A multimodal medical image contrastive learning algorithm with domain adaptive denormalization / 生物医学工程学杂志

Recently, deep learning has achieved impressive results in medical image tasks. However, this method usually requires large-scale annotated data, and medical images are expensive to annotate, so it is a challenge to learn efficiently from the limited annotated data. Currently, the two commonly used methods are transfer learning and self-supervised learning. However, these two methods have been little studied in multimodal medical images, so this study proposes a contrastive learning method for multimodal medical images. The method takes images of different modalities of the same patient as positive samples, which effectively increases the number of positive samples in the training process and helps the model to fully learn the similarities and differences of lesions on images of different modalities, thus improving the model's understanding of medical images and diagnostic accuracy. The commonly used data augmentation methods are not suitable for multimodal images, so this paper proposes a domain adaptive denormalization method to transform the source domain images with the help of statistical information of the target domain. In this study, the method is validated with two different multimodal medical image classification tasks: in the microvascular infiltration recognition task, the method achieves an accuracy of (74.79 ± 0.74)% and an F1 score of (78.37 ± 1.94)%, which are improved as compared with other conventional learning methods; for the brain tumor pathology grading task, the method also achieves significant improvements. The results show that the method achieves good results on multimodal medical images and can provide a reference solution for pre-training multimodal medical images.

A meta-learning based method for segmentation of few-shot magnetic resonance images / 生物医学工程学杂志

When applying deep learning algorithms to magnetic resonance (MR) image segmentation, a large number of annotated images are required as data support. However, the specificity of MR images makes it difficult and costly to acquire large amounts of annotated image data. To reduce the dependence of MR image segmentation on a large amount of annotated data, this paper proposes a meta-learning U-shaped network (Meta-UNet) for few-shot MR image segmentation. Meta-UNet can use a small amount of annotated image data to complete the task of MR image segmentation and obtain good segmentation results. Meta-UNet improves U-Net by introducing dilated convolution, which can increase the receptive field of the model to improve the sensitivity to targets of different scales. We introduce the attention mechanism to improve the adaptability of the model to different scales. We introduce the meta-learning mechanism, and employ a composite loss function for well-supervised and effective bootstrapping of model training. We use the proposed Meta-UNet model to train on different segmentation tasks, and then use the trained model to evaluate on a new segmentation task, where the Meta-UNet model achieves high-precision segmentation of target images. Meta-UNet has a certain improvement in mean Dice similarity coefficient (DSC) compared with voxel morph network (VoxelMorph), data augmentation using learned transformations (DataAug) and label transfer network (LT-Net). Experiments show that the proposed method can effectively perform MR image segmentation using a small number of samples. It provides a reliable aid for clinical diagnosis and treatment.

Hepatocellular carcinoma segmentation and pathological differentiation degree prediction method based on multi-task learning / 生物医学工程学杂志

Hepatocellular carcinoma (HCC) is the most common liver malignancy, where HCC segmentation and prediction of the degree of pathological differentiation are two important tasks in surgical treatment and prognosis evaluation. Existing methods usually solve these two problems independently without considering the correlation of the two tasks. In this paper, we propose a multi-task learning model that aims to accomplish the segmentation task and classification task simultaneously. The model consists of a segmentation subnet and a classification subnet. A multi-scale feature fusion method is proposed in the classification subnet to improve the classification accuracy, and a boundary-aware attention is designed in the segmentation subnet to solve the problem of tumor over-segmentation. A dynamic weighted average multi-task loss is used to make the model achieve optimal performance in both tasks simultaneously. The experimental results of this method on 295 HCC patients are superior to other multi-task learning methods, with a Dice similarity coefficient (Dice) of (83.9 ± 0.88)% on the segmentation task, while the average recall is (86.08 ± 0.83)% and an F1 score is (80.05 ± 1.7)% on the classification task. The results show that the multi-task learning method proposed in this paper can perform the classification task and segmentation task well at the same time, which can provide theoretical reference for clinical diagnosis and treatment of HCC patients.

Left ventricle segmentation in echocardiography based on adaptive mean shift / 生物医学工程学杂志

The use of echocardiography ventricle segmentation can obtain ventricular volume parameters, and it is helpful to evaluate cardiac function. However, the ultrasound images have the characteristics of high noise and difficulty in segmentation, bringing huge workload to segment the object region manually. Meanwhile, the automatic segmentation technology cannot guarantee the segmentation accuracy. In order to solve this problem, a novel algorithm framework is proposed to segment the ventricle. Firstly, faster region-based convolutional neural network is used to locate the object to get the region of interest. Secondly, -means is used to pre-segment the image; then a mean shift with adaptive bandwidth of kernel function is proposed to segment the region of interest. Finally, the region growing algorithm is used to get the object region. By this framework, ventricle is obtained automatically without manual localization. Experiments prove that this framework can segment the object accurately, and the algorithm of adaptive mean shift is more stable and accurate than the mean shift with fixed bandwidth on quantitative evaluation. These results show that the method in this paper is helpful for automatic segmentation of left ventricle in echocardiography.

Medical computer-aided detection method based on deep learning / 生物医学工程学杂志

This paper performs a comprehensive study on the computer-aided detection for the medical diagnosis with deep learning. Based on the region convolution neural network and the prior knowledge of target, this algorithm uses the region proposal network, the region of interest pooling strategy, introduces the multi-task loss function: classification loss, bounding box localization loss and object rotation loss, and optimizes it by end-to-end. For medical image it locates the target automatically, and provides the localization result for the next stage task of segmentation. For the detection of left ventricular in echocardiography, proposed additional landmarks such as mitral annulus, endocardial pad and apical position, were used to estimate the left ventricular posture effectively. In order to verify the robustness and effectiveness of the algorithm, the experimental data of ultrasonic and nuclear magnetic resonance images are selected. Experimental results show that the algorithm is fast, accurate and effective.

The clinical investigation of the potential complications of H-UPPP surgery in removing the partial pharyngeal muscle / 临床耳鼻咽喉头颈外科杂志

OBJECTIVE@#In order to improve the postoperative effect of modified UPPP, removing the partial pharyngeal muscle in surgery, we investigate the postoperative effect, the characteristics of pharyngeal cavity and the potential complications in OSAHS patients.@*METHOD@#To choose 82 OSAHS patients with obstructive oropharyngeal plane diagnosed by Apneagraphy (AG), Fibre nasopharyngoscope combined with Müller examination and nasopharyngeal 3D-CT, which had completed clinical data inpatients in the anesthesia underwent of the partial pharyngeal muscles in the postoperative, divided into a control group of 26 cases, operating the H-UPPP surgery which did not remove partial pharyngeal muscle; The experimental group of 56 cases did a H-UPPP surgical which removed partial pharyngeal muscle of possible concurrent symptoms such as nasal regurgitation, Eustachian tube dysfunction and other follow-up study in six months after the monthly telephone follow-up or outpatient exams to understand the disease. Patients were evaluated the sleepiness by ESS(Epworth sleepiness scale) in 6 months after the surgery, compared with the preoperative ESS scores, do a t test for statistical analysis. AG can be used to evaluate effects of the UPPP after 6 months. By measuring uvula length (L1), extent from free edge of soft palate to postpharyngeal (L2) and stenosis of nasopharynx width (L3) mean, we investigate the characteristics of pharyngeal cavity using the multiple linear regression to do the hypothesis test and evaluate the association between measuring mean and effect. Using SPSS19.0 software do the preoperative contrast analysis.@*RESULT@#After 6 months in surgery, 56 cases in the experimental group, effect in 50 cases (89.29%), effective in 6 cases (10.71%); ESS score: Preoperative 11.74 +/- 2.48, after the first 6 months 3.84 +/- 2.05. Twenty-six cases in control group,effect in 19 cases (73.08%), effective in 7 cases (26.92%); ESS score: Preoperative 11.91 +/- 2.40, after the first 6 months 6.92 +/- 2.47, t-test P value of less than 0.05 between the experimental group and the control group; There are no ear fullness, hearing loss, increase their own sound which reflect eustachian tube dysfunction and other complications in two groups; The function of pharyngeal cavity could be recovered normal lever after 6 months; After 6 months of the operation, in the experimental group and the control group L1 mean was respectively (5.91 +/- 3.38) mm and (6.20 +/- 3.76) mm (P>0.05); L2 mean was respectively (15.70 +/- 3.29)mm and (15.35 +/- 1.44) mm (P> 0.05); L3 mean was respectively (20.54 +/- 3.33) mm and (16.43 +/- 2.21) mm (P<0.05). Nasal fauces pitch mean was significantly widened. By the multiple linear regression analysis, the postoperative effect has the linear correlation between L2 and 1,3 residual mean with the negative correlation. Due to the standardized coefficient, L3 residual mean has the most influence on the postoperative effect.@*CONCLUSION@#Modified UPPP surgery removing the partial pharyngeal muscle is in favor of upgrading the postoperative effect with significantly increasing the width of postoperative nasal pharyngeal isthmus area, then there are not occur the eustachian tube dysfunction, the soft palate function, swallowing and articulation function disabled.

Research on failure reasons of surgical treatment of obstructive sleep apnea hypopnea syndrome and reoperation method / 临床耳鼻咽喉头颈外科杂志

OBJECTIVE@#To analyze failure reasons of surgical treatment of obstructive sleep apnea hypopnea syndrome (OSAHS), and explore the methods of reoperation.@*METHOD@#By selecting 27 patients, who accepted surgical treatment for OSAHS and recurred, we analyzed failure reasons and obstructive location by apneagraph, nasopharyngeal 3D-CT, electronic nasopharynlaryngoscope. Among them, 14 patients accepted reoperation, such as uvulopalatopharyngoplasty (UPPP), nasoendoscopic surgery, adenoidectomy, partial glossectomy, tracheotomy were applied matching to differential obstructive location. AHI, lowest SaO2, Epworth sleepiness scale (ESS), complication were recorded after 6 months.@*RESULT@#After 6 months, their AHI decreased from 48.19 +/- 13.11 to 11.32 +/- 4. 42, ESS scores decreased from 12.93 +/- 4.60 to 4.93 +/- 1.44, P<0.05. Two of the 14 patients were cured, while the other 12 were efficient. No complications were observed.@*CONCLUSION@#Obstructive location judgement and proper surgical operation are the keys of the treatment. Preoperative AG sleep monitoring, nasopharyngeal 3D CT, electronic nasopharynlaryngoscope examination for determining blocking plane, the decision of surgery which is significant.

Significance of AG in OSAHS operation treatment guidance and evaluation of postoperative efficacy / 临床耳鼻咽喉头颈外科杂志

OBJECTIVE@#Apneagraph can be used to discuss which the best operation scheme is for OSAHS. Effects of uvulopalatopharyngoplasty can be assessed by Apneagraph in obstructive sleep apnea hypopnea syndrome (OSAHS) patients.@*METHOD@#Fifty-six patients with OSAHS received the modified UPPP operation were randomly selected in our hospital. The AG and PSG were applied for diagnosis and evaluation of operation effects. The sleepiness state was assessed by ESS (Epworth sleepiness scale) 6 months after the surgery, compared with the preoperative ESS scores using attest for statistical analysis. We used the SPSS19.0 software to carry our data analysis.@*RESULT@#After 6 months, the evaluation of postoperative efficacy came out to be completely controlled in 42 cases (75%), significantly effective in 14 cases (25%), and uncured in 0 cases. Correlation between the transpalatal obstruction proportion and the AHI reduction percentage was significantly positive (r = 0.667). There were 38 patients with oropharynx obstruction percentage more than 73.35% presented completely controlled in 34 cases (89.47%), significantly effective in 4 cases (10.33%), and uncured in 0 cases.@*CONCLUSION@#AG has the dual functions of analyzing sleep-related respiratory disturbance events and determining upper airway obstruction sites. AG application in the postoperative evaluation of modified uppp has significantly objective guide significance. The modified UPPP for treatment of OSAHS can improve the operation effect. Patients with oropharynx obstruction percentage more than 73.5% don't need to receive the operation for treatment of retroglottal region.