Annotation Cost Minimization for Ultrasound Image Segmentation using Cross-domain Transfer Learning.

Deep learning techniques can help minimize inter-physician analysis variability and the medical expert workloads, thereby enabling more accurate diagnoses. However, their implementation requires large-scale annotated dataset whose acquisition incurs heavy time and human-expertise costs. Hence, to significantly minimize the annotation cost, this study presents a novel framework that enables the deployment of deep learning methods in ultrasound (US) image segmentation requiring only very limited manually annotated samples. We propose SegMix, a fast and efficient approach that exploits a segment-paste-blend concept to generate large number of annotated samples based on a few manually acquired labels. Besides, a series of US-specific augmentation strategies built upon image enhancement algorithms are introduced to make maximum use of the available limited number of manually delineated images. The feasibility of the proposed framework is validated on the left ventricle (LV) segmentation and fetal head (FH) segmentation tasks, respectively. Experimental results demonstrate that using only 10 manually annotated images, the proposed framework can achieve a Dice and JI of 82.61% and 83.92%, and 88.42% and 89.27% for LV segmentation and FH segmentation, respectively. Compared with training using the entire training set, there is over 98% of annotation cost reduction while achieving comparable segmentation performance. This indicates that the proposed framework enables satisfactory deep leaning performance when very limited number of annotated samples is available. Therefore, we believe that it can be a reliable solution for annotation cost reduction in medical image analysis.

US-Net: A lightweight network for simultaneous speckle suppression and texture enhancement in ultrasound images.

BACKGROUND: Numerous traditional filtering approaches and deep learning-based methods have been proposed to improve the quality of ultrasound (US) image data. However, their results tend to suffer from over-smoothing and loss of texture and fine details. Moreover, they perform poorly on images with different degradation levels and mainly focus on speckle reduction, even though texture and fine detail enhancement are of crucial importance in clinical diagnosis. METHODS: We propose an end-to-end framework termed US-Net for simultaneous speckle suppression and texture enhancement in US images. The architecture of US-Net is inspired by U-Net, whereby a feature refinement attention block (FRAB) is introduced to enable an effective learning of multi-level and multi-contextual representative features. Specifically, FRAB aims to emphasize high-frequency image information, which helps boost the restoration and preservation of fine-grained and textural details. Furthermore, our proposed US-Net is trained essentially with real US image data, whereby real US images embedded with simulated multi-level speckle noise are used as an auxiliary training set. RESULTS: Extensive quantitative and qualitative experiments indicate that although trained with only one US image data type, our proposed US-Net is capable of restoring images acquired from different body parts and scanning settings with different degradation levels, while exhibiting favorable performance against state-of-the-art image enhancement approaches. Furthermore, utilizing our proposed US-Net as a pre-processing stage for COVID-19 diagnosis results in a gain of 3.6% in diagnostic accuracy. CONCLUSIONS: The proposed framework can help improve the accuracy of ultrasound diagnosis.

An efficient annotated data generation method for echocardiographic image segmentation.

BACKGROUND: In recent years, deep learning techniques have demonstrated promising performances in echocardiography (echo) data segmentation, which constitutes a critical step in the diagnosis and prognosis of cardiovascular diseases (CVDs). However, their successful implementation requires large number and high-quality annotated samples, whose acquisition is arduous and expertise-demanding. To this end, this study aims at circumventing the tedious, time-consuming and expertise-demanding data annotation involved in deep learning-based echo data segmentation. METHODS: We propose a two-phase framework for fast generation of annotated echo data needed for implementing intelligent cardiac structure segmentation systems. First, multi-size and multi-orientation cardiac structures are simulated leveraging polynomial fitting method. Second, the obtained cardiac structures are embedded onto curated endoscopic ultrasound images using Fourier Transform algorithm, resulting in pairs of annotated samples. The practical significance of the proposed framework is validated through using the generated realistic annotated images as auxiliary dataset to pretrain deep learning models for automatic segmentation of left ventricle and left ventricle wall in real echo data, respectively. RESULTS: Extensive experimental analyses indicate that compared with training from scratch, fine-tuning after pretraining with the generated dataset always results in significant performance improvement whereby the improvement margins in terms of Dice and IoU can reach 12.9% and 7.74%, respectively. CONCLUSION: The proposed framework has great potential to overcome the shortage of labeled data hampering the deployment of deep learning approaches in echo data analysis.

Adaptive Multiple Subtraction Based on an Accelerating Iterative Curvelet Thresholding Method.

In the seismic exploration, recorded data contain primaries and multiples, where primaries, as signals of interest, can be used to image the subsurface geology. Surface-related multiple elimination (SRME), one important class of multiple attenuation algorithms, operates in two stages, multiple prediction and subtraction. Due to the phase and amplitude errors in the predicted multiples, adaptive multiple subtraction (AMS) is the key step of SRME. The main challenge of this technique resides in removing multiples without distorting primaries. The curvelet-based AMS methods, which exploit the sparsity of primary and multiple in curvelet domain and the misfit between the original and estimated signals in data domain, have shown outstanding performances in real seismic data processing. These methods are realized by using the iterative curvelet thresholding (ICT), which has heavy computation burden since it includes two forward/inverse curvelet transform (CuT) pairs in each iteration. To ameliorate the computational cost, we propose an accelerating ICT method by exploiting the misfit between the original and estimated signals in curvelet domain directly. Since the proposed method only needs do one forward/inverse CuT pair, it is faster than the traditional ICT method. Considering that the error of the predicted multiple is frequency-dependent, we furthermore introduce the joint constraints within different frequency bands to stabilize and improve the multiple attenuation. Synthetic and field examples demonstrate that the proposed method outperforms the traditional ICT method. In addition, the proposed method has shown to be suitable for refining other AMS methods' results, yielding a SNR improvement of 0.5-2.8 dB.

Arbuscular mycorrhizal fungi alleviate boron toxicity in Puccinellia tenuiflora under the combined stresses of salt and drought.

To investigate the effect of arbuscular mycorrhizal fungi (AMF) on boron (B) toxicity in plants under the combined stresses of salt and drought, Puccinellia tenuiflora was grown in the soil with the inoculation of Funneliformis mosseae and Claroideoglomus etunicatum. After three weeks of treatment, the plants were harvested to determine mycorrhizal colonization rates, plant biomass, as well as tissue B, phosphorus, sodium, and potassium concentrations. The results show that the combined stresses reduced mycorrhizal colonization. Mycorrhizal inoculation significantly increased plant biomass while reduced shoot B concentrations. Mycorrhizal inoculation also slightly increased shoot phosphorus and potassium concentrations, and reduced shoot sodium concentrations. F. mosseae and C. etunicatum were able to alleviate the combined stresses of B, salt, and drought. The two fungal species and their combination showed no significant difference in the alleviation of B toxicity. It is inferred that AMF is able to alleviate B toxicity in P. tenuiflora by increasing biomass and reducing tissue B concentrations. The increase in plant phosphorus and potassium, as well as the decrease in sodium accumulation that induced by AMF, can help plant tolerate the combined stresses of salt and drought. Our findings suggest that F. mosseae and C. etunicatum are potential candidates for facilitating the phytoremediation of B-contaminated soils with salt and drought stress.

Blind image deconvolution using the zero-lag slice of higher-order statistics.

A novel algorithm for blind image deconvolution using the zero-lag slice (ZLS) of higher-order statistics only is presented. This method first estimates the point-spread function (PSF) using the ZLS of its third-order moment (TOM) and then uses it with one of the known classical image deconvolution methods. The proposed method has simple computations for PSF estimation because it solves a nonlinear problem by using an iterative method with fast convergence. In each iteration, one need only calculate the ZLS of the TOM and estimate the PSF using simple two-dimensional operations. Furthermore, the method presented achieves good results, since the ZLS estimate obtained from the degraded image exhibits high reliability. The good performance of the proposed algorithm is demonstrated by applying it to synthetic and real data sets.

[Syndrome inappropriate ADH secretion after allogeneic hematopoietic stem cell transplantation: a case report and literature review].

OBJECTIVE: To study the syndrome of inappropriate ADH secretion (SIADH) after allogeneic hematopoietic stem cell transplantation (allo-HSCT) and the possible etiology. METHODS: The clinical manifestation, laboratory examination, treatment and outcome of a patient with refractory anemia with excess blasts after allo-HSCT were presented. RESULTS: Hyperacute graft-versus-host disease (GVHD) was developed in the patient after allo-HSCT followed by severe hyponatraemia (lowest serum sodium 103.7 mmol/L), natriuresis, hypo-osmolality of plasma, hyper-osmolality of urine, coma and twitch at day 17 after allo-HSCT. SIADH was diagnosed. The clinical condition was improved after restriction of water and administration of hypertonic saline, but SIADH was not controlled completely. Afterwards, graft failure was developed. Hyperacute GVHD and graft rejection occurred again after the second transplant. The patient died of secondary infection. CONCLUSION: SIADH after allo-HSCT is a rare fatal acute complication of central nervous system. Numerous transplant-related causes are probably associated with the development of SIADH. Early accurate diagnosis and treatment promptly is of great importance.

Automatic removal of the eye blink artifact from EEG using an ICA-based template matching approach.

Independent component analysis (ICA) proves to be effective in the removing the ocular artifact from electroencephalogram recordings (EEG). While using ICA in ocular artifact correction, a crucial step is to correctly identify the artifact components among the decomposed independent components. In most previous works, this step of selecting the artifact components was manually implemented, which is time consuming and inconvenient when dealing with a large amount of EEG data. We present a new method which automatically selects the eye blink artifact components based on the pattern of their scalp topographies, which can be exemplified as a template matching approach. The feasibility of using a fixed template for singling out the eye blink component after ICA decomposition was validated by an experiment in which 18 subjects among the 21 subjects involved exhibited a highly consistent pattern of eye blink scalp topographies. Since only the spatial feature is employed for singling out the eye blink component, the proposed method is very efficient and easy to implement. Objective evaluation of the real results shows that the proposed algorithm can remove the eye blink artifact from the EEG while causing little distortion to the underlying brain activities.

A technique for on-board CT reconstruction using both kilovoltage and megavoltage beam projections for 3D treatment verification.

The technologies with kilovoltage (kV) and megavoltage (MV) imaging in the treatment room are now available for image-guided radiation therapy to improve patient setup and target localization accuracy. However, development of strategies to efficiently and effectively implement these technologies for patient treatment remains challenging. This study proposed an aggregated technique for on-board CT reconstruction using combination of kV and MV beam projections to improve the data acquisition efficiency and image quality. These projections were acquired in the treatment room at the patient treatment position with a new kV imaging device installed on the accelerator gantry, orthogonal to the existing MV portal imaging device. The projection images for a head phantom and a contrast phantom were acquired using both the On-Board Imager kV imaging device and the MV portal imager mounted orthogonally on the gantry of a Varian Clinac 21EX linear accelerator. MV projections were converted into kV information prior to the aggregated CT reconstruction. The multilevel scheme algebraic-reconstruction technique was used to reconstruct CT images involving either full, truncated, or a combination of both full and truncated projections. An adaptive reconstruction method was also applied, based on the limited numbers of kV projections and truncated MV projections, to enhance the anatomical information around the treatment volume and to minimize the radiation dose. The effects of the total number of projections, the combination of kV and MV projections, and the beam truncation of MV projections on the details of reconstructed kV/MV CT images were also investigated.

Adaptive algebraic reconstruction technique.

Algebraic reconstruction techniques (ART) are iterative procedures for reconstructing objects from their projections. It is proven that ART can be computationally efficient by carefully arranging the order in which the collected data are accessed during the reconstruction procedure and adaptively adjusting the relaxation parameters. In this paper, an adaptive algebraic reconstruction technique (AART), which adopts the same projection access scheme in multilevel scheme algebraic reconstruction technique (MLS-ART), is proposed. By introducing adaptive adjustment of the relaxation parameters during the reconstruction procedure, one-iteration AART can produce reconstructions with better quality, in comparison with one-iteration MLS-ART. Furthermore, AART outperforms MLS-ART with improved computational efficiency.