A Two-stage Diagnostic Framework for Post-ablation Treatment Response Assessment in Patients with Hepatocellular Carcinoma.

Contrast-enhanced ultrasound (CEUS) video plays an important role in post-ablation treatment response assessment in patients with hepatocellular carcinoma (HCC). However, the assessment of treatment response using CEUS video is challenging due to issues such as high inter-frame data repeatability, small ablation area and poor imaging quality of CEUS video. To address these issues, we propose a two-stage diagnostic framework for post-ablation treatment response assessment in patients with HCC using CEUS video. The first stage is a location stage, which is used to locate the ablation area. At this stage, we propose a Yolov5-SFT to improve the location results of the ablation area and a similarity comparison module (SCM) to reduce data repeatability. The second stage is an assessment stage, which is used for the evaluation of postoperative efficacy. At this stage, we design an EfficientNet-SK to improve assessment accuracy. The Experimental results on the self-collected data show that the proposed framework outperforms other selected algorithms, and can effectively assist doctors in the assessment of post-ablation treatment response.

Corneal Lamb wave imaging for quantitative assessment of collagen cross-linking treatment based on comb-push ultrasound shear elastography.

Keratoconus, a serious corneal disorder, often causes highly irregular astigmatism and different degrees of visual impairment. Riboflavin/UVA corneal collagen cross-linking(CXL) is currently approved for effective treatment of keratoconus by enhancing the mechanical strength of collagen fibers in the cornea. However, few methods are capable of quantitatively and non-destructively assessing the mechanical properties of the cornea before and after CXL treatments. This study developed a corneal viscoelasticity imaging method based on comb-push ultrasound shear elastography (CUSE) and implemented this method on a Verasonics™ Vantage 256 ultrasound open system with a high-frequency linear array ultrasound transducer. Push beams were generated by three teeth each consisting of 10 elements (working frequency = 10.41 MHz) for inducing Lamb wave propagation in the cornea, and then the system immediately switched to the plane wave imaging mode using 60 elements in the middle (working frequency = 18 MHz). This method can provide a high-resolution 2D Lamb wave velocity image overlapping with a B-mode image as well as quantitative viscoelasticity estimation according to experimentally obtained phase velocity dispersion of Lamb waves. The validation experiments were performed on ex vivo porcine corneas, and the accuracy of elasticity estimation was verified by a tensile test. The results showed that the shear elasticity increased and the viscosity decreased after CXL treatment. The shear elasticity results (reported as mean ±â€¯standard deviation) of one control group with no CXL treatment and three CXL-treated groups named as 10 min, 30 min, and 60 min groups according to UV irradiation time were 14.62 ±â€¯3.38 kPa, 49.47 ±â€¯3.63 kPa, 116.54 ±â€¯23.99 kPa, and 197.89 ±â€¯39.64 kPa, respectively, which was in agreement with the results of tensile tests. The ultrasound safety measurement indicated that this method could have acceptable safety, but further to ocular tissue and vision function. The study demonstrated the possibility of using a commercial ultrasound system to obtain high-resolution images of corneal mechanical properties as well as the ability to quantify changes induced by CXL treatment. Therefore, the proposed method could serve as a helpful tool in the studies related in corneal biomechanics.

Multi-task learning for quality assessment of fetal head ultrasound images.

It is essential to measure anatomical parameters in prenatal ultrasound images for the growth and development of the fetus, which is highly relied on obtaining a standard plane. However, the acquisition of a standard plane is, in turn, highly subjective and depends on the clinical experience of sonographers. In order to deal with this challenge, we propose a new multi-task learning framework using a faster regional convolutional neural network (MF R-CNN) architecture for standard plane detection and quality assessment. MF R-CNN can identify the critical anatomical structure of the fetal head and analyze whether the magnification of the ultrasound image is appropriate, and then performs quality assessment of ultrasound images based on clinical protocols. Specifically, the first five convolution blocks of the MF R-CNN learn the features shared within the input data, which can be associated with the detection and classification tasks, and then extend to the task-specific output streams. In training, in order to speed up the different convergence of different tasks, we devise a section train method based on transfer learning. In addition, our proposed method also uses prior clinical and statistical knowledge to reduce the false detection rate. By identifying the key anatomical structure and magnification of the ultrasound image, we score the ultrasonic plane of fetal head to judge whether it is a standard image or not. Experimental results on our own-collected dataset show that our method can accurately make a quality assessment of an ultrasound plane within half a second. Our method achieves promising performance compared with state-of-the-art methods, which can improve the examination effectiveness and alleviate the measurement error caused by improper ultrasound scanning.

In vivo assessment of the mechanical properties of crystalline lenses in a rabbit model using ultrasound elastography: Effects of ultrasound frequency and age.

PURPOSE: The occurrence of presbyopia and cataract is closely related to changes in the mechanical properties of the crystalline lens. There are no established methods so far for in vivo assessment. By introducing ultrasound elastography, we proposed group velocity (Vg) of an induced shear wave as a new biomarker to characterize the mechanical properties of the lens in our previous study. Here, we investigated the effect of the ultrasound frequency on measurement accuracy and validated the results with a conventional ex vivo compression testing. We also demonstrated a change trend in Vg and its correlation with age in a rabbit model. METHODS: Eight New Zealand white rabbits were fed normally from the fourth to seventh month. An ultrasound elastography system was developed to measure Vgin vivo on every eye once per month. The performances when using a high-frequency (L22-11v) and low-frequency (L11-4v) probe were compared. Rabbits were sacrificed after in vivo measurements by the end of the seventh month and this was followed by ex vivo ultrasound measurements and conventional compression tests on the extracted lenses. RESULTS: The results demonstrated that there were no significance differences in Vg between measurements with high-frequency (USE-HF) and low-frequency (USE-LF) probes in the same month-age group. The mean Vg and the standard deviation of four rabbits that were 7 months old were 2.37 ±â€¯0.24  m/s, 2.36 ±â€¯0.25 m/s, 2.43 ±â€¯0.26 m/s and 2.44 ±â€¯0.38 m/s, with USE-HF for ex vivo and in vivo measurements and USE-LF for ex vivo and in vivo measurements, respectively. There were no significant differences (p > 0.05) and they were all in agreement with the results of compression tests, which was 16.16 ±â€¯1.84 kPa in Young's modulus. The results also showed that Vg increased with age. In combination with the results of our previous study, Vg showed a relatively sharp increase from 2 to 5 months, while it had a slight increase from 5 to 7 months. CONCLUSIONS: The USE-HF and USE-LF has comparable accuracy in Vg measurements while USE-HF had an advantage regarding better spatial resolution. The change trend of Vg was in accord with the growth phase of New Zealand white rabbits, which usually results in sexual maturity at 5 months old. This implies that Vg can be used as a biomarker parameter for evaluating the mechanical properties of the lens undergoing physiological changes.

Role of acoustic radiation force impulse imaging elastography in the assessment of steatohepatitis and fibrosis in rat models.

Acoustic radiation force impulse (ARFI) elastography is a non-invasive method for performing liver assessment via liver shear wave velocity (SWV) measurements. The aim of this study was to evaluate the performance of the ARFI technique in the diagnosis of nonalcoholic steatohepatitis (NASH) and fibrosis and to investigate the effect of steatosis and inflammation on liver fibrosis SWV measurements in a rat model of nonalcoholic fatty liver disease (NAFLD). The ex vivo right liver lobes from 110 rats were processed and embedded in a fabricated gelatin phantom, and the other lobes were used for histologic assessment. The SWV induced by acoustic radiation force was derived to evaluate liver stiffness. The experimental results showed that the liver SWV value could be used to differentiate non-NASH rats from NASH-presenting rats and NASH from cirrhosis, and these comparisons showed areas under the receiver operating characteristic curves (AUROC) of 0.951 and 0.980, respectively. The diagnostic performances of ARFI elastography in predicting severe fibrosis (F ≥ 3) and cirrhosis (F ≥ 4) showed AUROC values of 0.997 and 0.993, respectively. In rats with mild fibrosis (F0-F1), severe steatosis had a significant effect on the mean SWV values. In rats with significant fibrosis (F2-F4), severe lobular inflammation had significant effects on the mean SWV values. Our findings indicate that ARFI elastography is a promising method for differentiating non-NASH rats from NASH rats and for staging hepatic fibrosis in NASH. The presence of severe steatosis and severe lobular inflammation are significant factors for evaluating fibrosis stages.

Noninvasive assessment of age-related stiffness of crystalline lenses in a rabbit model using ultrasound elastography.

BACKGROUND: The pathological or physiological changes of a crystalline lens directly affect the eye accommodation and transmittance, and then they increase the risk of presbyopia and cataracts for people in the middle and old age groups. There is no universally accepted quantitative method to measure the lens' mechanical properties in vivo so far. This study aims to investigate the possibility of assessing the age-related stiffness change of crystalline lens by acoustic-radiation-force-based ultrasound elastography (ARF-USE) in a rabbit model in vivo. METHODS: There were 13 New Zealand white rabbits that were divided into four groups and fed normally until they were 60 (n = 4), 90 (n = 2), 120 (n = 4), and 150 (n = 3) days old, respectively. An ARF-USE platform was built based on the Verasonics™ Vantage 256 system. The shear waves were excited and traced in the lens by a linear ultrasound probe after a rabbit was anaesthetized. RESULTS: The average group velocities were 1.38 ± 0.2 m/s, 2.06 ± 0.3 m/s, 2.07 ± 0.29 m/s, and 2.30 ± 0.28 m/s, respectively, for the four groups of rabbits. The results shows that the group velocity has a strong correlation with the day age (r = 0.84, p < 1 × 10-7) and the weight (r = 0.83, p < 1×10-7) of the rabbits while the maximum displacement has no correlations with the day age (r = 0.27, p > 0.1) and the weight (r = 0.32, p > 0.1). CONCLUSION: This study demonstrated that the group velocity measured by ARF-USE had a strong correlation with age-related stiffness in a rabbit model, suggesting that group velocity is a good biomarker to characterize the stiffness of a crystalline lens. This study also demonstrated the feasibility of using this USE technique to assess the mechanical properties of the lens in vivo for clinical or research purposes.

Automation-assisted cervical cancer screening in manual liquid-based cytology with hematoxylin and eosin staining.

Current automation-assisted technologies for screening cervical cancer mainly rely on automated liquid-based cytology slides with proprietary stain. This is not a cost-efficient approach to be utilized in developing countries. In this article, we propose the first automation-assisted system to screen cervical cancer in manual liquid-based cytology (MLBC) slides with hematoxylin and eosin (H&E) stain, which is inexpensive and more applicable in developing countries. This system consists of three main modules: image acquisition, cell segmentation, and cell classification. First, an autofocusing scheme is proposed to find the global maximum of the focus curve by iteratively comparing image qualities of specific locations. On the autofocused images, the multiway graph cut (GC) is performed globally on the a* channel enhanced image to obtain cytoplasm segmentation. The nuclei, especially abnormal nuclei, are robustly segmented by using GC adaptively and locally. Two concave-based approaches are integrated to split the touching nuclei. To classify the segmented cells, features are selected and preprocessed to improve the sensitivity, and contextual and cytoplasm information are introduced to improve the specificity. Experiments on 26 consecutive image stacks demonstrated that the dynamic autofocusing accuracy was 2.06 µm. On 21 cervical cell images with nonideal imaging condition and pathology, our segmentation method achieved a 93% accuracy for cytoplasm, and a 87.3% F-measure for nuclei, both outperformed state of the art works in terms of accuracy. Additional clinical trials showed that both the sensitivity (88.1%) and the specificity (100%) of our system are satisfyingly high. These results proved the feasibility of automation-assisted cervical cancer screening in MLBC slides with H&E stain, which is highly desirable in community health centers and small hospitals.