Deep image prior
Dotaz
Zobrazit nápovědu
Annotation of multiple regions of interest across the whole mouse brain is an indispensable process for quantitative evaluation of a multitude of study endpoints in neuroscience digital pathology. Prior experience and domain expert knowledge are the key aspects for image annotation quality and consistency. At present, image annotation is often achieved manually by certified pathologists or trained technicians, limiting the total throughput of studies performed at neuroscience digital pathology labs. It may also mean that simpler and quicker methods of examining tissue samples are used by non-pathologists, especially in the early stages of research and preclinical studies. To address these limitations and to meet the growing demand for image analysis in a pharmaceutical setting, we developed AnNoBrainer, an open-source software tool that leverages deep learning, image registration, and standard cortical brain templates to automatically annotate individual brain regions on 2D pathology slides. Application of AnNoBrainer to a published set of pathology slides from transgenic mice models of synucleinopathy revealed comparable accuracy, increased reproducibility, and a significant reduction (~ 50%) in time spent on brain annotation, quality control and labelling compared to trained scientists in pathology. Taken together, AnNoBrainer offers a rapid, accurate, and reproducible automated annotation of mouse brain images that largely meets the experts' histopathological assessment standards (> 85% of cases) and enables high-throughput image analysis workflows in digital pathology labs.
Deep learning (DL) methods have in recent years yielded impressive results in medical imaging, with the potential to function as clinical aid to radiologists. However, DL models in medical imaging are often trained on public research cohorts with images acquired with a single scanner or with strict protocol harmonization, which is not representative of a clinical setting. The aim of this study was to investigate how well a DL model performs in unseen clinical datasets-collected with different scanners, protocols and disease populations-and whether more heterogeneous training data improves generalization. In total, 3117 MRI scans of brains from multiple dementia research cohorts and memory clinics, that had been visually rated by a neuroradiologist according to Scheltens' scale of medial temporal atrophy (MTA), were included in this study. By training multiple versions of a convolutional neural network on different subsets of this data to predict MTA ratings, we assessed the impact of including images from a wider distribution during training had on performance in external memory clinic data. Our results showed that our model generalized well to datasets acquired with similar protocols as the training data, but substantially worse in clinical cohorts with visibly different tissue contrasts in the images. This implies that future DL studies investigating performance in out-of-distribution (OOD) MRI data need to assess multiple external cohorts for reliable results. Further, by including data from a wider range of scanners and protocols the performance improved in OOD data, which suggests that more heterogeneous training data makes the model generalize better. To conclude, this is the most comprehensive study to date investigating the domain shift in deep learning on MRI data, and we advocate rigorous evaluation of DL models on clinical data prior to being certified for deployment.
To stimulate progress in automating the reconstruction of neural circuits, we organized the first international challenge on 2D segmentation of electron microscopic (EM) images of the brain. Participants submitted boundary maps predicted for a test set of images, and were scored based on their agreement with a consensus of human expert annotations. The winning team had no prior experience with EM images, and employed a convolutional network. This "deep learning" approach has since become accepted as a standard for segmentation of EM images. The challenge has continued to accept submissions, and the best so far has resulted from cooperation between two teams. The challenge has probably saturated, as algorithms cannot progress beyond limits set by ambiguities inherent in 2D scoring and the size of the test dataset. Retrospective evaluation of the challenge scoring system reveals that it was not sufficiently robust to variations in the widths of neurite borders. We propose a solution to this problem, which should be useful for a future 3D segmentation challenge.
- Publikační typ
- časopisecké články MeSH
OBJECTIVES: There is increasing evidence that serum lipoprotein cholesterol biomarkers are associated with disease progression in clinically isolated syndromes (CIS). Apolipoproteins (Apo) are recognition ligands that mediate the physiological interactions of cholesterol-containing lipoproteins. The objective of this study was to investigate whether serum Apo levels are associated with CIS disease progression. METHODS: ApoB, ApoAI, ApoAII, ApoE and lipoprotein (a) (Lpa) levels were measured in serum samples obtained prior to the start of treatment from 181 CIS patients (123 women, 58 men, 68% women; mean age: 28.1±SD 8.1 years). All patients were treated with intramuscular interferon-β as part of the prospective study. Clinical and MRI assessments were obtained at baseline, 6, 12 and 24 months after start of interferon-β treatment. RESULTS: Greater ApoB levels were associated with increased number of new T2 lesions (p<0.001) and increased number of new or enlarging T2 lesions (p<0.001) over 2 years. Each 10 mg/dL of greater baseline ApoB is associated with a 16% increase in the number of new T2 lesions over 2 years. ApoAI, ApoAII, ApoE and Lpa were not associated with T2 lesions. Greater ApoE levels were associated with greater deep grey matter atrophy (partial correlation rp=-0.28, p<0.001). Each 1 mg/dL increment in ApoE levels was associated with a 1% increase in deep grey matter atrophy over 2 years. CONCLUSIONS: Serum ApoB levels are associated with new lesion accumulation whereas ApoE levels are associated with deep grey matter atrophy in high risk CIS patients treated with interferon β-1a.
- MeSH
- antiflogistika terapeutické užití MeSH
- apolipoproteiny E krev MeSH
- apolipoproteiny metabolismus MeSH
- atrofie MeSH
- dospělí MeSH
- interpretace statistických dat MeSH
- lidé středního věku MeSH
- lidé MeSH
- lipidy krev MeSH
- longitudinální studie MeSH
- magnetická rezonanční tomografie MeSH
- methylprednisolon terapeutické užití MeSH
- mladiství MeSH
- mladý dospělý MeSH
- mozek patologie MeSH
- počítačové zpracování obrazu MeSH
- progrese nemoci MeSH
- prospektivní studie MeSH
- roztroušená skleróza metabolismus patologie MeSH
- Check Tag
- dospělí MeSH
- lidé středního věku MeSH
- lidé MeSH
- mladiství MeSH
- mladý dospělý MeSH
- mužské pohlaví MeSH
- ženské pohlaví MeSH
- Publikační typ
- časopisecké články MeSH
- multicentrická studie MeSH
- pozorovací studie MeSH
- Research Support, U.S. Gov't, Non-P.H.S. MeSH
During implantation of deep-brain stimulation (DBS) electrodes in the target structure, neurosurgeons and neurologists commonly observe a "microlesion effect" (MLE), which occurs well before initiating subthalamic DBS. This phenomenon typically leads to a transitory improvement of motor symptoms of patients suffering from Parkinson's disease (PD). Mechanisms behind MLE remain poorly understood. In this work, we exploited the notion of ranking to assess spontaneous brain activity in PD patients examined by resting-state functional magnetic resonance imaging in response to penetration of DBS electrodes in the subthalamic nucleus. In particular, we employed a hypothesis-free method, eigenvector centrality (EC), to reveal motor-communication-hubs of the highest rank and their reorganization following the surgery; providing a unique opportunity to evaluate the direct impact of disrupting the PD motor circuitry in vivo without prior assumptions. Penetration of electrodes was associated with increased EC of functional connectivity in the brainstem. Changes in connectivity were quantitatively related to motor improvement, which further emphasizes the clinical importance of the functional integrity of the brainstem. Surprisingly, MLE and DBS were associated with anatomically different EC maps despite their similar clinical benefit on motor functions. The DBS solely caused an increase in connectivity of the left premotor region suggesting separate pathophysiological mechanisms of both interventions. While the DBS acts at the cortical level suggesting compensatory activation of less affected motor regions, the MLE affects more fundamental circuitry as the dysfunctional brainstem predominates in the beginning of PD. These findings invigorate the overlooked brainstem perspective in the understanding of PD and support the current trend towards its early diagnosis.
- MeSH
- dospělí MeSH
- echoplanární zobrazování MeSH
- hluboká mozková stimulace škodlivé účinky MeSH
- kyslík krev MeSH
- lidé středního věku MeSH
- lidé MeSH
- magnetická rezonanční tomografie * MeSH
- mapování mozku MeSH
- mozkový kmen krevní zásobení fyziologie MeSH
- nucleus subthalamicus krevní zásobení patologie MeSH
- odpočinek * MeSH
- Parkinsonova nemoc * patologie patofyziologie terapie MeSH
- počítačové zpracování obrazu MeSH
- stupeň závažnosti nemoci MeSH
- Check Tag
- dospělí MeSH
- lidé středního věku MeSH
- lidé MeSH
- mužské pohlaví MeSH
- ženské pohlaví MeSH
- Publikační typ
- časopisecké články MeSH
- práce podpořená grantem MeSH
Although the cerebellum contributes to higher-order cognitive and emotional functions relevant to posttraumatic stress disorder (PTSD), prior research on cerebellar volume in PTSD is scant, particularly when considering subregions that differentially map on to motor, cognitive, and affective functions. In a sample of 4215 adults (PTSD n = 1642; Control n = 2573) across 40 sites from the ENIGMA-PGC PTSD working group, we employed a new state-of-the-art deep-learning based approach for automatic cerebellar parcellation to obtain volumetric estimates for the total cerebellum and 28 subregions. Linear mixed effects models controlling for age, gender, intracranial volume, and site were used to compare cerebellum volumes in PTSD compared to healthy controls (88% trauma-exposed). PTSD was associated with significant grey and white matter reductions of the cerebellum. Compared to controls, people with PTSD demonstrated smaller total cerebellum volume, as well as reduced volume in subregions primarily within the posterior lobe (lobule VIIB, crus II), vermis (VI, VIII), flocculonodular lobe (lobule X), and corpus medullare (all p-FDR < 0.05). Effects of PTSD on volume were consistent, and generally more robust, when examining symptom severity rather than diagnostic status. These findings implicate regionally specific cerebellar volumetric differences in the pathophysiology of PTSD. The cerebellum appears to play an important role in higher-order cognitive and emotional processes, far beyond its historical association with vestibulomotor function. Further examination of the cerebellum in trauma-related psychopathology will help to clarify how cerebellar structure and function may disrupt cognitive and affective processes at the center of translational models for PTSD.
- MeSH
- bílá hmota patologie diagnostické zobrazování MeSH
- deep learning MeSH
- dospělí MeSH
- lidé středního věku MeSH
- lidé MeSH
- magnetická rezonanční tomografie * metody MeSH
- mozeček * patologie diagnostické zobrazování MeSH
- posttraumatická stresová porucha * patologie patofyziologie diagnostické zobrazování MeSH
- šedá hmota patologie MeSH
- velikost orgánu MeSH
- Check Tag
- dospělí MeSH
- lidé středního věku MeSH
- lidé MeSH
- mužské pohlaví MeSH
- ženské pohlaví MeSH
- Publikační typ
- časopisecké články MeSH
- metaanalýza MeSH
Diabetic retinopathy is a diabetes complication that affects the eyes, caused by damage to the blood vessels of the light-sensitive tissue of the retina. At the onset, diabetic retinopathy may cause no symptoms or only mild vision problems, but eventually it can cause blindness. Totally automated segmentation of Eye Fundus Images (EFI) is a necessary step for accurate and early quantification of lesions, useful in the future for better automated diagnosis of degree of diabetic retinopathy and damage caused by the disease. Deep Learning segmentation networks are the state-of-the-art, but quality, limitations and comparison of architectures of segmentation networks is necessary. We build off-theshelf deep learning architectures and evaluate them on a publicly available dataset, to conclude the strengths and limitations of the approaches and to compare architectures. Results show that the segmentation networks score high on important metrics, such as 87.5% weighted IoU on FCN. We also show that network architecture is very important, with DeepLabV3 and FCN outperforming other networks tested by more than 30 pp. We also show that DeepLabV3 outperforms prior related work using deep learning to detect lesions. Finally, we identify and investigate the problem of very low IoU and precision scores, such as 17% IoU of microaneurisms in DeepLabV3, concluding it is due to a large number of false positives. This leads us to discuss the challenges that lie ahead to improve the limitations that we identified
Electrode implantation into the subthalamic nucleus for deep brain stimulation in Parkinson's disease (PD) is associated with a temporary motor improvement occurring prior to neurostimulation. We studied this phenomenon by functional magnetic resonance imaging (fMRI) when considering the Unified Parkinson's Disease Rating Scale (UPDRS-III) and collateral oedema. Twelve patients with PD (age 55.9± (SD)6.8 years, PD duration 9-15 years) underwent bilateral electrode implantation into the subthalamic nucleus. The fMRI was carried out after an overnight withdrawal of levodopa (OFF condition): (i) before and (ii) within three days after surgery in absence of neurostimulation. The motor task involved visually triggered finger tapping. The OFF/UPDRS-III score dropped from 33.8±8.7 before to 23.3±4.8 after the surgery (p<0.001), correlating with the postoperative oedema score (p<0.05). During the motor task, bilateral activation of the thalamus and basal ganglia, motor cortex and insula were preoperatively higher than after surgery (p<0.001). The results became more enhanced after compensation for the oedema and UPDRS-III scores. In addition, the rigidity and axial symptoms score correlated inversely with activation of the putamen and globus pallidus (p<0.0001). One month later, the OFF/UPDRS-III score had returned to the preoperative level (35.8±7.0, p = 0.4).In conclusion, motor improvement induced by insertion of an inactive electrode into the subthalamic nucleus caused an acute microlesion which was at least partially related to the collateral oedema and associated with extensive impact on the motor network. This was postoperatively manifested as lowered movement-related activation at the cortical and subcortical levels and differed from the known effects of neurostimulation or levodopa. The motor system finally adapted to the microlesion within one month as suggested by loss of motor improvement and good efficacy of deep brain stimulation.
- MeSH
- edém patologie MeSH
- elektrostimulační terapie metody MeSH
- globus pallidus účinky léků patologie chirurgie MeSH
- hluboká mozková stimulace metody MeSH
- implantované elektrody MeSH
- levodopa terapeutické užití MeSH
- lidé středního věku MeSH
- lidé MeSH
- magnetická rezonanční tomografie metody MeSH
- motorické korové centrum účinky léků patologie chirurgie MeSH
- nucleus subthalamicus účinky léků patologie chirurgie MeSH
- Parkinsonova nemoc farmakoterapie patologie chirurgie MeSH
- thalamus účinky léků patologie chirurgie MeSH
- Check Tag
- lidé středního věku MeSH
- lidé MeSH
- mužské pohlaví MeSH
- ženské pohlaví MeSH
- Publikační typ
- časopisecké články MeSH
- práce podpořená grantem MeSH
- randomizované kontrolované studie MeSH
BACKGROUND: Risk of subsequent deep vein thrombosis (DVT) following superficial vein thrombosis (SVT) is not fully appreciated. Mechanisms, time relations and risk factors for DVT arising upon earlier SVT remain unclear. The aim of this study was to analyze time relations between local symptoms of lower limb superficial vein thrombosis, duplex findings and onset of deep vein thrombosis during clinically evident course of SVT. PATIENTS AND METHODS: 46 patients with early (onset less than 72 hours prior to inclusion) clinical diagnosis of SVT, confirmed ultrasonographically were included in this prospective, multicenter study. Progress of pain, erythema and swelling in relation to subsequent ultrasound changes in size and localization of thrombus at 0, 7, 14 and 21 day of study has been recorded. RESULTS: Local symptoms subsided completely during 3 weeks. At that time thrombus disappeared completely only in 26% of cases, in remaining cases decreased in size from average 117.5 mm to 43.0 mm. Thrombus regression was similar to venous blood outflow direction--proximal to femoral area. Thrombus propagation was observed following regression of local symptoms of SVT. 4 cases of DVT (8.7%) were diagnosed at 2-11 days. CONCLUSIONS: Local, clinically detectable symptoms of SVT regress incomparably quicker than thrombus in affected veins. Risk of further thrombus propagation extends well beyond the period of intensive local symptoms of SVT. Regression of thrombus in femoral area requires significantly more time than in popliteal or calf segment. Thrombus propagation is directed with blood flow towards femoral segment.
- MeSH
- dospělí MeSH
- hodnocení rizik metody MeSH
- lidé středního věku MeSH
- lidé MeSH
- progrese nemoci MeSH
- rizikové faktory MeSH
- rozložení podle pohlaví MeSH
- senioři MeSH
- stupeň závažnosti nemoci * MeSH
- ultrasonografie MeSH
- žilní trombóza klasifikace diagnostické zobrazování epidemiologie MeSH
- Check Tag
- dospělí MeSH
- lidé středního věku MeSH
- lidé MeSH
- mužské pohlaví MeSH
- senioři MeSH
- ženské pohlaví MeSH
- Publikační typ
- časopisecké články MeSH
- klinické zkoušky MeSH
- multicentrická studie MeSH
- práce podpořená grantem MeSH
- randomizované kontrolované studie MeSH
- Geografické názvy
- Česká republika MeSH
- Polsko epidemiologie MeSH
Aim: Synovial sarcoma is a clinically and morphologically distinct neoplasm of uncertain histogenesis predominantly affecting the extremities of adolescents and young adults. It is the fourth commonest soft tissue sarcoma in adults. It is regarded as an aggressive neoplasm with a reported 5-year survival rate ranging from 40-76%. The aim of our presentation is to verify the unusual presentation of long-standing pain at the site of a deep small tumour, without the development of swelling. Patients - methods and results: An 18-year-old young man suffered with localised pain in the medial lower part of the right thigh for 5 years. The pain appeared after prolonged exercise (cycling), and worsened with palpation. He had no pain at rest. There was no palpable mass in the thigh. A small spherical lesion was diagnosed as an obliterated venous varix one year prior to admission for an MRI. A small deep localised tumour was revealed by an ultrasonography and MRI shortly before surgery. Surgical excision was performed using preoperative ultrasonography. A solid 1 cm diameter demarcated tumour was removed from the deep inter-muscular septum. Surprisingly, the histological diagnosis of a monophasic synovial sarcoma was established. Thus the reoperation of the surgical margin excision was indicated. No malignant cells were found in the excised soft tissues. The patient is alive and well 1 year after the surgery. The controlled MRI of the thigh did not reveal any signs of a viable tumour. Discussion: The occurrence of long-standing pain at a tumour site prior to the development of swelling is uncommon. The pathophysiology of this pain is unclear. Awareness of this unusual presentation and appropriate investigation may enable the detection of synovial sarcoma at an early prognostically favourable stage.
- MeSH
- bolest diagnóza etiologie MeSH
- lidé MeSH
- magnetická rezonanční tomografie MeSH
- mladý dospělý MeSH
- stehno patologie MeSH
- synoviom chirurgie patologie MeSH
- Check Tag
- lidé MeSH
- mladý dospělý MeSH
- mužské pohlaví MeSH
- Publikační typ
- kazuistiky MeSH
