 2016
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)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.
|
|
| 7. |
- Bondesson, Johan, 1991
(författare)
-
Geometric Modeling of Thoracic Aortic Surface Morphology - Implications for Pathophysiology and Clinical Interventions
- 2021
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Vascular disease risk factors such as hypertension, hyperlipidemia and old age are all results of modern-day lifestyle, and these diseases are getting more and more common. One treatment option for vascular diseases such as aneurysms and dissections is endovascular aortic repair introduced in the early 1990s. This treatment uses tubular fabric covered metallic structures (endografts) that are implanted using a minimally invasive approach and placed to serve as an articial vessel in a damaged portion of the vasculature. To ensure that the interventions are successful, the endograft must be placed in the correct location, and designed to sustain the hostile biological, chemical, and mechanical conditions in the body for many years. This is an interaction that goes both ways, and keeping in mind that the endograft is a foreign object placed in the sensitive vascular system, it is also important that it does not disrupt the native conditions more than necessary. This thesis presents a segmentation and quantication methodology to accurately describe the complex morphology and motion of diseased blood vessels in vivo through a natural and intuitive description of their luminal surfaces. After methodology validation, a series of important clinical applications are performed, all based on non-invasive imaging. Firstly, it is shown that explicit surface curvature quantication is necessary when compared to relying solely on centerline curvature and estimation methods. Secondly, it is shown that endograft malapposition severity can be predicted from preoperative geometric analysis of thoracic aortic surfaces. Thirdly, a multiaxial dynamics analysis of cardiac induced thoracic aortic surface motion shows how thoracic endovascular aortic repair affects the deformations of the dierent portions of the thoracic aorta. Fourthly, the helical propagation pattern of type B aortic dissection is determined, and two distinct modes of chirality are revealed, i.e., achiral and right-handed chiral groups. Finally, the effects of thoracic endovascular aortic repair on helical and cross-sectional morphology of type B dissections are investigated revealing how acuity and chirality affects the alteration due to intraluminal lining with endografts. Thus, the work presented in this thesis contributes by adding knowledge about pathology and pathophysiology through better geometric description of surface conditions of diseased thoracic aortas. This gives clinicians insights to use in their treatment planning and provides more nuanced boundary conditions for endograft manufacturers. Comprehensive knowledge about diseases, better treatment planning, and better devices are all crucial in order to improve the outcomes of performed interventions and ultimately the quality of life for the treated patients.
|
|
| 8. |
- Forsgren, Mikael
(författare)
-
The Non-Invasive Liver Biopsy : Determining Hepatic Function in Diffuse and Focal LiverDisease
- 2017
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- The liver is one of the largest organs within the human body and it handles many vital tasks such as nutrient processing, toxin removal, and synthesis of important proteins. The number of people suffering from chronic liver disease is on the rise, likely due to the present ‘western’ lifestyle. As disease develops in the liver there are pathophysiological manifestations within the liver parenchyma that are both common and important to monitor. These manifestations include inflammation, fatty infiltration (steatosis), excessive scar tissue formation (fibrosis and cirrhosis), and iron loading. Importantly, as the disease progresses there is concurrent loss of liver function. Furthermore, postoperative liver function insufficiency is an important concern when planning surgical treatment of the liver, because it is associated with both morbidity and mortality. Liver function can also be hampered due to drug-induced injuries, an important aspect to consider in drug-development.Currently, an invasive liver needle biopsy is required to determine the aetiology and to stage or grade the pathophysiological manifestations. There are important limitations with the biopsy, which include, risk of serious complications, mortality, morbidity, inter- and intra-observer variability, sampling error, and sampling variability. Cleary, it would be beneficial to be able investigate the pathophysiological manifestations accurately, non-invasively, and on regional level.Current available laboratory liver function blood panels are typically insufficient and often only indicate damage at a late stage. Thus, it would be beneficial to have access to biomarkers that are both sensitive and responds to early changes in liver function in both clinical settings and for the pharmaceutical industry and regulatory agencies.The main aim of this thesis was to develop and evaluate methods that can be used for a ‘non-invasive liver biopsy’ using magnetic resonance (MR). We also aimed to develop sensitive methods for measure liver function based on gadoxetate-enhanced MR imaging (MRI).The presented work is primarily based on a prospective study on c. 100 patients suffering from chronic liver disease of varying aetiologies recruited due to elevated liver enzyme levels, without clear signs of decompensated cirrhosis. Our results show that the commonly used liver fat cut-off for diagnosing steatosis should be lowered from 5% to 3% when using MR proton-density fat fraction (PDFF). We also show that MR elastography (MRE) is superior in staging fibrosis.Finally we presented a framework for quantifying liver function based on gadoxetate-enhanced MRI. The method is based on clinical images and a clinical approved contrast agent (gadoxetate). The framework consists of; state-of the-art image reconstruction and correction methods, a mathematical model, and a precise model parametrization method. The model was developed and validated on healthy subjects. Thereafter the model was found applicable on the chronic liver disease cohort as well as validated using gadoxetate levels in biopsy samples and blood samples. The liver function parameters correlated with clinical markers for liver function and liver fibrosis (used as a surrogate marker for liver function).In summary, it should be possible to perform a non-invasive liver biopsy using: MRI-PDFF for liver fat and iron loading, MRE for liver fibrosis and possibly also inflammation, and measure liver function using the presented framework for analysing gadoxetate-enhanced MRI. With the exception of an MREtransducer no additional hardware is required on the MR scanner. The liver function method is likely to be useful both in a clinical setting and in pharmaceutical trials.
|
|
| 9. |
- Hortensius, Lisa M., et al.
(författare)
-
Serum docosahexaenoic acid levels are associated with brain volumes in extremely preterm born infants
- 2021
-
Ingår i: Pediatric Research. - : Springer Science and Business Media LLC. - 0031-3998 .- 1530-0447. ; 90:6
-
Tidskriftsartikel (refereegranskat)abstract
- Background: Docosahexaenoic acid (DHA) and arachidonic acid (AA) are important for fetal brain growth and development. Our aim was to evaluate the association between serum DHA and AA levels and brain volumes in extremely preterm infants. Methods: Infants born at <28 weeks gestational age in 2013–2015, a cohort derived from a randomized controlled trial comparing two types of parenteral lipid emulsions, were included (n = 90). Serum DHA and AA levels were measured at postnatal days 1, 7, 14, and 28, and the area under the curve was calculated. Magnetic resonance (MR) imaging was performed at term-equivalent age (n = 66), and volumes of six brain regions were automatically generated. Results: After MR image quality assessment and area under the curve calculation, 48 infants were included (gestational age mean [SD] 25.5 [1.4] weeks). DHA levels were positively associated with total brain (B = 7.966, p = 0.012), cortical gray matter (B = 3.653, p = 0.036), deep gray matter (B = 0.439, p = 0.014), cerebellar (B = 0.932, p = 0.003), and white matter volume (B = 3.373, p = 0.022). AA levels showed no association with brain volumes. Conclusions: Serum DHA levels during the first 28 postnatal days were positively associated with volumes of several brain structures in extremely preterm infants at term-equivalent age. Impact: Higher serum levels of DHA in the first 28 postnatal days are positively associated with brain volumes at term-equivalent age in extremely preterm born infants.Especially the most immature infants suffer from low DHA levels in the first 28 postnatal days, with little increase over time.Future research is needed to explore whether postnatal fatty acid supplementation can improve brain development and may serve as a nutritional preventive and therapeutic treatment option in extremely preterm infants.
|
|
| 10. |
- Jacobson, Lena, et al.
(författare)
-
Retinal ganglion cell topography predicts visual field function in spastic cerebral palsy
- 2020
-
Ingår i: Developmental Medicine and Child Neurology. - : Wiley. - 0012-1622 .- 1469-8749. ; 62:9, s. 1100-1106
-
Tidskriftsartikel (refereegranskat)abstract
- The aim of this study was to evaluate the use of optical coherence tomography (OCT) to identify and assess visual field defects caused by primary damage to the optic radiation in individuals with spastic cerebral palsy (CP). Ten individuals with spastic CP (six females, four males, with a median age of 21 years [range 17–38y]) had their brain lesions documented with conventional magnetic resonance imaging (MRI) and diffusion-weighted MRI fibre tractography. Their macular ganglion cell layer (GCL) and inner plexiform layer (IPL) were examined with OCT and their visual fields were plotted. All participants had good visual acuity and were able to cooperate with the MRI and OCT examinations, as well as undergoing reliable perimetry. We found focal thinning of the GCL+IPL and corresponding homonymous visual field defects in individuals with brain damage affecting the optic radiation. We used GCL+IPL sector asymmetry as a sensitive OCT parameter to identify focal visual field defects. We observed no such sector asymmetry in GCL+IPL, or focal visual field defects, in individuals with normal MRI optic radiation imaging. Lesions affecting the optic radiation cause retrograde trans-synaptic degeneration of retinal ganglion cells. OCT examination of the GCL in the macula identified corresponding focal damage to the optic radiation in individuals with spastic CP and can be used to predict focal visual field defects. What this paper adds: Spastic cerebral palsy (CP) may be associated with damage to the optic radiation. Damage to the optic radiation causes retrograde trans-synaptic degeneration (RTSD). RTSD can be mapped using optical coherence tomography. Ganglion cell topography can predict visual field defects in individuals with spastic CP.
|
|
