Computer-aided detection of radiation-induced cerebral microbleeds on susceptibility-weighted MR images

Deep-Learning-Based MRI Microbleeds Detection for Cerebral Small Vessel Disease on Quantitative Susceptibility Mapping

BACKGROUND

Cerebral microbleeds (CMB) are indicators of severe cerebral small vessel disease (CSVD) that can be identified through hemosiderin-sensitive sequences in MRI. Specifically, quantitative susceptibility mapping (QSM) and deep learning were applied to detect CMBs in MRI.

PURPOSE

To automatically detect CMB on QSM, we proposed a two-stage deep learning pipeline.

STUDY TYPE

Retrospective.

SUBJECTS

A total number of 1843 CMBs from 393 patients (69 ± 12) with cerebral small vessel disease were included in this study. Seventy-eight subjects (70 ± 13) were used as external testing.

FIELD STRENGTH/SEQUENCE

3 T/QSM.

ASSESSMENT

The proposed pipeline consisted of two stages. In stage I, 2.5D fast radial symmetry transform (FRST) algorithm along with a one-layer convolutional network was used to identify CMB candidate regions in QSM images. In stage II, the V-Net was utilized to reduce false positives. The V-Net was trained using CMB and non CMB labels, which allowed for high-level feature extraction and differentiation between CMBs and CMB mimics like vessels. The location of CMB was assessed according to the microbleeds anatomical rating scale (MARS) system.

STATISTICAL TESTS

The sensitivity and positive predicative value (PPV) were reported to evaluate the performance of the model. The number of false positive per subject was presented.

RESULTS

Our pipeline demonstrated high sensitivities of up to 94.9% at stage I and 93.5% at stage II. The overall sensitivity was 88.9%, and the false positive rate per subject was 2.87. With respect to MARS, sensitivities of above 85% were observed for nine different brain regions.

DATA CONCLUSION

We have presented a deep learning pipeline for detecting CMB in the CSVD cohort, along with a semi-automated MARS scoring system using the proposed method. Our results demonstrated the successful application of deep learning for CMB detection on QSM and outperformed previous handcrafted methods.

LEVEL OF EVIDENCE

2 TECHNICAL EFFICACY: Stage 2.

Machine and deep learning algorithms for classifying different types of dementia: A literature review

The cognitive impairment known as dementia affects millions of individuals throughout the globe. The use of machine learning (ML) and deep learning (DL) algorithms has shown great promise as a means of early identification and treatment of dementia. Dementias such as Alzheimer's Dementia, frontotemporal dementia, Lewy body dementia, and vascular dementia are all discussed in this article, along with a literature review on using ML algorithms in their diagnosis. Different ML algorithms, such as support vector machines, artificial neural networks, decision trees, and random forests, are compared and contrasted, along with their benefits and drawbacks. As discussed in this article, accurate ML models may be achieved by carefully considering feature selection and data preparation. We also discuss how ML algorithms can predict disease progression and patient responses to therapy. However, overreliance on ML and DL technologies should be avoided without further proof. It's important to note that these technologies are meant to assist in diagnosis but should not be used as the sole criteria for a final diagnosis. The research implies that ML algorithms may help increase the precision with which dementia is diagnosed, especially in its early stages. The efficacy of ML and DL algorithms in clinical contexts must be verified, and ethical issues around the use of personal data must be addressed, but this requires more study.

Preliminary evidence of high prevalence of cerebral microbleeds in astronauts with spaceflight experience

Long-duration spaceflight poses a variety of health risks to astronauts, largely resulting from extended exposure to microgravity and radiation. Here, we assessed the prevalence and incidence of cerebral microbleeds in sixteen astronauts before and after a typical 6-month mission on board the International Space Station Cerebral microbleeds are microhemorrhages in the brain, which are typically interpreted as early evidence of small vessel disease and have been associated with cognitive impairment. We identified evidence of higher-than-expected microbleed prevalence in astronauts with prior spaceflight experience. However, we did not identify a statistically significant increase in microbleed burden up to 7 months after spaceflight. Altogether, these preliminary findings suggest that spaceflight exposure may increase microbleed burden, but this influence may be indirect or occur over time courses that exceed 1 year. For health monitoring purposes, it may be valuable to acquire neuroimaging data that are able to detect the occurrence of microbleeds in astronauts following their spaceflight missions.

Automated Detection of Cerebral Microbleeds on Two-dimensional Gradient-recalled Echo T2* Weighted Images Using a Morphology Filter Bank and Convolutional Neural Network

PURPOSE

We present a novel algorithm for the automated detection of cerebral microbleeds (CMBs) on 2D gradient-recalled echo T2* weighted images (T2*WIs). This approach combines a morphology filter bank with a convolutional neural network (CNN) to improve the efficiency of CMB detection. A technical evaluation was performed to ascertain the algorithm's accuracy.

METHODS

In this retrospective study, 60 patients with CMBs on T2*WIs were included. The gold standard was set by three neuroradiologists based on the Microbleed Anatomic Rating Scale guidelines. Images with CMBs were extracted from the training dataset comprising 30 cases using a morphology filter bank, and false positives (FPs) were removed based on the threshold of size and signal intensity. The extracted images were used to train the CNN (Vgg16). To determine the effectiveness of the morphology filter bank, the outcomes of the following two methods for detecting CMBs from the 30-case test dataset were compared: (a) employing the morphology filter bank and additional FP removal and (b) comprehensive detection without filters. The trained CNN processed both sets of initial CMB candidates, and the final CMB candidates were compared with the gold standard. The sensitivity and FPs per patient of both methods were compared.

RESULTS

After CNN processing, the morphology-filter-bank-based method had a 95.0% sensitivity with 4.37 FPs per patient. In contrast, the comprehensive method had a 97.5% sensitivity with 25.87 FPs per patient.

CONCLUSION

Through effective CMB candidate refinement with a morphology filter bank and FP removal with a CNN, we achieved a high CMB detection rate and low FP count. Combining a CNN and morphology filter bank may facilitate the accurate automated detection of CMBs on T2*WIs.

Naïve Bayes classifier assisted automated detection of cerebral microbleeds in susceptibility-weighted imaging brain images

Cerebral microbleeds (CMBs) in the brain are the essential indicators of critical brain disorders such as dementia and ischemic stroke. Generally, CMBs are detected manually by experts, which is an exhaustive task with limited productivity. Since CMBs have complex morphological nature, manual detection is prone to errors. This paper presents a machine learning-based automated CMB detection technique in the brain susceptibility-weighted imaging (SWI) scans based on statistical feature extraction and classification. The proposed method consists of three steps: (1) removal of the skull and extraction of the brain; (2) thresholding for the extraction of initial candidates; and (3) extracting features and applying classification models such as random forest and naïve Bayes classifiers for the detection of true positive CMBs. The proposed technique is validated on a dataset consisting of 20 subjects. The dataset is divided into training data that consist of 14 subjects with 104 microbleeds and testing data that consist of 6 subjects with 63 microbleeds. We were able to achieve 85.7% sensitivity using the random forest classifier with 4.2 false positives per CMB, and the naïve Bayes classifier achieved 90.5% sensitivity with 5.5 false positives per CMB. The proposed technique outperformed many state-of-the-art methods proposed in previous studies.

Deep learning-assisted IoMT framework for cerebral microbleed detection

The Internet of Things (IoT), big data, and artificial intelligence (AI) are all key technologies that influence the formation and implementation of digital medical services. Building Internet of Medical Things (IoMT) systems that combine advanced sensors with AI-powered insights is critical for intelligent medical systems. This paper presents an IoMT framework for brain magnetic resonance imaging (MRI) analysis to lessen the unavoidable diagnosis and therapy faults that occur in human clinical settings for the accurate detection of cerebral microbleeds (CMBs). The problems in accurate CMB detection include that CMBs are tiny dots 5-10 mm in diameter; they are similar to healthy tissues and are exceedingly difficult to identify, necessitating specialist guidance in remote and underdeveloped medical centers. Secondly, in the existing studies, computer-aided diagnostic (CAD) systems are designed for accurate CMB detection, however, their proposed approaches consist of two stages. Potential candidate CMBs from the complete MRI image are selected in the first stage and then passed to the phase of false-positive reduction. These pre-and post-processing steps make it difficult to build a completely automated CAD system for CMB that can produce results without human intervention. Hence, as a key goal of this work, an end-to-end enhanced UNet-based model for effective CMB detection and segmentation for IoMT devices is proposed. The proposed system requires no pre-processing or post-processing steps for CMB segmentation, and no existing research localizes each CMB pixel from the complete MRI image input. The findings indicate that the suggested method outperforms in detecting CMBs in the presence of contrast variations and similarities with other normal tissues and yields a good dice score of 0.70, an accuracy of 99 %, as well as a false-positive rate of 0.002 %. © 2017 Elsevier Inc. All rights reserved.

DeepSWI: Using Deep Learning to Enhance Susceptibility Contrast on T2*-Weighted MRI

BACKGROUND

Although susceptibility-weighted imaging (SWI) is the gold standard for visualizing cerebral microbleeds (CMBs) in the brain, the required phase data are not always available clinically. Having a postprocessing tool for generating SWI contrast from T2*-weighted magnitude images is therefore advantageous.

PURPOSE

To create synthetic SWI images from clinical T2*-weighted magnitude images using deep learning and evaluate the resulting images in terms of similarity to conventional SWI images and ability to detect radiation-associated CMBs.

STUDY TYPE

Retrospective.

POPULATION

A total of 145 adults (87 males/58 females; 43.9 years old) with radiation-associated CMBs were used to train (16,093 patches/121 patients), validate (484 patches/4 patients), and test (2420 patches/20 patients) our networks.

FIELD STRENGTH/SEQUENCE

3D T2*-weighted, gradient-echo acquired at 3 T.

ASSESSMENT

Structural similarity index (SSIM), peak signal-to-noise-ratio (PSNR), normalized mean-squared-error (nMSE), CMB counts, and line profiles were compared among magnitude, original SWI, and synthetic SWI images. Three blinded raters (J.E.V.M., M.A.M., B.B. with 8-, 6-, and 4-years of experience, respectively) independently rated and classified test-set images.

STATISTICAL TESTS

Kruskall-Wallis and Wilcoxon signed-rank tests were used to compare SSIM, PSNR, nMSE, and CMB counts among magnitude, original SWI, and predicted synthetic SWI images. Intraclass correlation assessed interrater variability. P values <0.005 were considered statistically significant.

RESULTS

SSIM values of the predicted vs. original SWI (0.972, 0.995, 0.9864) were statistically significantly higher than that of the magnitude vs. original SWI (0.970, 0.994, 0.9861) for whole brain, vascular structures, and brain tissue regions, respectively; 67% (19/28) CMBs detected on original SWI images were also detected on the predicted SWI, whereas only 10 (36%) were detected on magnitude images. Overall image quality was similar between the synthetic and original SWI images, with less artifacts on the former.

CONCLUSIONS

This study demonstrated that deep learning can increase the susceptibility contrast present in neurovasculature and CMBs on T2*-weighted magnitude images, without residual susceptibility-induced artifacts. This may be useful for more accurately estimating CMB burden from magnitude images alone.

EVIDENCE LEVEL

3.

TECHNICAL EFFICACY

Stage 2.

Automated detection of cerebral microbleeds on MR images using knowledge distillation framework

Introduction

Cerebral microbleeds (CMBs) are associated with white matter damage, and various neurodegenerative and cerebrovascular diseases. CMBs occur as small, circular hypointense lesions on T2*-weighted gradient recalled echo (GRE) and susceptibility-weighted imaging (SWI) images, and hyperintense on quantitative susceptibility mapping (QSM) images due to their paramagnetic nature. Accurate automated detection of CMBs would help to determine quantitative imaging biomarkers (e.g., CMB count) on large datasets. In this work, we propose a fully automated, deep learning-based, 3-step algorithm, using structural and anatomical properties of CMBs from any single input image modality (e.g., GRE/SWI/QSM) for their accurate detections.

Methods

In our method, the first step consists of an initial candidate detection step that detects CMBs with high sensitivity. In the second step, candidate discrimination step is performed using a knowledge distillation framework, with a multi-tasking teacher network that guides the student network to classify CMB and non-CMB instances in an offline manner. Finally, a morphological clean-up step further reduces false positives using anatomical constraints. We used four datasets consisting of different modalities specified above, acquired using various protocols and with a variety of pathological and demographic characteristics.

Results

On cross-validation within datasets, our method achieved a cluster-wise true positive rate (TPR) of over 90% with an average of <2 false positives per subject. The knowledge distillation framework improves the cluster-wise TPR of the student model by 15%. Our method is flexible in terms of the input modality and provides comparable cluster-wise TPR and better cluster-wise precision compared to existing state-of-the-art methods. When evaluating across different datasets, our method showed good generalizability with a cluster-wise TPR >80 % with different modalities. The python implementation of the proposed method is openly available.

Diagnosis of cerebral microbleed via VGG and extreme learning machine trained by Gaussian map bat algorithm

Cerebral microbleed (CMB) is a serious public health concern. It is associated with dementia, which can be detected with brain magnetic resonance image (MRI). CMBs often appear as tiny round dots on MRIs, and they can be spotted anywhere over brain. Therefore, manual inspection is tedious and lengthy, and the results are often short in reproducible. In this paper, a novel automatic CMB diagnosis method was proposed based on deep learning and optimization algorithms, which used the brain MRI as the input and output the diagnosis results as CMB and non-CMB. Firstly, sliding window processing was employed to generate the dataset from brain MRIs. Then, a pre-trained VGG was employed to obtain the image features from the dataset. Finally, an ELM was trained by Gaussian-map bat algorithm (GBA) for identification. Results showed that the proposed method VGG-ELM-GBA provided better generalization performance than several state-of-the-art approaches.

CMB-HUNT: Automatic detection of cerebral microbleeds using a deep neural network

Cerebral microbleeds (CMBs) are gaining increasing interest due to their importance in diagnosing cerebral small vessel diseases. However, manual inspection of CMBs is time-consuming and prone to human error. Existing automated or semi-automated solutions still have insufficient detection sensitivity and specificity. Furthermore, they frequently use more than one magnetic resonance imaging modality, but these are not always available. The majority of AI-based solutions use either numeric or image data, which may not provide sufficient information about the true nature of CMBs. This paper proposes a deep neural network with multi-type input data for automated CMB detection (CMB-HUNT) using only susceptibility-weighted imaging data (SWI). Combination of SWIs and radiomic-type numerical features allowed us to identify CMBs with high accuracy without the need for additional imaging modalities or complex predictive models. Two independent datasets were used: one with 304 patients (39 with CMBs) for training and internal system validation and one with 61 patients (21 with CMBs) for external validation. For the hold-out testing dataset, CMB-HUNT reached a sensitivity of 90.0%. As results of testing showed, CMB-HUNT outperforms existing methods in terms of the number of FPs per case, which is the lowest reported thus far (0.54 FPs/patient). The proposed system was successfully applied to the independent validation set, reaching a sensitivity of 91.5% with 1.9 false positives per patient and proving its generalization potential. The results were comparable to previous studies. Our research confirms the usefulness of deep learning solutions for CMB detection based only on one MRI modality.

Using transfer learning for automated microbleed segmentation

INTRODUCTION

Cerebral microbleeds are small perivascular hemorrhages that can occur in both gray and white matter brain regions. Microbleeds are a marker of cerebrovascular pathology and are associated with an increased risk of cognitive decline and dementia. Microbleeds can be identified and manually segmented by expert radiologists and neurologists, usually from susceptibility-contrast MRI. The latter is hard to harmonize across scanners, while manual segmentation is laborious, time-consuming, and subject to interrater and intrarater variability. Automated techniques so far have shown high accuracy at a neighborhood ("patch") level at the expense of a high number of false positive voxel-wise lesions. We aimed to develop an automated, more precise microbleed segmentation tool that can use standardizable MRI contrasts.

METHODS

We first trained a ResNet50 network on another MRI segmentation task (cerebrospinal fluid vs. background segmentation) using T1-weighted, T2-weighted, and T2* MRIs. We then used transfer learning to train the network for the detection of microbleeds with the same contrasts. As a final step, we employed a combination of morphological operators and rules at the local lesion level to remove false positives. Manual segmentation of microbleeds from 78 participants was used to train and validate the system. We assessed the impact of patch size, freezing weights of the initial layers, mini-batch size, learning rate, and data augmentation on the performance of the Microbleed ResNet50 network.

RESULTS

The proposed method achieved high performance, with a patch-level sensitivity, specificity, and accuracy of 99.57, 99.16, and 99.93%, respectively. At a per lesion level, sensitivity, precision, and Dice similarity index values were 89.1, 20.1, and 0.28% for cortical GM; 100, 100, and 1.0% for deep GM; and 91.1, 44.3, and 0.58% for WM, respectively.

DISCUSSION

The proposed microbleed segmentation method is more suitable for the automated detection of microbleeds with high sensitivity.

Computer-Aided Extraction of Select MRI Markers of Cerebral Small Vessel Disease: A Systematic Review

Cerebral small vessel disease (CSVD) is a major vascular contributor to cognitive impairment in ageing, including dementias. Imaging remains the most promising method for in vivo studies of CSVD. To replace the subjective and laborious visual rating approaches, emerging studies have applied state-of-the-art artificial intelligence to extract imaging biomarkers of CSVD from MRI scans. We aimed to summarise published computer-aided methods for the examination of three imaging biomarkers of CSVD, namely cerebral microbleeds (CMB), dilated perivascular spaces (PVS), and lacunes of presumed vascular origin. Seventy classical image processing, classical machine learning, and deep learning studies were identified. Transfer learning and weak supervision techniques have been applied to accommodate the limitations in the training data. While good performance metrics were achieved in local datasets, there have not been generalisable pipelines validated in different research and/or clinical cohorts. Future studies could consider pooling data from multiple sources to increase data size and diversity, and evaluating performance using both image processing metrics and associations with clinical measures.

Construction of a Medical Micro-Object Cascade Network for Automated Segmentation of Cerebral Microbleeds in Susceptibility Weighted Imaging

Aim: The detection and segmentation of cerebral microbleeds (CMBs) images are the focus of clinical diagnosis and treatment. However, segmentation is difficult in clinical practice, and missed diagnosis may occur. Few related studies on the automated segmentation of CMB images have been performed, and we provide the most effective CMB segmentation to date using an automated segmentation system.

Materials and Methods: From a research perspective, we focused on the automated segmentation of CMB targets in susceptibility weighted imaging (SWI) for the first time and then constructed a deep learning network focused on the segmentation of micro-objects. We collected and marked clinical datasets and proposed a new medical micro-object cascade network (MMOC-Net). In the first stage, U-Net was utilized to select the region of interest (ROI). In the second stage, we utilized a full-resolution network (FRN) to complete fine segmentation. We also incorporated residual atrous spatial pyramid pooling (R-ASPP) and a new joint loss function.

Results: The most suitable segmentation result was achieved with a ROI size of 32 × 32. To verify the validity of each part of the method, ablation studies were performed, which showed that the best segmentation results were obtained when FRN, R-ASPP and the combined loss function were used simultaneously. Under these conditions, the obtained Dice similarity coefficient (DSC) value was 87.93% and the F2-score (F2) value was 90.69%. We also innovatively developed a visual clinical diagnosis system that can provide effective support for clinical diagnosis and treatment decisions.

Conclusions: We created the MMOC-Net method to perform the automated segmentation task of CMBs in an SWI and obtained better segmentation performance; hence, this pioneering method has research significance.

An Integrated Analysis of Clinical, Genomic, and Imaging Features Reveals Predictors of Neurocognitive Outcomes in a Longitudinal Cohort of Pediatric Cancer Survivors, Enriched with CNS Tumors (Rad ART Pro)

Background

Neurocognitive deficits in pediatric cancer survivors occur frequently; however, individual outcomes are unpredictable. We investigate clinical, genetic, and imaging predictors of neurocognition in pediatric cancer survivors, with a focus on survivors of central nervous system (CNS) tumors exposed to radiation.

Methods

One hundred eighteen patients with benign or malignant cancers (median diagnosis age: 7; 32% embryonal CNS tumors) were selected from an existing multi-institutional cohort (RadART Pro) if they had: 1) neurocognitive evaluation; 2) available DNA; 3) standard imaging. Utilizing RadART Pro, we collected clinical history, genomic sequencing, CNS imaging, and neurocognitive outcomes. We performed single nucleotide polymorphism (SNP) genotyping for candidate genes associated with neurocognition: COMT, BDNF, KIBRA, APOE, KLOTHO. Longitudinal neurocognitive testing were performed using validated computer-based CogState batteries. The imaging cohort was made of patients with available iron-sensitive (n = 28) and/or T2 FLAIR (n = 41) sequences. Cerebral microbleeds (CMB) were identified using a semi-automated algorithm. Volume of T2 FLAIR white matter lesions (WML) was measured using an automated method based on a convolutional neural network. Summary statistics were performed for patient characteristics, neurocognitive assessments, and imaging. Linear mixed effects and hierarchical models assessed patient characteristics and SNP relationship with neurocognition over time. Nested case-control analysis was performed to compare candidate gene carriers to non-carriers.

Results

CMB presence at baseline correlated with worse performance in 3 of 7 domains, including executive function. Higher baseline WML volumes correlated with worse performance in executive function and verbal learning. No candidate gene reliably predicted neurocognitive outcomes; however, APOE ϵ4 carriers trended toward worse neurocognitive function over time compared to other candidate genes and carried the highest odds of low neurocognitive performance across all domains (odds ratio 2.85, P=0.002). Hydrocephalus and seizures at diagnosis were the clinical characteristics most frequently associated with worse performance in neurocognitive domains (5 of 7 domains). Overall, executive function and verbal learning were the most frequently negatively impacted neurocognitive domains.

Conclusion

Presence of CMB, APOE ϵ4 carrier status, hydrocephalus, and seizures correlate with worse neurocognitive outcomes in pediatric cancer survivors, enriched with CNS tumors exposed to radiation. Ongoing research is underway to verify trends in larger cohorts.

Cerebral Microbleeds Assessment and Quantification in COVID-19 Patients With Neurological Manifestations

It is increasingly acknowledged that Coronavirus Disease 2019 (COVID-19) can have neurological manifestations, and cerebral microbleeds (CMBs) have been observed in this setting. The aim of this study was to characterize CMBs patterns on susceptibility-weighted imaging (SWI) in hospitalized patients with COVID-19 with neurological manifestations. CMBs volume was quantified and correlated with clinical and laboratory parameters. The study included patients who were hospitalized due to COVID-19, exhibited neurological manifestations, and underwent a brain MRI between March and May 2020. Neurological, clinical, and biochemical variables were reported. The MRI was acquired using a 3T scanner, with a standardized protocol including SWI. Patients were divided based on radiological evidence of CMBs or their absence. The CMBs burden was also assessed with a semi-automatic SWI processing procedure specifically developed for the purpose of this study. Odds ratios (OR) for CMBs were calculated using age, sex, clinical, and laboratory data by logistic regression analysis. Of the 1,760 patients with COVID-19 admitted to the ASST Papa Giovanni XXIII Hospital between 1 March and 31 May 2020, 116 exhibited neurological symptoms requiring neuroimaging evaluation. Of these, 63 patients underwent brain MRI and were therefore included in the study. A total of 14 patients had radiological evidence of CMBs (CMBs+ group). CMBs+ patients had a higher prevalence of CSF inflammation (p = 0.020), a higher white blood cell count (p = 0.020), and lower lymphocytes (p = 0.010); the D-dimer (p = 0.026), LDH (p = 0.004), procalcitonin (p = 0.002), and CRP concentration (p < 0.001) were higher than in the CMBs- group. In multivariable logistic regression analysis, CRP (OR = 1.16, p = 0.011) indicated an association with CMBs. Estimated CMBs volume was higher in females than in males and decreased with age (Rho = −0.38; p = 0.18); it was positively associated with CRP (Rho = 0.36; p = 0.22), and negatively associated with lymphocytes (Rho = −0.52; p = 0.07). CMBs are a frequent imaging finding in hospitalized patients with COVID-19 with neurological manifestations and seem to be related to pro-inflammatory status.

Automated Detection of Candidate Subjects With Cerebral Microbleeds Using Machine Learning

Cerebral microbleeds (CMBs) appear as small, circular, well defined hypointense lesions of a few mm in size on T2^*-weighted gradient recalled echo (T2^*-GRE) images and appear enhanced on susceptibility weighted images (SWI). Due to their small size, contrast variations and other mimics (e.g., blood vessels), CMBs are highly challenging to detect automatically. In large datasets (e.g., the UK Biobank dataset), exhaustively labelling CMBs manually is difficult and time consuming. Hence it would be useful to preselect candidate CMB subjects in order to focus on those for manual labelling, which is essential for training and testing automated CMB detection tools on these datasets. In this work, we aim to detect CMB candidate subjects from a larger dataset, UK Biobank, using a machine learning-based, computationally light pipeline. For our evaluation, we used 3 different datasets, with different intensity characteristics, acquired with different scanners. They include the UK Biobank dataset and two clinical datasets with different pathological conditions. We developed and evaluated our pipelines on different types of images, consisting of SWI or GRE images. We also used the UK Biobank dataset to compare our approach with alternative CMB preselection methods using non-imaging factors and/or imaging data. Finally, we evaluated the pipeline's generalisability across datasets. Our method provided subject-level detection accuracy > 80% on all the datasets (within-dataset results), and showed good generalisability across datasets, providing a consistent accuracy of over 80%, even when evaluated across different modalities.

Generative Model of Brain Microbleeds for MRI Detection of Vascular Marker of Neurodegenerative Diseases

Cerebral microbleeds (CMB) are increasingly present with aging and can reveal vascular pathologies associated with neurodegeneration. Deep learning-based classifiers can detect and quantify CMB from MRI, such as susceptibility imaging, but are challenging to train because of the limited availability of ground truth and many confounding imaging features, such as vessels or infarcts. In this study, we present a novel generative adversarial network (GAN) that has been trained to generate three-dimensional lesions, conditioned by volume and location. This allows one to investigate CMB characteristics and create large training datasets for deep learning-based detectors. We demonstrate the benefit of this approach by achieving state-of-the-art CMB detection of real CMB using a convolutional neural network classifier trained on synthetic CMB. Moreover, we showed that our proposed 3D lesion GAN model can be applied on unseen dataset, with different MRI parameters and diseases, to generate synthetic lesions with high diversity and without needing laboriously marked ground truth.

7-T Magnetic Resonance Imaging in the Management of Brain Tumors

This article provides an overview of the current status of ultrahigh-field 7-T magnetic resonance (MR) imaging in neuro-oncology, specifically for the management of patients with brain tumors. It includes a discussion of areas across the pretherapeutic, peritherapeutic, and posttherapeutic stages of patient care where 7-T MR imaging is currently being exploited and holds promise. This discussion includes existing technical challenges, barriers to clinical integration, as well as our impression of the future role of 7-T MR imaging as a clinical tool in neuro-oncology.

Cerebral Microbleed Detection via Convolutional Neural Network and Extreme Learning Machine

Aim: Cerebral microbleeds (CMBs) are small round dots distributed over the brain which contribute to stroke, dementia, and death. The early diagnosis is significant for the treatment. Method: In this paper, a new CMB detection approach was put forward for brain magnetic resonance images. We leveraged a sliding window to obtain training and testing samples from input brain images. Then, a 13-layer convolutional neural network (CNN) was designed and trained. Finally, we proposed to utilize an extreme learning machine (ELM) to substitute the last several layers in the CNN for detection. We carried out an experiment to decide the optimal number of layers to be substituted. The parameters in ELM were optimized by a heuristic algorithm named bat algorithm. The evaluation of our approach was based on hold-out validation, and the final predictions were generated by averaging the performance of five runs. Results: Through the experiments, we found replacing the last five layers with ELM can get the optimal results. Conclusion: We offered a comparison with state-of-the-art algorithms, and it can be revealed that our method was accurate in CMB detection.

Synthetic microbleeds generation for classifier training without ground truth

BACKGROUND AND OBJECTIVE

Cerebral microbleeds (CMB) are important biomarkers of cerebrovascular diseases and cognitive dysfunctions. Susceptibility weighted imaging (SWI) is a common MRI sequence where CMB appear as small hypointense blobs. The prevalence of CMB in the population and in each scan is low, resulting in tedious and time-consuming visual assessment. Automated detection methods would be of value but are challenged by the CMB low prevalence, the presence of mimics such as blood vessels, and the difficulty to obtain sufficient ground truth for training and testing. In this paper, synthetic CMB (sCMB) generation using an analytical model is proposed for training and testing machine learning methods. The main aim is creating perfect synthetic ground truth as similar as reals, in high number, with a high diversity of shape, volume, intensity, and location to improve training of supervised methods.

METHOD

sCMB were modelled with a random Gaussian shape and added to healthy brain locations. We compared training on our synthetic data to standard augmentation techniques. We performed a validation experiment using sCMB and report result for whole brain detection using a 10-fold cross validation design with an ensemble of 10 neural networks.

RESULTS

Performance was close to state of the art (~9 false positives per scan), when random forest was trained on synthetic only and tested on real lesion. Other experiments showed that top detection performance could be achieved when training on synthetic CMB only. Our dataset is made available, including a version with 37,000 synthetic lesions, that could be used for benchmarking and training.

CONCLUSION

Our proposed synthetic microbleeds model is a powerful data augmentation approach for CMB classification with and should be considered for training automated lesion detection system from MRI SWI.

DEEPMIR: a deep neural network for differential detection of cerebral microbleeds and iron deposits in MRI

Lobar cerebral microbleeds (CMBs) and localized non-hemorrhage iron deposits in the basal ganglia have been associated with brain aging, vascular disease and neurodegenerative disorders. Particularly, CMBs are small lesions and require multiple neuroimaging modalities for accurate detection. Quantitative susceptibility mapping (QSM) derived from in vivo magnetic resonance imaging (MRI) is necessary to differentiate between iron content and mineralization. We set out to develop a deep learning-based segmentation method suitable for segmenting both CMBs and iron deposits. We included a convenience sample of 24 participants from the MESA cohort and used T2-weighted images, susceptibility weighted imaging (SWI), and QSM to segment the two types of lesions. We developed a protocol for simultaneous manual annotation of CMBs and non-hemorrhage iron deposits in the basal ganglia. This manual annotation was then used to train a deep convolution neural network (CNN). Specifically, we adapted the U-Net model with a higher number of resolution layers to be able to detect small lesions such as CMBs from standard resolution MRI. We tested different combinations of the three modalities to determine the most informative data sources for the detection tasks. In the detection of CMBs using single class and multiclass models, we achieved an average sensitivity and precision of between 0.84-0.88 and 0.40-0.59, respectively. The same framework detected non-hemorrhage iron deposits with an average sensitivity and precision of about 0.75-0.81 and 0.62-0.75, respectively. Our results showed that deep learning could automate the detection of small vessel disease lesions and including multimodal MR data (particularly QSM) can improve the detection of CMB and non-hemorrhage iron deposits with sensitivity and precision that is compatible with use in large-scale research studies.

Novel Approaches to Detection of Cerebral Microbleeds: Single Deep Learning Model to Achieve a Balanced Performance

PURPOSE

Cerebral microbleeds (CMBs) are considered essential indicators for the diagnosis of cerebrovascular disease and cognitive disorders. Traditionally, CMBs are manually interpreted based on criteria including the shape, diameter, and signal characteristics after an MR examination, such as susceptibility-weighted imaging or gradient echo imaging (GRE). In this paper, an efficient method for CMB detection in GRE scans is presented.

MATERIALS AND METHODS

The proposed framework consists of the following phases: (1) pre-processing (skull extraction), (2) the first training with the ground truth labeled using CMB, (3) the second training with the ground truth labeled with CMB mimicking the same subjects, and (4) post-processing (cerebrospinal fluid (CSF) filtering). The proposed technique was validated on a dataset of 1133 CBMs that consisted of 5284 images for training and 1737 images for testing. We applied a two-stage approach using a region-based CNN method based on You Only Look Once (YOLO) to investigate a novel CMB detection technique.

RESULTS

The sensitivity, precision, F1-score and false positive per person (FP_avg) were evaluated as 80.96, 60.98, 69.57 and 6.57, 59.69, 62.70, 61.16 and 4.5, 66.90, 79.75, 72.76 and 2.15 for YOLO with a single label, YOLO with double labels, and YOLO + CSF filtering, respectively, and YOLO + CSF filtering showed the highest precision performance, F1-score and lowest FP_avg.

CONCLUSIONS

Using proposed framework, we developed an optimized CMB learning model with low false positives and a balanced performance in clinical practice.

Relationship between 7T MR-angiography features of vascular injury and cognitive decline in young brain tumor patients treated with radiation therapy

PURPOSE

Although radiation therapy (RT) is a common treatment for pediatric brain tumors, it is associated with detrimental long-term effects such as impaired cognition, vascular injury, and increased stroke risk. This study aimed to develop metrics that describe vascular injury and relate them to the presence of cerebral microbleeds (CMBs) and cognitive performance scores.

METHODS

Twenty-five young adult survivors of pediatric brain tumors treated with either whole-brain (n = 12), whole-ventricular (n = 7), or no RT (n = 6) underwent 7T MRI and neurocognitive testing. Simultaneously acquired MR angiography and susceptibility-weighted images were used to segment CMBs and vessels and quantify their radii and volume.

RESULTS

Patients treated with whole-brain RT had significantly lower arterial volumes (p = 0.003) and a higher proportion of smaller vessels (p = 0.003) compared to the whole-ventricular RT and non-irradiated control patients. Normalized arterial volume decreased with increasing CMB count (R = - 0.66, p = 0.003), and decreasing trends were observed with time since RT and at longitudinal follow-up. Global cognition and verbal memory significantly decreased with smaller normalized arterial volume (p ≤ 0.05).

CONCLUSIONS

Arterial volume is reduced with increasing CMB presence and is influenced by the total brain volume exposed to radiation. This work highlights the potential use of vascular-derived metrics as non-invasive markers of treatment-induced injury and cognitive impairment in pediatric brain tumor patients.

Identification of occult cerebral microbleeds in adults with immune thrombocytopenia

Management of symptoms and prevention of life-threatening hemorrhage in immune thrombocytopenia (ITP) must be balanced against adverse effects of therapies. Because current treatment guidelines based on platelet count are confounded by variable bleeding phenotypes, there is a need to identify new objective markers of disease severity for treatment stratification. In this cross-sectional prospective study of 49 patients with ITP and nadir platelet counts <30 × 109/L and 18 aged-matched healthy controls, we used susceptibility-weighted magnetic resonance imaging to detect cerebral microbleeds (CMBs) as a marker of occult hemorrhage. CMBs were detected using a semiautomated method and correlated with clinical metadata using multivariate regression analysis. No CMBs were detected in health controls. In contrast, lobar CMBs were identified in 43% (21 of 49) of patients with ITP; prevalence increased with decreasing nadir platelet count (0/4, ≥15 × 109/L; 2/9, 10-14 × 109/L; 4/11, 5-9 × 109/L; 15/25 <5 × 109/L) and was associated with longer disease duration (P = 7 × 10-6), lower nadir platelet count (P = .005), lower platelet count at time of neuroimaging (P = .029), and higher organ bleeding scores (P = .028). Mucosal and skin bleeding scores, number of previous treatments, age, and sex were not associated with CMBs. Occult cerebral microhemorrhage is common in patients with moderate to severe ITP. Strong associations with ITP duration may reflect CMB accrual over time or more refractory disease. Further longitudinal studies in children and adults will allow greater understanding of the natural history and clinical and prognostic significance of CMBs.

Automated detection of cerebral microbleeds on T2*-weighted MRI

Cerebral microbleeds, observed as small, spherical hypointense regions on gradient echo (GRE) or susceptibility weighted (SWI) magnetic resonance imaging (MRI) sequences, reflect small hemorrhagic infarcts, and are associated with conditions such as vascular dementia, small vessel disease, cerebral amyloid angiopathy, and Alzheimer’s disease. The current gold standard for detecting and rating cerebral microbleeds in a research context is visual inspection by trained raters, a process that is both time consuming and subject to poor reliability. We present here a novel method to automate microbleed detection on GRE and SWI images. We demonstrate in a community-based cohort of older adults that the method is highly sensitive (greater than 92% of all microbleeds accurately detected) across both modalities, with reasonable precision (fewer than 20 and 10 false positives per scan on GRE and SWI, respectively). We also demonstrate that the algorithm can be used to identify microbleeds over longitudinal scans with a higher level of sensitivity than visual ratings (50% of longitudinal microbleeds correctly labeled by the algorithm, while manual ratings was 30% or lower). Further, the algorithm identifies the anatomical localization of microbleeds based on brain atlases, and greatly reduces time spent completing visual ratings (43% reduction in visual rating time). Our automatic microbleed detection instrument is ideal for implementation in large-scale studies that include cross-sectional and longitudinal scanning, as well as being capable of performing well across multiple commonly used MRI modalities.

A Two Cascaded Network Integrating Regional-based YOLO and 3D-CNN for Cerebral Microbleeds Detection

Cerebral Microbleeds (CMBs) are small chronic brain hemorrhages, which have been considered as diagnostic indicators for different cerebrovascular diseases including stroke, dysfunction, dementia, and cognitive impairment. In this paper, we propose a fully automated two-stage integrated deep learning approach for efficient CMBs detection, which combines a regional-based You Only Look Once (YOLO) stage for potential CMBs candidate detection and three-dimensional convolutional neural networks (3D-CNN) stage for false positives reduction. Both stages are conducted using the 3D contextual information of microbleeds from the MR susceptibility-weighted imaging (SWI) and phase images. However, we average the adjacent slices of SWI and complement the phase images independently and utilize them as a two- channel input for the regional-based YOLO method. The results in the first stage show that the proposed regional-based YOLO efficiently detected the CMBs with an overall sensitivity of 93.62% and an average number of false positives per subject (FP_avg) of 52.18 throughout the five-folds cross-validation. The 3D-CNN based second stage further improved the detection performance by reducing the FP_avg to 1.42. The outcomes of this work might provide useful guidelines towards applying deep learning algorithms for automatic CMBs detection.

Automated detection of cerebral microbleeds in MR images: A two-stage deep learning approach

Cerebral Microbleeds (CMBs) are small chronic brain hemorrhages, which have been considered as diagnostic indicators for different cerebrovascular diseases including stroke, dysfunction, dementia, and cognitive impairment. However, automated detection and identification of CMBs in Magnetic Resonance (MR) images is a very challenging task due to their wide distribution throughout the brain, small sizes, and the high degree of visual similarity between CMBs and CMB mimics such as calcifications, irons, and veins. In this paper, we propose a fully automated two-stage integrated deep learning approach for efficient CMBs detection, which combines a regional-based You Only Look Once (YOLO) stage for potential CMBs candidate detection and three-dimensional convolutional neural networks (3D-CNN) stage for false positives reduction. Both stages are conducted using the 3D contextual information of microbleeds from the MR susceptibility-weighted imaging (SWI) and phase images. However, we average the adjacent slices of SWI and complement the phase images independently and utilize them as a two-channel input for the regional-based YOLO method. This enables YOLO to learn more reliable and representative hierarchal features and hence achieve better detection performance. The proposed work was independently trained and evaluated using high and low in-plane resolution data, which contained 72 subjects with 188 CMBs and 107 subjects with 572 CMBs, respectively. The results in the first stage show that the proposed regional-based YOLO efficiently detected the CMBs with an overall sensitivity of 93.62% and 78.85% and an average number of false positives per subject (FPavg) of 52.18 and 155.50 throughout the five-folds cross-validation for both the high and low in-plane resolution data, respectively. These findings outperformed results by previously utilized techniques such as 3D fast radial symmetry transform, producing fewer FPavg and lower computational cost. The 3D-CNN based second stage further improved the detection performance by reducing the FPavg to 1.42 and 1.89 for the high and low in-plane resolution data, respectively. The outcomes of this work might provide useful guidelines towards applying deep learning algorithms for automatic CMBs detection.

Rate of radiation-induced microbleed formation on 7T MRI relates to cognitive impairment in young patients treated with radiation therapy for a brain tumor

BACKGROUND

Radiation therapy (RT) is essential to the management of many brain tumors, but has been known to lead to cognitive decline and vascular injury in the form of cerebral microbleeds (CMBs).

PURPOSE

In a subset of children, adolescents, and young adults recruited from a larger trial investigating arteriopathy and stroke risk after RT, we evaluated the prevalence of CMBs after RT, examined risk factors for CMBs and cognitive impairment, and related their longitudinal development to cognitive performance changes.

METHODS

Twenty-five patients (mean 17 years, range: 10-25 years) underwent 7-Tesla MRI and cognitive assessment. Nineteen patients were treated with whole-brain or focal RT 1-month to 20-years prior, while 6 non-irradiated patients with posterior-fossa tumors served as controls. CMBs were detected on 7T susceptibility-weighted imaging (SWI) using semi-automated software, a first use in this population.

RESULTS

CMB detection sensitivity with 7T SWI was higher than previously reported at lower field strengths, with one or more CMBs detected in 100% of patients treated with RT at least 1-year prior. CMBs were localized to dose-targeted brain volumes with risk factors including whole-brain RT (p = 0.05), a higher RT dose (p = 0.01), increasing time since RT (p = 0.03), and younger age during RT (p = 0.01). Apart from RT dose, these factors were associated with impaired memory performance. Follow-up data in a subset of patients revealed a proportional increase in CMB count with worsening verbal memory performance (r = -0.85, p = 0.03).

CONCLUSIONS

Treatment with RT during youth is associated with the chronic development of CMBs that evolve with memory impairment over time.

Non-Human Primates Receiving High-Dose Total-Body Irradiation are at Risk of Developing Cerebrovascular Injury Years Postirradiation

Nuclear accidents and acts of terrorism have the potential to expose thousands of people to high-dose total-body iradiation (TBI). Those who survive the acute radiation syndrome are at risk of developing chronic, degenerative radiation-induced injuries [delayed effects of acute radiation (DEARE)] that may negatively affect quality of life. A growing body of literature suggests that the brain may be vulnerable to radiation injury at survivable doses, yet the long-term consequences of high-dose TBI on the adult brain are unclear. Herein we report the occurrence of lesions consistent with cerebrovascular injury, detected by susceptibility-weighted magnetic resonance imaging (MRI), in a cohort of non-human primate [(NHP); rhesus macaque, Macaca mulatta] long-term survivors of high-dose TBI (1.1-8.5 Gy). Animals were monitored longitudinally with brain MRI (approximately once every three years). Susceptibility-weighted images (SWI) were reviewed for hypointensities (cerebral microbleeds and/or focal necrosis). SWI hypointensities were noted in 13% of irradiated NHP; lesions were not observed in control animals. A prior history of exposure was correlated with an increased risk of developing a lesion detectable by MRI (P = 0.003). Twelve of 16 animals had at least one brain lesion present at the time of the first MRI evaluation; a subset of animals (n = 7) developed new lesions during the surveillance period (3.7-11.3 years postirradiation). Lesions occurred with a predilection for white matter and the gray-white matter junction. The majority of animals with lesions had one to three SWI hypointensities, but some animals had multifocal disease (n = 2). Histopathologic evaluation of deceased animals within the cohort (n = 3) revealed malformation of the cerebral vasculature and remodeling of the blood vessel walls. There was no association between comorbid diabetes mellitus or hypertension with SWI lesion status. These data suggest that long-term TBI survivors may be at risk of developing cerebrovascular injury years after irradiation.

Toward Automatic Detection of Radiation-Induced Cerebral Microbleeds Using a 3D Deep Residual Network

Cerebral microbleeds, which are small focal hemorrhages in the brain that are prevalent in many diseases, are gaining increasing attention due to their potential as surrogate markers of disease burden, clinical outcomes, and delayed effects of therapy. Manual detection is laborious and automatic detection and labeling of these lesions is challenging using traditional algorithms. Inspired by recent successes of deep convolutional neural networks in computer vision, we developed a 3D deep residual network that can distinguish true microbleeds from false positive mimics of a previously developed technique based on traditional algorithms. A dataset of 73 patients with radiation-induced cerebral microbleeds scanned at 7 T with susceptibility-weighted imaging was used to train and evaluate our model. With the resulting network, we maintained 95% of the true microbleeds in 12 test patients and the average number of false positives was reduced by 89%, achieving a detection precision of 71.9%, higher than existing published methods. The likelihood score predicted by the network was also evaluated by comparing to a neuroradiologist's rating, and good correlation was observed.

Cerebral Micro-Bleeding Detection Based on Densely Connected Neural Network

Cerebral micro-bleedings (CMBs) are small chronic brain hemorrhages that have many side effects. For example, CMBs can result in long-term disability, neurologic dysfunction, cognitive impairment and side effects from other medications and treatment. Therefore, it is important and essential to detect CMBs timely and in an early stage for prompt treatment. In this research, because of the limited labeled samples, it is hard to train a classifier to achieve high accuracy. Therefore, we proposed employing Densely connected neural network (DenseNet) as the basic algorithm for transfer learning to detect CMBs. To generate the subsamples for training and test, we used a sliding window to cover the whole original images from left to right and from top to bottom. Based on the central pixel of the subsamples, we could decide the target value. Considering the data imbalance, the cost matrix was also employed. Then, based on the new model, we tested the classification accuracy, and it achieved 97.71%, which provided better performance than the state of art methods.

Risk factors of radiotherapy-induced cerebral microbleeds and serial analysis of their size compared with white matter changes: A 7T MRI study in 113 adult patients with brain tumors

BACKGROUND

Although radiation therapy (RT) contributes to survival benefit in many brain tumor patients, it has also been associated with long-term brain injury. Cerebral microbleeds (CMBs) represent an important manifestation of radiation-related injury.

PURPOSE

To characterize the change in size and number of CMBs over time and to evaluate their relationship to white matter structural integrity as measured using diffusion MRI indices.

STUDY TYPE

Longitudinal, retrospective, human cohort.

POPULATION

In all, 113 brain tumor patients including patients treated with focal RT (n = 91, 80.5%) and a subset of nonirradiated controls (n = 22, 19.5%).

FIELD STRENGTH/SEQUENCE

Single and multiecho susceptibility-weighted imaging (SWI) and multiband, shell, and direction diffusion tensor imaging (DTI) at 7 T.

ASSESSMENT

Patients were scanned either once or serially. CMBs were detected and quantified on SWI images using a semiautomated approach. Local and global fractional anisotropy (FA) were measured from DTI data for a subset of 35 patients.

STATISTICAL TESTS

Potential risk factors for CMB development were determined by multivariate linear regression and using linear mixed-effect models. Longitudinal FA was quantitatively and qualitatively evaluated for trends.

RESULTS

All patients scanned at 1 or more years post-RT had CMBs. A history of multiple surgical resections was a risk factor for development of CMBs. The total number and volume of CMBs increased by 18% and 11% per year, respectively, although individual CMBs decreased in volume over time. Simultaneous to these microvascular changes, FA decreased by a median of 6.5% per year. While the majority of nonirradiated controls had no CMBs, four control patients presented with fewer than five CMBs.

DATA CONCLUSION

Identifying patients who are at the greatest risk for CMB development, with its likely associated long-term cognitive impairment, is an important step towards developing and piloting preventative and/or rehabilitative measures for patients undergoing RT.

LEVEL OF EVIDENCE

3 Technical Efficacy: Stage 4 J. Magn. Reson. Imaging 2019;50:868-877.

Macroscale White Matter Alterations Due to Traumatic Cerebral Microhemorrhages Are Revealed by Diffusion Tensor Imaging

With the advent of susceptibility-weighted imaging (SWI), the ability to identify cerebral microbleeds (CMBs) associated with mild traumatic brain injury (mTBI) has become increasingly commonplace. Nevertheless, the clinical significance of post-traumatic CMBs remains controversial partly because it is unclear whether mTBI-related CMBs entail brain circuitry disruptions which, although structurally subtle, are functionally significant. This study combines magnetic resonance and diffusion tensor imaging (MRI and DTI) to map white matter (WM) circuitry differences across 6 months in 26 healthy control volunteers and in 26 older mTBI victims with acute CMBs of traumatic etiology. Six months post-mTBI, significant changes (p < 0.001) in the mean fractional anisotropy of perilesional WM bundles were identified in 21 volunteers, and an average of 47% (σ = 21%) of TBI-related CMBs were associated with such changes. These results suggest that CMBs can be associated with lasting changes in perilesional WM properties, even relatively far from CMB locations. Future strategies for mTBI care will likely rely on the ability to assess how subtle circuitry changes impact neural/cognitive function. Thus, assessing CMB effects upon the structural connectome can play a useful role when studying CMB sequelae and their potential impact upon the clinical outcome of individuals with concussion.

A user-guided tool for semi-automated cerebral microbleed detection and volume segmentation: Evaluating vascular injury and data labelling for machine learning

Background and purpose

With extensive research efforts in place to address the clinical relevance of cerebral microbleeds (CMBs), there remains a need for fast and accurate methods to detect and quantify CMB burden. Although some computer-aided detection algorithms have been proposed in the literature with high sensitivity, their specificity remains consistently poor. More sophisticated machine learning methods appear to be promising in their ability to minimize false positives (FP) through high-level feature extraction and the discrimination of hard-mimics. To achieve superior performance, these methods require sizable amounts of precisely labelled training data. Here we present a user-guided tool for semi-automated CMB detection and volume segmentation, offering high specificity for routine use and FP labelling capabilities to ease and expedite the process of generating labelled training data.

Materials and methods

Existing computer-aided detection methods reported by our group were extended to include fully-automated segmentation and user-guided CMB classification with FP labelling. The algorithm's performance was evaluated on a test set of ten patients exhibiting radiotherapy-induced CMBs on MR images.

Results

The initial algorithm's base sensitivity was maintained at 86.7%. FP's were reduced to inter-rater variations and segmentation results were in 98% agreement with ground truth labelling. There was an approximate 5-fold reduction in the time users spent evaluating CMB burden with the algorithm versus without computer aid. The Intra-class Correlation Coefficient for inter-rater agreement was 0.97 CI[0.92,0.99].

Conclusions

This development serves as a valuable tool for routine evaluation of CMB burden and data labelling to improve CMB classification with machine learning. The algorithm is available to the public on GitHub (https://github.com/LupoLab-UCSF/CMB_labeler).

Cerebral Microbleeds: Imaging and Clinical Significance

Cerebral microbleeds (CMBs), also referred to as microhemorrhages, appear on magnetic resonance (MR) images as hypointense foci notably at T2*-weighted or susceptibility-weighted (SW) imaging. CMBs are detected with increasing frequency because of the more widespread use of high magnetic field strength and of newer dedicated MR imaging techniques such as three-dimensional gradient-echo T2*-weighted and SW imaging. The imaging appearance of CMBs is mainly because of changes in local magnetic susceptibility and reflects the pathologic iron accumulation, most often in perivascular macrophages, because of vasculopathy. CMBs are depicted with a true-positive rate of 48%-89% at 1.5 T or 3.0 T and T2*-weighted or SW imaging across a wide range of diseases. False-positive "mimics" of CMBs occur at a rate of 11%-24% and include microdissections, microaneurysms, and microcalcifications; the latter can be differentiated by using phase images. Compared with postmortem histopathologic analysis, at least half of CMBs are missed with premortem clinical MR imaging. In general, CMB detection rate increases with field strength, with the use of three-dimensional sequences, and with postprocessing methods that use local perturbations of the MR phase to enhance T2* contrast. Because of the more widespread availability of high-field-strength MR imaging systems and growing use of SW imaging, CMBs are increasingly recognized in normal aging, and are even more common in various disorders such as Alzheimer dementia, cerebral amyloid angiopathy, stroke, and trauma. Rare causes include endocarditis, cerebral autosomal dominant arteriopathy with subcortical infarcts, leukoencephalopathy, and radiation therapy. The presence of CMBs in patients with stroke is increasingly recognized as a marker of worse outcome. Finally, guidelines for adjustment of anticoagulant therapy in patients with CMBs are under development. ^© RSNA, 2018.

Automated algorithm for counting microbleeds in patients with familial cerebral cavernous malformations

PURPOSE

Familial cerebral cavernous malformation (CCM) patients present with multiple lesions that can grow both in number and size over time and are reliably detected on susceptibility-weighted imaging (SWI). Manual counting of lesions is arduous and subject to high variability. We aimed to develop an automated algorithm for counting CCM microbleeds (lesions <5 mm in diameter) on SWI images.

METHODS

Fifty-seven familial CCM type-1 patients were included in this institutional review board-approved study. Baseline SWI (n = 57) and follow-up SWI (n = 17) were performed on a 3T Siemens MR scanner with lesions counted manually by the study neuroradiologist. We modified an algorithm for detecting radiation-induced microbleeds on SWI images in brain tumor patients, using a training set of 22 manually delineated CCM microbleeds from two random scans. Manual and automated counts were compared using linear regression with robust standard errors, intra-class correlation (ICC), and paired t tests. A validation analysis comparing the automated counting algorithm and a consensus read from two neuroradiologists was used to calculate sensitivity, the proportion of microbleeds correctly identified by the automated algorithm.

RESULTS

Automated and manual microbleed counts were in strong agreement in both baseline (ICC = 0.95, p < 0.001) and longitudinal (ICC = 0.88, p < 0.001) analyses, with no significant difference between average counts (baseline p = 0.11, longitudinal p = 0.29). In the validation analysis, the algorithm correctly identified 662 of 1325 microbleeds (sensitivity=50%), again with strong agreement between approaches (ICC = 0.77, p < 0.001).

CONCLUSION

The automated algorithm is a consistent method for counting microbleeds in familial CCM patients that can facilitate lesion quantification and tracking.

Automatic cerebral microbleeds detection from MR images via Independent Subspace Analysis based hierarchical features

With the development of susceptibility weighted imaging (SWI) technology, cerebral microbleed (CMB) detection is increasingly essential in cerebrovascular diseases diagnosis and cognitive impairment assessment. Clinical CMB detection is based on manual rating which is subjective and time-consuming with limited reproducibility. In this paper, we propose a computer-aided system for automatic detection of CMBs from brain SWI images. Our approach detects the CMBs within three stages: (i) candidates screening based on intensity values (ii) compact 3D hierarchical features extraction via a stacked convolutional Independent Subspace Analysis (ISA) network (iii) false positive candidates removal with a support vector machine (SVM) classifier based on the learned representation features from ISA. Experimental results on 19 subjects (161 CMBs) achieve a high sensitivity of 89.44% with an average of 7.7 and 0.9 false positives per subject and per CMB, respectively, which validate the efficacy of our approach.

Automated detection of cerebral microbleeds in patients with Traumatic Brain Injury

In this paper a Computer Aided Detection (CAD) system is presented to automatically detect Cerebral Microbleeds (CMBs) in patients with Traumatic Brain Injury (TBI). It is believed that the presence of CMBs has clinical prognostic value in TBI patients. To study the contribution of CMBs in patient outcome, accurate detection of CMBs is required. Manual detection of CMBs in TBI patients is a time consuming task that is prone to errors, because CMBs are easily overlooked and are difficult to distinguish from blood vessels.

This study included 33 TBI patients. Because of the laborious nature of manually annotating CMBs, only one trained expert manually annotated the CMBs in all 33 patients. A subset of ten TBI patients was annotated by six experts. Our CAD system makes use of both Susceptibility Weighted Imaging (SWI) and T1 weighted magnetic resonance images to detect CMBs. After pre-processing these images, a two-step approach was used for automated detection of CMBs. In the first step, each voxel was characterized by twelve features based on the dark and spherical nature of CMBs and a random forest classifier was used to identify CMB candidate locations. In the second step, segmentations were made from each identified candidate location. Subsequently an object-based classifier was used to remove false positive detections of the voxel classifier, by considering seven object-based features that discriminate between spherical objects (CMBs) and elongated objects (blood vessels). A guided user interface was designed for fast evaluation of the CAD system result. During this process, an expert checked each CMB detected by the CAD system.

A Fleiss' kappa value of only 0.24 showed that the inter-observer variability for the TBI patients in this study was very large. An expert using the guided user interface reached an average sensitivity of 93%, which was significantly higher (p = 0.03) than the average sensitivity of 77% (sd 12.4%) that the six experts manually detected. Furthermore, with the use of this CAD system the reading time was substantially reduced from one hour to 13 minutes per patient, because the CAD system only detects on average 25.9 false positives per TBI patient, resulting in 0.29 false positives per definite CMB finding.

•

We present an automated detection system for microbleeds in MRIs of trauma patients.

•

When using the system, detection time goes down from one hour to 13 min.

•

The system enables the user to detect a significantly higher number of CMBs.

•

The inter-observer variability for detecting CMBs is very large in TBI patients.

Automatic Detection of Cerebral Microbleeds From MR Images via 3D Convolutional Neural Networks

Cerebral microbleeds (CMBs) are small haemorrhages nearby blood vessels. They have been recognized as important diagnostic biomarkers for many cerebrovascular diseases and cognitive dysfunctions. In current clinical routine, CMBs are manually labelled by radiologists but this procedure is laborious, time-consuming, and error prone. In this paper, we propose a novel automatic method to detect CMBs from magnetic resonance (MR) images by exploiting the 3D convolutional neural network (CNN). Compared with previous methods that employed either low-level hand-crafted descriptors or 2D CNNs, our method can take full advantage of spatial contextual information in MR volumes to extract more representative high-level features for CMBs, and hence achieve a much better detection accuracy. To further improve the detection performance while reducing the computational cost, we propose a cascaded framework under 3D CNNs for the task of CMB detection. We first exploit a 3D fully convolutional network (FCN) strategy to retrieve the candidates with high probabilities of being CMBs, and then apply a well-trained 3D CNN discrimination model to distinguish CMBs from hard mimics. Compared with traditional sliding window strategy, the proposed 3D FCN strategy can remove massive redundant computations and dramatically speed up the detection process. We constructed a large dataset with 320 volumetric MR scans and performed extensive experiments to validate the proposed method, which achieved a high sensitivity of 93.16% with an average number of 2.74 false positives per subject, outperforming previous methods using low-level descriptors or 2D CNNs by a significant margin. The proposed method, in principle, can be adapted to other biomarker detection tasks from volumetric medical data.

Computer-aided detection of cerebral microbleeds in susceptibility-weighted imaging

Susceptibility-weighted imaging (SWI) is recognized as the preferred MRI technique for visualizing cerebral vasculature and related pathologies such as cerebral microbleeds (CMBs). Manual identification of CMBs is time-consuming, has limited reliability and reproducibility, and is prone to misinterpretation. In this paper, a novel computer-aided microbleed detection technique based on machine learning is presented: First, spherical-like objects (potential CMB candidates) with their corresponding bounding boxes were detected using a novel multi-scale Laplacian of Gaussian technique. A set of robust 3-dimensional Radon- and Hessian-based shape descriptors within each bounding box were then extracted to train a cascade of binary random forests (RF). The cascade consists of consecutive independent RF classifiers with low to high posterior probability constraints to handle imbalanced training sets (CMBs and non-CMBs), and to progressively improve detection rates. The proposed method was validated on 66 subjects whose CMBs were manually stratified into "possible" and "definite" by two medical experts. The proposed technique achieved a sensitivity of 87% and an average false detection rate of 27.1 CMBs per subject on the "possible and definite" set. A sensitivity of 93% and false detection rate of 10 CMBs per subject was also achieved on the "definite" set. The proposed automated approach outperforms state of the art methods, and promises to enhance manual expert screening. Benefits include improved reliability, minimization of intra-rater variability and a reduction in assessment time.

Simultaneous imaging of radiation-induced cerebral microbleeds, arteries and veins, using a multiple gradient echo sequence at 7 Tesla

BACKGROUND

The purpose of this study was to implement and evaluate the utility of a multi-echo sequence at 7 Tesla (T) for simultaneous time-of-flight (TOF) MR-angiography (MRA) and susceptibility-weighted imaging (SWI) of radiation-induced cerebral microbleeds (CMBs), intracranial arteries, and veins.

METHODS

A four-echo gradient-echo sequence was implemented on a 7T scanner. The first echo was used to create TOF-MRA images, while the remaining echoes were combined to visualize CMBs and veins on SWI images. The sequence was evaluated on eight brain tumor patients with known radiation-induced CMBs. Single-echo images were also acquired to visually and quantitatively compare the contrast-to-noise ratio (CNR) of small- and intermediate-vessels between acquisitions. The number of CMBs detected with each acquisition was also quantified. Statistical significance was determined using a Wilcoxon signed-rank test.

RESULTS

Compared with the single-echo sequences, the CNR of small and intermediate arteries increased 7.6% (P < 0.03) and 9.5% (P = 0.06), respectively, while the CNR of small and intermediate veins were not statistically different between sequences (P = 0.95 and P = 0.46, respectively). However, these differences were not discernible by visual inspection. Also the multi-echo sequence detected 18.3% more CMBs (P < 0.008) due to higher slice resolution.

CONCLUSION

The proposed 7T multi-echo sequence depicts arteries, veins, and CMBs on a single image to facilitate quantitative evaluation of radiation-induced vascular injury.

A CAD System for Hemorrhagic Stroke

Computer-aided detection/diagnosis (CAD) is a key component of routine clinical practice, increasingly used for detection, interpretation, quantification and decision support. Despite a critical need, there is no clinically accepted CAD system for stroke yet. Here we introduce a CAD system for hemorrhagic stroke. This CAD system segments, quantifies, and displays hematoma in 2D/3D, and supports evacuation of hemorrhage by thrombolytic treatment monitoring progression and quantifying clot removal. It supports seven-step workflow: select patient, add a new study, process patient's scans, show segmentation results, plot hematoma volumes, show 3D synchronized time series hematomas, and generate report. The system architecture contains four components: library, tools, application with user interface, and hematoma segmentation algorithm. The tools include a contour editor, 3D surface modeler, 3D volume measure, histogramming, hematoma volume plot, and 3D synchronized time-series hematoma display. The CAD system has been designed and implemented in C++. It has also been employed in the CLEAR and MISTIE phase-III, multicenter clinical trials. This stroke CAD system is potentially useful in research and clinical applications, particularly for clinical trials.

Semi-Automated Detection of Cerebral Microbleeds on 3.0 T MR Images

Cerebral microbleeds are associated with vascular disease and dementia. They can be detected on MRI and receive increasing attention. Visual rating is the current standard for microbleed detection, but is rater dependent, has limited reproducibility, modest sensitivity, and can be time-consuming. The goal of the current study is to present a tool for semi-automated detection of microbleeds that can assist human raters in the rating procedure. The radial symmetry transform is originally a technique to highlight circular-shaped objects in two-dimensional images. In the current study, the three-dimensional radial symmetry transform was adapted to detect spherical microbleeds in a series of 72 patients from our hospital, for whom a ground truth visual rating was made by four raters. Potential microbleeds were automatically identified on T2*-weighted 3.0 T MRI scans and the results were visually checked to identify microbleeds. Final ratings of the radial symmetry transform were compared to human ratings. After implementing and optimizing the radial symmetry transform, the method achieved a high sensitivity, while maintaining a modest number of false positives. Depending on the settings, sensitivities ranged from 65%-84% compared to the ground truth rating. Rating of the processed images required 1-2 minutes per participant, in which 20-96 false positive locations per participant were censored. Sensitivities of individual raters ranged from 39%-86% compared to the ground truth and required 5-10 minutes per participant per rater. The sensitivities that were achieved by the radial symmetry transform are similar to those of individual experienced human raters, demonstrating its feasibility and usefulness for semi-automated microbleed detection.