Probing the intravascular and interstitial compartments of remodeled myocardium in heart failure patients with preserved and reduced ejection fraction: a CMR study

Intelligent skin-removal photoacoustic computed tomography for human based on deep learning

Photoacoustic computed tomography (PACT) has centimeter-level imaging ability and can be used to detect the human body. However, strong photoacoustic signals from skin cover deep tissue information, hindering the frontal display and analysis of photoacoustic images of deep regions of interest. Therefore, we propose a 2.5 D deep learning model based on feature pyramid structure and single-type skin annotation to extract the skin region, and design a mask generation algorithm to remove skin automatically. PACT imaging experiments on the human periphery blood vessel verified the correctness our proposed skin-removal method. Compared with previous studies, our method exhibits high robustness to the uneven illumination, irregular skin boundary, and reconstruction artifacts in the images, and the reconstruction errors of PACT images decreased by 20% ~ 90% with a 1.65 dB improvement in the signal-to-noise ratio at the same time. This study may provide a promising way for high-definition PACT imaging of deep tissues.

Tile-based microscopic image processing for malaria screening using a deep learning approach

BACKGROUND

Manual microscopic examination remains the golden standard for malaria diagnosis. But it is laborious, and pathologists with experience are needed for accurate diagnosis. The need for computer-aided diagnosis methods is driven by the enormous workload and difficulties associated with manual microscopy based examination. While the importance of computer-aided diagnosis is increasing at an enormous pace, fostered by the advancement of deep learning algorithms, there are still challenges in detecting small objects such as malaria parasites in microscopic images of blood films. The state-of-the-art (SOTA) deep learning-based object detection models are inefficient in detecting small objects accurately because they are underrepresented on benchmark datasets. The performance of these models is affected by the loss of detailed spatial information due to in-network feature map downscaling. This is due to the fact that the SOTA models cannot directly process high-resolution images due to their low-resolution network input layer.

METHODS

In this study, an efficient and robust tile-based image processing method is proposed to enhance the performance of malaria parasites detection SOTA models. Three variants of YOLOV4-based object detectors are adopted considering their detection accuracy and speed. These models were trained using tiles generated from 1780 high-resolution P. falciparum-infected thick smear microscopic images. The tiling of high-resolution images improves the performance of the object detection models. The detection accuracy and the generalization capability of these models have been evaluated using three datasets acquired from different regions.

RESULTS

The best-performing model using the proposed tile-based approach outperforms the baseline method significantly (Recall, [95.3%] vs [57%] and Average Precision, [87.1%] vs [76%]). Furthermore, the proposed method has outperformed the existing approaches that used different machine learning techniques evaluated on similar datasets.

CONCLUSIONS

The experimental results show that the proposed method significantly improves P. falciparum detection from thick smear microscopic images while maintaining real-time detection speed. Furthermore, the proposed method has the potential to assist and reduce the workload of laboratory technicians in malaria-endemic remote areas of developing countries where there is a critical skill gap and a shortage of experts.

Identification of normal and depression EEG signals in variational mode decomposition domain

Early detection of depression is critical in assisting patients in receiving the best therapy possible to avoid negative repercussions. Depression detection using electroencephalogram (EEG) signals is a simple, low-cost, convenient, and accurate approach. This paper proposes a six-stage novel method for detecting depression using EEG signals. First, EEG signals are recorded from 44 subjects, with 22 subjects being normal and 22 subjects being depressed. Second, a simple notch filter with EEG signals differencing approach is employed for effective preprocessing. Third, the variational mode decomposition (VMD) approach is implemented for nonlinear and non-stationary EEG signals analysis, resulting in many modes. Fourth, mutual information-based novel modes selection criterion is proposed to select the most informative modes. In the fifth step, a combination of linear and nonlinear features are extracted from selected modes and at last, classification is performed with neural networks. In this study, a novel single feature is also proposed, which is made using Log energy, norm entropies and fluctuation index, which delivers 100% classification accuracy, sensitivity and specificity. By using these features, a novel depression diagnostic index is also proposed. This integrated index would assist in quicker and more objective identification of normal and depression EEG signals. The proposed computerized framework and the DDI can help health workers, large enterprises, and product developers build a real-time system.

A rare low-grade myofibroblastic sarcoma in lower jaw with the resemblance to benign lesions

Background

Low-grade myofibroblastic sarcoma (LGMS) is a rare solid infiltrative soft tissue tumor with a predilection for the head and neck region.

Case presentation

We report the diagnostic steps of a fast-growing lesion of the lower left jaw in a 45-year-old otherwise healthy woman. A first biopsy and subsequent histopathological examination showed potential differentials of a benign myofibroma, benign nodular fasciitis or an LGMS. This diagnostic overlap was a challenge for the decision of the further treatment approach. The treatment consisted of a segmental en bloc resection of the mandible including the second premolar, first and second molar. Histopathological examination of the resected tumor confirmed an LGMS.

Conclusion

The histopathologic resemblance of LGMS to a range of benign and reactive tumors may lead to misdiagnosis and mistreatment. The rarity of LGMS explains the lack of established treatment protocols. This case shows the importance of adequate clinical decisions, expertise in the histopathology of rare tumors and interdisciplinary exchange to achieve state-of-the-art patient management.

A systematic review: normative reference values of the median nerve cross-sectional area using ultrasonography in healthy individuals

Median nerve cross-sectional area (CSA) was used for screening and diagnosis of neuropathy, but few studies have suggested reference range. Hence, this systematic review was performed to evaluate a normative values of median nerve CSA at various landmarks of upper limb based on ultrasonography. PubMed and Web of science were used to search relevant articles from 2000 to 2020. Forty-one eligible articles (2504 nerves) were included to access median nerve CSA at different landmarks (mid-arm, elbow, mid-forearm, carpal tunnel (CT) inlet and CT outlet). Data was also stratified based on age, sex, ethnicity, geographical location, and method of measurement. Random effects model was used to calculate pooled weighted mean (95% confidence interval (CI), [upper bound, lower bound]) at mid-arm, elbow, mid-forearm, CT inlet and outlet which found to be 8.81 mm², CI [8.10, 9.52]; 8.57 mm² [8.00, 9.14]; 7.07 mm² [6.41, 7.73]; 8.74 mm² [8.45, 9.03] and 9.02 mm² [8.08, 9.95] respectively. Median nerve CSA varies with age, geographical location, and sex at all landmarks. A low (I² < 25%) to considerable heterogeneity (I² > 75%) was observed, indicating the variation among the included studies. These findings show that median nerve CSA is varying not only along its course but also in other sub-variables.

The Role of Systemic Microvascular Dysfunction in Heart Failure with Preserved Ejection Fraction

Heart failure with preserved ejection fraction (HFpEF) is a condition with increasing incidence, leading to a health care problem of epidemic proportions for which no curative treatments exist. Consequently, an urge exists to better understand the pathophysiology of HFpEF. Accumulating evidence suggests a key pathophysiological role for coronary microvascular dysfunction (MVD), with an underlying mechanism of low-grade pro-inflammatory state caused by systemic comorbidities. The systemic entity of comorbidities and inflammation in HFpEF imply that patients develop HFpEF due to systemic mechanisms causing coronary MVD, or systemic MVD. The absence or presence of peripheral MVD in HFpEF would reflect HFpEF being predominantly a cardiac or a systemic disease. Here, we will review the current state of the art of cardiac and systemic microvascular dysfunction in HFpEF (Graphical Abstract), resulting in future perspectives on new diagnostic modalities and therapeutic strategies.

A novel classification method of lymph node metastasis in colorectal cancer

Colorectal cancer lymph node metastasis, which is highly associated with the patient's cancer recurrence and survival rate, has been the focus of many therapeutic strategies that are highly associated with the patient's cancer recurrence and survival rate. The popular methods for classification of lymph node metastasis by neural networks, however, show limitations as the available low-level features are inadequate for classification, and the radiologists are unable to quickly review the images. Identifying lymph node metastasis in colorectal cancer is a key factor in the treatment of patients with colorectal cancer. In the present work, an automatic classification method based on deep transfer learning was proposed. Specifically, the method resolved the problem of repetition of low-level features and combined these features with high-level features into a new feature map for classification; and a merged layer which merges all transmitted features from previous layers into a map of the first full connection layer. With a dataset collected from Harbin Medical University Cancer Hospital, the experiment involved a sample of 3,364 patients. Among these samples, 1,646 were positive, and 1,718 were negative. The experiment results showed the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were 0.8732, 0.8746, 0.8746 and 0.8728, respectively, and the accuracy and AUC were 0.8358 and 0.8569, respectively. These demonstrated that our method significantly outperformed the previous classification methods for colorectal cancer lymph node metastasis without increasing the depth and width of the model.

Optimization of psoriasis assessment system based on patch images

Psoriasis is a chronic inflammatory skin disease that occurs in various forms throughout the body and is associated with certain conditions such as heart disease, diabetes, and depression. The psoriasis area severity index (PASI) score, a tool used to evaluate the severity of psoriasis, is currently used in clinical trials and clinical research. The determination of severity is based on the subjective judgment of the clinician. Thus, the disease evaluation deviations are induced. Therefore, we propose optimal algorithms that can effectively segment the lesion area and classify the severity. In addition, a new dataset on psoriasis was built, including patch images of erythema and scaling. We performed psoriasis lesion segmentation and classified the disease severity. In addition, we evaluated the best-performing segmentation method and classifier and analyzed features that are highly related to the severity of psoriasis. In conclusion, we presented the optimal techniques for evaluating the severity of psoriasis. Our newly constructed dataset improved the generalization performance of psoriasis diagnosis and evaluation. It proposed an optimal system for specific evaluation indicators of the disease and a quantitative PASI scoring method. The proposed system can help to evaluate the severity of localized psoriasis more accurately.

Hahn-PCNN-CNN: an end-to-end multi-modal brain medical image fusion framework useful for clinical diagnosis

Background

In medical diagnosis of brain, the role of multi-modal medical image fusion is becoming more prominent. Among them, there is no lack of filtering layered fusion and newly emerging deep learning algorithms. The former has a fast fusion speed but the fusion image texture is blurred; the latter has a better fusion effect but requires higher machine computing capabilities. Therefore, how to find a balanced algorithm in terms of image quality, speed and computing power is still the focus of all scholars.

Methods

We built an end-to-end Hahn-PCNN-CNN. The network is composed of feature extraction module, feature fusion module and image reconstruction module. We selected 8000 multi-modal brain medical images downloaded from the Harvard Medical School website to train the feature extraction layer and image reconstruction layer to enhance the network’s ability to reconstruct brain medical images. In the feature fusion module, we use the moments of the feature map combined with the pulse-coupled neural network to reduce the information loss caused by convolution in the previous fusion module and save time.

Results

We choose eight sets of registered multi-modal brain medical images in four diease to verify our model. The anatomical structure images are from MRI and the functional metabolism images are SPECT and 18F-FDG. At the same time, we also selected eight representative fusion models as comparative experiments. In terms of objective quality evaluation, we select six evaluation metrics in five categories to evaluate our model.

Conclusions

The fusion image obtained by our model can retain the effective information in source images to the greatest extent. In terms of image fusion evaluation metrics, our model is superior to other comparison algorithms. In terms of time computational efficiency, our model also performs well. In terms of robustness, our model is very stable and can be generalized to multi-modal image fusion of other organs.

A generative adversarial network-based abnormality detection using only normal images for model training with application to digital breast tomosynthesis

Deep learning has shown tremendous potential in the task of object detection in images. However, a common challenge with this task is when only a limited number of images containing the object of interest are available. This is a particular issue in cancer screening, such as digital breast tomosynthesis (DBT), where less than 1% of cases contain cancer. In this study, we propose a method to train an inpainting generative adversarial network to be used for cancer detection using only images that do not contain cancer. During inference, we removed a part of the image and used the network to complete the removed part. A significant error in completing an image part was considered an indication that such location is unexpected and thus abnormal. A large dataset of DBT images used in this study was collected at Duke University. It consisted of 19,230 reconstructed volumes from 4348 patients. Cancerous masses and architectural distortions were marked with bounding boxes by radiologists. Our experiments showed that the locations containing cancer were associated with a notably higher completion error than the non-cancer locations (mean error ratio of 2.77). All data used in this study has been made publicly available by the authors.

MR-guided stereotactic body radiation therapy for primary cardiac sarcomas

BACKGROUND

Primary cardiac tumors are an extremely rare disease with limited prognosis. The treatment of choice is surgery. Other treatment options include chemotherapy and radiation therapy, which historically represented a palliative approach in patients who were not eligible for surgery. The development of hybrid MR-guided radiation therapy makes it possible to better visualize cardiac lesions and to apply high doses per fraction in sensible organs such as the heart.

CASE PRESENTATION

Patients affected by inoperable primary cardiac sarcomas and treated at two different institutions were considered for this analysis and retrospectively analyzed. All patients were treated using a 0.35 T hybrid MR Linac system (MRIdian, ViewRay Inc., Mountain View, CA). In the present study we investigated the feasibility, early outcome and toxicity of MR-guided RT in primary cardiac sarcomas. Four consecutive non-metastasized patients who were treated between 05-09/2020 were analyzed. The cardiac sarcomas were mostly located in the right atrium (50%) and one patient presented with 3 epicardial lesions. All patients received MRgRT as a salvage treatment for recurrent cardiac sarcoma after initial surgery, after a mean interval of 12 months (range 1-29 months). Regarding the treatment characteristics, the mean GTV size was 22.9 cc (range 2.5-56.9 cc) and patients were treated with a mean GTV dose of 38.9 Gy (range 30.1-41.1 Gy) in 5 fractions. Regarding feasibility, all treatments were completed as planned and all patients tolerated the treatment very well and showed only mild grade 1 or 2 symptoms like fatigue, dyspnea or mild chest pain at early follow-up.

CONCLUSION

To the best of our knowledge, in this retrospective analysis we present the first and largest series of patients presenting with primary cardiac sarcomas treated with online adaptive MRgRT. However, further studies are needed to evaluate the impact of this new methodology on the outcome of this very rare disease.

Deep learning approaches for automated classification and segmentation of head and neck cancers and brain tumors in magnetic resonance images: a meta-analysis study

PURPOSE

Deep learning (DL) has led to widespread changes in automated segmentation and classification for medical purposes. This study is an attempt to use statistical methods to analyze studies related to segmentation and classification of head and neck cancers (HNCs) and brain tumors in MRI images.

METHODS

PubMed, Web of Science, Embase, and Scopus were searched to retrieve related studies published from January 2016 to January 2020. Studies that evaluated the performance of DL-based models in the segmentation, and/or classification and/or grading of HNCs and/or brain tumors were included. Selected studies for each analysis were statistically evaluated based on the diagnostic performance metrics.

RESULTS

The search results retrieved 1,664 related studies, of which 30 studies were eligible for meta-analysis. The overall performance of DL models for the complete tumor in terms of the pooled Dice score, sensitivity, and specificity was 0.8965 (95% confidence interval (95% CI): 0.76-0.9994), 0.9132 (95% CI: 0.71-0.994) and 0.9164 (95% CI: 0.78-1.00), respectively. The DL methods achieved the highest performance for classifying three types of glioma, meningioma, and pituitary tumors with overall accuracies of 96.01%, 99.73%, and 96.58%, respectively. Stratification of glioma tumors by high and low grading revealed overall accuracies of 94.32% and 94.23% for the DL methods, respectively.

CONCLUSION

Based on the obtained results, we can acknowledge the significant ability of DL methods in the mentioned applications. Poor reporting in these studies challenges the analysis process, so it is recommended that future studies report comprehensive results based on different metrics.

Reviewing Machine Learning and Image Processing Based Decision-Making Systems for Breast Cancer Imaging

Breast cancer (BC) is the leading cause of death among women worldwide. It affects in general women older than 40 years old. Medical images analysis is one of the most promising research areas since it provides facilities for diagnosis and decision-making of several diseases such as BC. This paper conducts a Structured Literature Review (SLR) of the use of Machine Learning (ML) and Image Processing (IP) techniques to deal with BC imaging. A set of 530 papers published between 2000 and August 2019 were selected and analyzed according to ten criteria: year and publication channel, empirical type, research type, medical task, machine learning techniques, datasets used, validation methods, performance measures and image processing techniques which include image pre-processing, segmentation, feature extraction and feature selection. Results showed that diagnosis was the most used medical task and that Deep Learning techniques (DL) were largely used to perform classification. Furthermore, we found out that classification was the most ML objective investigated followed by prediction and clustering. Most of the selected studies used Mammograms as imaging modalities rather than Ultrasound or Magnetic Resonance Imaging with the use of public or private datasets with MIAS as the most frequently investigated public dataset. As for image processing techniques, the majority of the selected studies pre-process their input images by reducing the noise and normalizing the colors, and some of them use segmentation to extract the region of interest with the thresholding method. For feature extraction, we note that researchers extracted the relevant features using classical feature extraction techniques (e.g. Texture features, Shape features, etc.) or DL techniques (e. g. VGG16, VGG19, ResNet, etc.), and finally few papers used feature selection techniques in particular the filter methods.

Longitudinal in-vivo OCM imaging of glioblastoma development in the mouse brain

We present in-vivo imaging of the mouse brain using custom made Gaussian beam optical coherence microscopy (OCM) with 800nm wavelength. We applied new instrumentation to longitudinal imaging of the glioblastoma (GBM) tumor microvasculature in the mouse brain. We have introduced new morphometric biomarkers that enable quantitative analysis of the development of GBM. We confirmed quantitatively an intensive angiogenesis in the tumor area between 3 and 14 days after GBM cells injection confirmed by considerably increased of morphometric parameters. Moreover, the OCM setup revealed heterogeneity and abnormality of newly formed vessels.

Deep semi-supervised learning for brain tumor classification

Background

This paper addresses issues of brain tumor, glioma, classification from four modalities of Magnetic Resonance Image (MRI) scans (i.e., T1 weighted MRI, T1 weighted MRI with contrast-enhanced, T2 weighted MRI and FLAIR). Currently, many available glioma datasets often contain some unlabeled brain scans, and many datasets are moderate in size.

Methods

We propose to exploit deep semi-supervised learning to make full use of the unlabeled data. Deep CNN features were incorporated into a new graph-based semi-supervised learning framework for learning the labels of the unlabeled data, where a new 3D-2D consistent constraint is added to make consistent classifications for the 2D slices from the same 3D brain scan. A deep-learning classifier is then trained to classify different glioma types using both labeled and unlabeled data with estimated labels. To alleviate the overfitting caused by moderate-size datasets, synthetic MRIs generated by Generative Adversarial Networks (GANs) are added in the training of CNNs.

Results

The proposed scheme has been tested on two glioma datasets, TCGA dataset for IDH-mutation prediction (molecular-based glioma subtype classification) and MICCAI dataset for glioma grading. Our results have shown good performance (with test accuracies 86.53% on TCGA dataset and 90.70% on MICCAI dataset).

Conclusions

The proposed scheme is effective for glioma IDH-mutation prediction and glioma grading, and its performance is comparable to the state-of-the-art.

Technical report: dynamic assessment of plantar fasciitis and plantar fascia tears utilising dorsiflexion of the great toe

Discrimination between plantar fasciitis and partial tears of the plantar fascia can be difficult on ultrasound given laxity of the plantar fascia in the region of its calcaneal insertion and anisotropy. Dynamic assessment with great toe dorsiflexion can improve visualisation of the proximal portion of the plantar fascia on ultrasound, by straightening the plantar fascia due to the windlass mechanism. This article describes the technique and its anatomical basis.