Evaluation of pore-fracture microstructure of gypsum rock fragments using micro-CT

This study utilized X-ray micro-computed tomography (micro-CT) to investigate weathered gypsum rocks which can or do serve as a rock substrate for endolithic organisms, focusing on their internal pore-fracture microstructure, estimating porosity, and quantitative comparison between various samples. Examining sections and reconstructed 3D models provides a more detailed insight into the overall structural conditions within rock fragments and the interconnectivity in pore networks, surpassing the limitations of analyzing individual 2D images. Results revealed diverse gypsum forms, cavities, fractures, and secondary features influenced by weathering. Using deep learning segmentation based on the U-Net models within the Dragonfly software enabled to identify and visualize the porous systems and determinate void space which was used to calculate porosity. This approach allowed to describe what type of microstructures and cavities is responsible for the porous spaces in different gypsum samples. A set of quantitative analysis of the detected void and modeled networks provided a needed information about the development of the pore system, connectivity, and pore size distribution. Comparison with mercury intrusion porosimetry showed that both methods consider different populations of pores. In our case, micro-CT typically detects larger pores (> 10 μm) which is related to the effective resolution of the scanned images. Still, micro-CT demonstrated to be an efficient tool in examining the internal microstructures of weathered gypsum rocks, with promising implications particularly in geobiology and microbiology for the characterization of lithic habitats.

Predicting dynamic, motion-related changes in B0 field in the brain at a 7T MRI using a subject-specific fine-trained U-net

PURPOSE

Subject movement during the MR examination is inevitable and causes not only image artifacts but also deteriorates the homogeneity of the main magnetic field (B0 ), which is a prerequisite for high quality data. Thus, characterization of changes to B0 , for example induced by patient movement, is important for MR applications that are prone to B0 inhomogeneities.

METHODS

We propose a deep learning based method to predict such changes within the brain from the change of the head position to facilitate retrospective or even real-time correction. A 3D U-net was trained on in vivo gradient-echo brain 7T MRI data. The input consisted of B0 maps and anatomical images at an initial position, and anatomical images at a different head position (obtained by applying a rigid-body transformation on the initial anatomical image). The output consisted of B0 maps at the new head positions. We further fine-trained the network weights to each subject by measuring a limited number of head positions of the given subject, and trained the U-net with these data.

RESULTS

Our approach was compared to established dynamic B0 field mapping via interleaved navigators, which suffer from limited spatial resolution and the need for undesirable sequence modifications. Qualitative and quantitative comparison showed similar performance between an interleaved navigator-equivalent method and proposed method.

CONCLUSION

It is feasible to predict B0 maps from rigid subject movement and, when combined with external tracking hardware, this information could be used to improve the quality of MR acquisitions without the use of navigators.

Automatic liver segmentation and assessment of liver fibrosis using deep learning with MR T1-weighted images in rats

OBJECTIVES

To validate the performance of nnU-Net in segmentation and CNN in classification for liver fibrosis using T1-weighted images.

MATERIALS AND METHODS

In this prospective study, animal models of liver fibrosis were induced by injecting subcutaneously a mixture of Carbon tetrachloride and olive oil. A total of 99 male Wistar rats were successfully induced and underwent MR scanning with no contrast agent to get T1-weighted images. The regions of interest (ROIs) of the whole liver were delineated layer by layer along the liver edge by 3D Slicer. For segmentation task, all T1-weighted images were randomly divided into training and test cohorts in a ratio of 7:3. For classification, images containing the hepatic maximum diameter of every rat were selected and 80% images of no liver fibrosis (NLF), early liver fibrosis (ELF) and progressive liver fibrosis (PLF) stages were randomly selected for training, while the rest were used for testing. Liver segmentation was performed by the nnU-Net model. The convolutional neural network (CNN) was used for classification task of liver fibrosis stages. The Dice similarity coefficient was used to evaluate the segmentation performance of nnU-Net. Confusion matrix, ROC curve and accuracy were used to show the classification performance of CNN.

RESULTS

A total of 2628 images were obtained from 99 Wistar rats by MR scanning. For liver segmentation by nnU-Net, the Dice similarity coefficient in the test set was 0.8477. The accuracies of CNN in staging NLF, ELF and PLF were 0.73, 0.89 and 0.84, respectively. The AUCs were 0.76, 0.88 and 0.79, respectively.

CONCLUSION

The nnU-Net architecture is of high accuracy for liver segmentation and CNN for assessment of liver fibrosis with T1-weighted images.

CSCA U-Net: A channel and space compound attention CNN for medical image segmentation

Image segmentation is one of the vital steps in medical image analysis. A large number of methods based on convolutional neural networks have emerged, which can extract abstract features from multiple-modality medical images, learn valuable information that is difficult to recognize by humans, and obtain more reliable results than traditional image segmentation approaches. U-Net, due to its simple structure and excellent performance, is widely used in medical image segmentation. In this paper, to further improve the performance of U-Net, we propose a channel and space compound attention (CSCA) convolutional neural network, CSCA U-Net in abbreviation, which increases the network depth and employs a double squeeze-and-excitation (DSE) block in the bottleneck layer to enhance feature extraction and obtain more high-level semantic features. Moreover, the characteristics of the proposed method are three-fold: (1) channel and space compound attention (CSCA) block, (2) cross-layer feature fusion (CLFF), and (3) deep supervision (DS). Extensive experiments on several available medical image datasets, including Kvasir-SEG, CVC-ClinicDB, CVC-ColonDB, ETIS, CVC-T, 2018 Data Science Bowl (2018 DSB), ISIC 2018, and JSUAH-Cerebellum, show that CSCA U-Net achieves competitive results and significantly improves generalization performance. The codes and trained models are available at https://github.com/xiaolanshu/CSCA-U-Net.

HmsU-Net: A hybrid multi-scale U-net based on a CNN and transformer for medical image segmentation

Accurate medical image segmentation is of great significance for subsequent diagnosis and analysis. The acquisition of multi-scale information plays an important role in segmenting regions of interest of different sizes. With the emergence of Transformers, numerous networks adopted hybrid structures incorporating Transformers and CNNs to learn multi-scale information. However, the majority of research has focused on the design and composition of CNN and Transformer structures, neglecting the inconsistencies in feature learning between Transformer and CNN. This oversight has resulted in the hybrid network's performance not being fully realized. In this work, we proposed a novel hybrid multi-scale segmentation network named HmsU-Net, which effectively fused multi-scale features. Specifically, HmsU-Net employed a parallel design incorporating both CNN and Transformer architectures. To address the inconsistency in feature learning between CNN and Transformer within the same stage, we proposed the multi-scale feature fusion module. For feature fusion across different stages, we introduced the cross-attention module. Comprehensive experiments conducted on various datasets demonstrate that our approach surpasses current state-of-the-art methods.

Comparison between U-shaped structural deep learning models to detect landslide traces

Landslides endanger lives and public infrastructure in mountainous areas. Monitoring landslide traces in real-time is difficult for scientists, sometimes costly and risky because of the harsh terrain and instability. Nowadays, modern technology may be able to identify landslide-prone locations and inform locals for hours or days when the weather worsens. This study aims to propose indicators to detect landslide traces on the fields and remote sensing images; build deep learning (DL) models to identify landslides from Sentinel-2 images automatically; and apply DL-trained models to detect this natural hazard in some particular areas of Vietnam. Nine DL models were trained based on three U-shaped architectures, including U-Net, U2-Net, and U-Net3+, and three options of input sizes. The multi-temporal Sentinel-2 images were chosen as input data for training all models. As a result, the U-Net, using an input image size of 32 × 32 and a performance of 97 % with a loss function of 0.01, can detect typical landslide traces in Vietnam. Meanwhile, the U-Net (64 × 64) can detect more considerable landslide traces. Based on multi-temporal remote sensing data, a different case study in Vietnam was chosen to see landslide traces over time based on the trained U-Net (32 × 32) model. The trained model allows mountain managers to track landslide occurrences during wet seasons. Thus, landslide incidents distant from residential areas may be discovered early to warn of flash floods.

Multimodal Biomedical Image Segmentation using Multi-Dimensional U-Convolutional Neural Network

Deep learning recently achieved advancement in the segmentation of medical images. In this regard, U-Net is the most predominant deep neural network, and its architecture is the most prevalent in the medical imaging society. Experiments conducted on difficult datasets directed us to the conclusion that the traditional U-Net framework appears to be deficient in certain respects, despite its overall excellence in segmenting multimodal medical images. Therefore, we propose several modifications to the existing cutting-edge U-Net model. The technical approach involves applying a Multi-Dimensional U-Convolutional Neural Network to achieve accurate segmentation of multimodal biomedical images, enhancing precision and comprehensiveness in identifying and analyzing structures across diverse imaging modalities. As a result of the enhancements, we propose a novel framework called Multi-Dimensional U-Convolutional Neural Network (MDU-CNN) as a potential successor to the U-Net framework. On a large set of multimodal medical images, we compared our proposed framework, MDU-CNN, to the classical U-Net. There have been small changes in the case of perfect images, and a huge improvement is obtained in the case of difficult images. We tested our model on five distinct datasets, each of which presented unique challenges, and found that it has obtained a better performance of 1.32%, 5.19%, 4.50%, 10.23% and 0.87%, respectively.

An improved supervised and attention mechanism-based U-Net algorithm for retinal vessel segmentation

The segmentation results of retinal blood vessels are crucial for automatically diagnosing ophthalmic diseases such as diabetic retinopathy, hypertension, cardiovascular and cerebrovascular diseases. To improve the accuracy of vessel segmentation and better extract information about small vessels and edges, we introduce the U-Net algorithm with a supervised attention mechanism for retinal vessel segmentation. We achieve this by introducing a decoder fusion module (DFM) in the encoding part, effectively combining different convolutional blocks to extract features comprehensively. Additionally, in the decoding part of U-Net, we propose the context squeeze and excitation (CSE) decoding module to enhance important contextual feature information and the detection of tiny blood vessels. For the final output, we introduce the supervised fusion mechanism (SFM), which combines multiple branches from shallow to deep layers, effectively fusing multi-scale features and capturing information from different levels, fully integrating low-level and high-level features to improve segmentation performance. Our experimental results on the public datasets of DRIVE, STARE, and CHASED_B1 demonstrate the excellent performance of our proposed network.

Clinical feasibility of MRI-based synthetic CT imaging in the diagnosis of lumbar disc herniation: a comparative study

BACKGROUND

Computed tomography (CT) and magnetic resonance imaging (MRI) are indicated for use in preoperative planning and may complicate diagnosis and place a burden on patients with lumbar disc herniation.

PURPOSE

To investigate the diagnostic potential of MRI-based synthetic CT with conventional CT in the diagnosis of lumbar disc herniation.

MATERIAL AND METHODS

After obtaining prior institutional review board approval, 19 patients who underwent conventional and synthetic CT imaging were enrolled in this prospective study. Synthetic CT images were generated from the MRI data using U-net. The two sets of images were compared and analyzed qualitatively by two musculoskeletal radiologists. The images were rated on a 4-point scale to determine their subjective quality. The agreement between the conventional and synthetic images for a diagnosis of lumbar disc herniation was determined independently using the kappa statistic. The diagnostic performances of conventional and synthetic CT images were evaluated for sensitivity, specificity, and accuracy, and the consensual results based on T2-weighted imaging were employed as the reference standard.

RESULTS

The inter-reader and intra-reader agreement were almost moderate for all evaluated modalities (κ = 0.57-0.79 and 0.47-0.75, respectively). The sensitivity, specificity, and accuracy for detecting lumbar disc herniation were similar for synthetic and conventional CT images (synthetic vs. conventional, reader 1: sensitivity = 91% vs. 81%, specificity = 83% vs. 100%, accuracy = 87% vs. 91%; P < 0.001; reader 2: sensitivity = 84% vs. 81%, specificity = 85% vs. 98%, accuracy = 84% vs. 90%; P < 0.001).

CONCLUSION

Synthetic CT images can be used in the diagnostics of lumbar disc herniation.

An open-source nnU-net algorithm for automatic segmentation of MRI scans in the male pelvis for adaptive radiotherapy

Background

Adaptive MRI-guided radiotherapy (MRIgRT) requires accurate and efficient segmentation of organs and targets on MRI scans. Manual segmentation is time-consuming and variable, while deformable image registration (DIR)-based contour propagation may not account for large anatomical changes. Therefore, we developed and evaluated an automatic segmentation method using the nnU-net framework.

Methods

The network was trained on 38 patients (76 scans) with localized prostate cancer and tested on 30 patients (60 scans) with localized prostate, metastatic prostate, or bladder cancer treated at a 1.5 T MRI-linac at our institution. The performance of the network was compared with the current clinical workflow based on DIR. The segmentation accuracy was evaluated using the Dice similarity coefficient (DSC), mean surface distance (MSD), and Hausdorff distance (HD) metrics.

Results

The trained network successfully segmented all 600 structures in the test set. High similarity was obtained for most structures, with 90% of the contours having a DSC above 0.9 and 86% having an MSD below 1 mm. The largest discrepancies were found in the sigmoid and colon structures. Stratified analysis on cancer type showed that the best performance was seen in the same type of patients that the model was trained on (localized prostate). Especially in patients with bladder cancer, the performance was lower for the bladder and the surrounding organs. A complete automatic delineation workflow took approximately 1 minute. Compared with contour transfer based on the clinically used DIR algorithm, the nnU-net performed statistically better across all organs, with the most significant gain in using the nnU-net seen for organs subject to more considerable volumetric changes due to variation in the filling of the rectum, bladder, bowel, and sigmoid.

Conclusion

We successfully trained and tested a network for automatically segmenting organs and targets for MRIgRT in the male pelvis region. Good test results were seen for the trained nnU-net, with test results outperforming the current clinical practice using DIR-based contour propagation at the 1.5 T MRI-linac. The trained network is sufficiently fast and accurate for clinical use in an online setting for MRIgRT. The model is provided as open-source.

CCRA: A colon cleanliness rating algorithm based on colonoscopy video analysis

Objective

A Colon Cleanliness Rating Algorithm (CCRA) based on colonoscopy image analysis is proposed in this paper, in order to solve the problem that the results of Colon Cleanliness (or Bowel Preparation Quality) rating caused by manual inspection are inconsistent.

Methods

Firstly, CCRA intercepts images from the colonoscopy video. Secondly, each colonoscopy image's stool area is segmented by U-Net to obtain the 2-classification segmentation results. Finally, the colon cleanliness is obtained by comparing the average area of the stool area with the standard proportion.

Results

After testing, the pixel accuracy of the U-Net model is 97.02 %, IoU is 83.67 %, accuracy is 92.17 %, recall is 90.21 %, F1-Score is 90.95 %. The accuracy of CCRA is 92.45 %-99.275.

Conclusion

The experimental results show that the CCRA proposed in this paper can quickly and accurately output the colon cleanliness rating of patients without manpower.

Utilizing historical maps in identification of long-term land use and land cover changes

Knowledge in the magnitude and historical trends in land use and land cover (LULC) is needed to understand the changing status of the key elements of the landscape and to better target management efforts. However, this information is not easily available before the start of satellite campaign missions. Scanned historical maps are a valuable but underused source of LULC information. As a case study, we used U-Net to automatically extract fields, mires, roads, watercourses, and water bodies from scanned historical maps, dated 1965, 1984 and 1985 for our 900 km[Formula: see text] study area in Southern Finland. We then used these data, along with the topographic databases from 2005 and 2022, to quantify the LULC changes for the past 57 years. For example, the total area of fields decreased by around 27 km[Formula: see text], and the total length of watercourses increased by around 2250 km in our study area.

Identifying core MRI sequences for reliable automatic brain metastasis segmentation

BACKGROUND

Many automatic approaches to brain tumor segmentation employ multiple magnetic resonance imaging (MRI) sequences. The goal of this project was to compare different combinations of input sequences to determine which MRI sequences are needed for effective automated brain metastasis (BM) segmentation.

METHODS

We analyzed preoperative imaging (T1-weighted sequence ± contrast-enhancement (T1/T1-CE), T2-weighted sequence (T2), and T2 fluid-attenuated inversion recovery (T2-FLAIR) sequence) from 339 patients with BMs from seven centers. A baseline 3D U-Net with all four sequences and six U-Nets with plausible sequence combinations (T1-CE, T1, T2-FLAIR, T1-CE + T2-FLAIR, T1-CE + T1 + T2-FLAIR, T1-CE + T1) were trained on 239 patients from two centers and subsequently tested on an external cohort of 100 patients from five centers.

RESULTS

The model based on T1-CE alone achieved the best segmentation performance for BM segmentation with a median Dice similarity coefficient (DSC) of 0.96. Models trained without T1-CE performed worse (T1-only: DSC = 0.70 and T2-FLAIR-only: DSC = 0.73). For edema segmentation, models that included both T1-CE and T2-FLAIR performed best (DSC = 0.93), while the remaining four models without simultaneous inclusion of these both sequences reached a median DSC of 0.81-0.89.

CONCLUSIONS

A T1-CE-only protocol suffices for the segmentation of BMs. The combination of T1-CE and T2-FLAIR is important for edema segmentation. Missing either T1-CE or T2-FLAIR decreases performance. These findings may improve imaging routines by omitting unnecessary sequences, thus allowing for faster procedures in daily clinical practice while enabling optimal neural network-based target definitions.

Combining neural networks and image synthesis to enable automatic thoracic cavity segmentation of hyperpolarized 129Xe MRI without proton scans

RATIONALE AND OBJECTIVES

Quantification of 129Xe MRI relies on accurate segmentation of the thoracic cavity, typically performed manually using a combination of 1H and 129Xe scans. This can be accelerated by using Convolutional Neural Networks (CNNs) that segment only the 129Xe scan. However, this task is complicated by peripheral ventilation defects, which requires training CNNs with large, diverse datasets. Here, we accelerate the creation of training data by synthesizing 129Xe images with a variety of defects. We use this to train a 3D model to provide thoracic cavity segmentation from 129Xe ventilation MRI alone.

MATERIALS AND METHODS

Training and testing data consisted of 22 and 33 3D 129Xe ventilation images. Training data were expanded to 484 using Template-based augmentation while an additional 298 images were synthesized using the Pix2Pix model. This data was used to train both a 2D U-net and 3D V-net-based segmentation model using a combination of Dice-Focal and Anatomical Constraint loss functions. Segmentation performance was compared using Dice coefficients calculated over the entire lung and within ventilation defects.

RESULTS

Performance of both U-net and 3D segmentation was improved by including synthetic training data. The 3D models performed significantly better than U-net, and the 3D model trained with synthetic 129Xe images exhibited the highest overall Dice score of 0.929. Moreover, addition of synthetic training data improved the Dice score in ventilation defect regions from 0.545 to 0.588 for U-net and 0.739 to 0.765 for the 3D model.

CONCLUSION

It is feasible to obtain high-quality segmentations from 129Xe scan alone using 3D models trained with additional synthetic images.

Smart Parking Locks Based on Extended UNET-GWO-SVM Algorithm

Due to the rapid increase in private car ownership in China, most cities face the problem of insufficient parking spaces, leading to frequent occurrences of parking space conflicts. There is a wide variety of parking locks available on the market. However, most of them lack advanced intelligence and cannot cater to the growing diverse needs of people. The present study attempts to devise a smart parking lock to tackle this issue. Specifically, the smart parking lock uses a Raspberry Pi as the core controller, senses the vehicle with an ultrasonic ranging module, and collects the license plate image with a camera. In addition, algorithms for license plate recognition based on traditional image-processing methods typically require a high pixel resolution, but their recognition accuracy is often low. Therefore, we propose a new algorithm called UNET-GWO-SVM to achieve higher accuracy in embedded systems. Moreover, we developed a WeChat mini program to control the smart parking lock. Field tests were conducted on campus to evaluate the performance of the parking locks. The test results show that the corresponding effective unlocking rate is 99.0% when the recognition error is less than two license plate characters. The average time consumption is controlled at about 2 s. It can meet real-time requirements.

Utility of Artificial Intelligence for Real-Time Anatomical Landmark Identification in Ultrasound-Guided Thoracic Paravertebral Block

Thoracic paravertebral block (TPVB) is a common method of inducing perioperative analgesia in thoracic and abdominal surgery. Identifying anatomical structures in ultrasound images is very important especially for inexperienced anesthesiologists who are unfamiliar with the anatomy. Therefore, our aim was to develop an artificial neural network (ANN) to automatically identify (in real-time) anatomical structures in ultrasound images of TPVB. This study is a retrospective study using ultrasound scans (both video and standard still images) that we acquired. We marked the contours of the paravertebral space (PVS), lung, and bone in the TPVB ultrasound image. Based on the labeled ultrasound images, we used the U-net framework to train and create an ANN that enabled real-time identification of important anatomical structures in ultrasound images. A total of 742 ultrasound images were acquired and labeled in this study. In this ANN, the Intersection over Union (IoU) and Dice similarity coefficient (DSC or Dice coefficient) of the paravertebral space (PVS) were 0.75 and 0.86, respectively, the IoU and DSC of the lung were 0.85 and 0.92, respectively, and the IoU and DSC of the bone were 0.69 and 0.83, respectively. The accuracies of the PVS, lung, and bone were 91.7%, 95.4%, and 74.3%, respectively. For tenfold cross validation, the median interquartile range for PVS IoU and DSC was 0.773 and 0.87, respectively. There was no significant difference in the scores for the PVS, lung, and bone between the two anesthesiologists. We developed an ANN for the real-time automatic identification of thoracic paravertebral anatomy. The performance of the ANN was highly satisfactory. We conclude that AI has good prospects for use in TPVB. Clinical registration number: ChiCTR2200058470 (URL: http://www.chictr.org.cn/showproj.aspx?proj=152839 ; registration date: 2022-04-09).

Establishing a machine learning model for predicting nutritional risk through facial feature recognition

Background

Malnutrition affects many worldwide, necessitating accurate and timely nutritional risk assessment. This study aims to develop and validate a machine learning model using facial feature recognition for predicting nutritional risk. This innovative approach seeks to offer a non-invasive, efficient method for early identification and intervention, ultimately improving health outcomes.

Methods

We gathered medical examination data and facial images from 949 patients across multiple hospitals to predict nutritional status. In this multicenter investigation, facial images underwent preprocessing via face alignment and cropping. Orbital fat pads were isolated using the U-net model, with the histogram of oriented gradient (HOG) method employed for feature extraction. Standardized HOG features were subjected to principal component analysis (PCA) for dimensionality reduction. A support vector machine (SVM) classification model was utilized for NRS-2002 detection. Our approach established a non-linear mapping between facial features and NRS-2002 nutritional risk scores, providing an innovative method for evaluating patient nutritional status.

Results

In context of orbital fat pad area segmentation with U-net model, the averaged dice coefficient is 88.3%. Our experimental results show that the proposed method to predict NRS-2002 scores achieves an accuracy of 73.1%. We also grouped the samples by gender, age, and the location of the hospital where the data were collected to evaluate the classification accuracy in different subsets. The classification accuracy rate for the elderly group was 85%, while the non-elderly group exhibited a classification accuracy rate of 71.1%; Furthermore, the classification accuracy rate for males and females were 69.2 and 78.6%, respectively. Hospitals located in remote areas, such as Tibet and Yunnan, yielded a classification accuracy rate of 76.5% for collected patient samples, whereas hospitals in non-remote areas achieved a classification accuracy rate of 71.1%.

Conclusion

The attained accuracy rate of 73.1% holds significant implications for the feasibility of the method. While not impeccable, this level of accuracy highlights the potential for further improvements. The development of this algorithm has the potential to revolutionize nutritional risk assessment by providing healthcare professionals and individuals with a non-invasive, cost-effective, and easily accessible tool.

CS-MRI Reconstruction Using an Improved GAN with Dilated Residual Networks and Channel Attention Mechanism

Compressed sensing (CS) MRI has shown great potential in enhancing time efficiency. Deep learning techniques, specifically generative adversarial networks (GANs), have emerged as potent tools for speedy CS-MRI reconstruction. Yet, as the complexity of deep learning reconstruction models increases, this can lead to prolonged reconstruction time and challenges in achieving convergence. In this study, we present a novel GAN-based model that delivers superior performance without the model complexity escalating. Our generator module, built on the U-net architecture, incorporates dilated residual (DR) networks, thus expanding the network's receptive field without increasing parameters or computational load. At every step of the downsampling path, this revamped generator module includes a DR network, with the dilation rates adjusted according to the depth of the network layer. Moreover, we have introduced a channel attention mechanism (CAM) to distinguish between channels and reduce background noise, thereby focusing on key information. This mechanism adeptly combines global maximum and average pooling approaches to refine channel attention. We conducted comprehensive experiments with the designed model using public domain MRI datasets of the human brain. Ablation studies affirmed the efficacy of the modified modules within the network. Incorporating DR networks and CAM elevated the peak signal-to-noise ratios (PSNR) of the reconstructed images by about 1.2 and 0.8 dB, respectively, on average, even at 10× CS acceleration. Compared to other relevant models, our proposed model exhibits exceptional performance, achieving not only excellent stability but also outperforming most of the compared networks in terms of PSNR and SSIM. When compared with U-net, DR-CAM-GAN's average gains in SSIM and PSNR were 14% and 15%, respectively. Its MSE was reduced by a factor that ranged from two to seven. The model presents a promising pathway for enhancing the efficiency and quality of CS-MRI reconstruction.

MRI-Based Deep Learning Method for Classification of IDH Mutation Status

Isocitrate dehydrogenase (IDH) mutation status has emerged as an important prognostic marker in gliomas. This study sought to develop deep learning networks for non-invasive IDH classification using T2w MR images while comparing their performance to a multi-contrast network. Methods: Multi-contrast brain tumor MRI and genomic data were obtained from The Cancer Imaging Archive (TCIA) and The Erasmus Glioma Database (EGD). Two separate 2D networks were developed using nnU-Net, a T2w-image-only network (T2-net) and a multi-contrast network (MC-net). Each network was separately trained using TCIA (227 subjects) or TCIA + EGD data (683 subjects combined). The networks were trained to classify IDH mutation status and implement single-label tumor segmentation simultaneously. The trained networks were tested on over 1100 held-out datasets including 360 cases from UT Southwestern Medical Center, 136 cases from New York University, 175 cases from the University of Wisconsin-Madison, 456 cases from EGD (for the TCIA-trained network), and 495 cases from the University of California, San Francisco public database. A receiver operating characteristic curve (ROC) was drawn to calculate the AUC value to determine classifier performance. Results: T2-net trained on TCIA and TCIA + EGD datasets achieved an overall accuracy of 85.4% and 87.6% with AUCs of 0.86 and 0.89, respectively. MC-net trained on TCIA and TCIA + EGD datasets achieved an overall accuracy of 91.0% and 92.8% with AUCs of 0.94 and 0.96, respectively. We developed reliable, high-performing deep learning algorithms for IDH classification using both a T2-image-only and a multi-contrast approach. The networks were tested on more than 1100 subjects from diverse databases, making this the largest study on image-based IDH classification to date.

Automated Stabilization, Enhancement and Capillaries Segmentation in Videocapillaroscopy

Oral capillaroscopy is a critical and non-invasive technique used to evaluate microcirculation. Its ability to observe small vessels in vivo has generated significant interest in the field. Capillaroscopy serves as an essential tool for diagnosing and prognosing various pathologies, with anatomic-pathological lesions playing a crucial role in their progression. Despite its importance, the utilization of videocapillaroscopy in the oral cavity encounters limitations due to the acquisition setup, encompassing spatial and temporal resolutions of the video camera, objective magnification, and physical probe dimensions. Moreover, the operator's influence during the acquisition process, particularly how the probe is maneuvered, further affects its effectiveness. This study aims to address these challenges and improve data reliability by developing a computerized support system for microcirculation analysis. The designed system performs stabilization, enhancement and automatic segmentation of capillaries in oral mucosal video sequences. The stabilization phase was performed by means of a method based on the coupling of seed points in a classification process. The enhancement process implemented was based on the temporal analysis of the capillaroscopic frames. Finally, an automatic segmentation phase of the capillaries was implemented with the additional objective of quantitatively assessing the signal improvement achieved through the developed techniques. Specifically, transfer learning of the renowned U-net deep network was implemented for this purpose. The proposed method underwent testing on a database with ground truth obtained from expert manual segmentation. The obtained results demonstrate an achieved Jaccard index of 90.1% and an accuracy of 96.2%, highlighting the effectiveness of the developed techniques in oral capillaroscopy. In conclusion, these promising outcomes encourage the utilization of this method to assist in the diagnosis and monitoring of conditions that impact microcirculation, such as rheumatologic or cardiovascular disorders.

C2FTFNet: Coarse-to-fine transformer network for joint optic disc and cup segmentation

Glaucoma is a leading cause of worldwide blindness and visual impairment, making early screening and diagnosis is crucial to prevent vision loss. Cup-to-Disk Ratio (CDR) evaluation serves as a widely applied approach for effective glaucoma screening. At present, deep learning methods have exhibited outstanding performance in optic disk (OD) and optic cup (OC) segmentation and maturely deployed in CAD system. However, owning to the complexity of clinical data, these techniques could be constrained. Therefore, an original Coarse-to-Fine Transformer Network (C2FTFNet) is designed to segment OD and OC jointly , which is composed of two stages. In the coarse stage, to eliminate the effects of irrelevant organization on the segmented OC and OD regions, we employ U-Net and Circular Hough Transform (CHT) to segment the Region of Interest (ROI) of OD. Meanwhile, a TransUnet3+ model is designed in the fine segmentation stage to extract the OC and OD regions more accurately from ROI. In this model, to alleviate the limitation of the receptive field caused by traditional convolutional methods, a Transformer module is introduced into the backbone to capture long-distance dependent features for retaining more global information. Then, a Multi-Scale Dense Skip Connection (MSDC) module is proposed to fuse the low-level and high-level features from different layers for reducing the semantic gap among different level features. Comprehensive experiments conducted on DRIONS-DB, Drishti-GS, and REFUGE datasets validate the superior effectiveness of the proposed C2FTFNet compared to existing state-of-the-art approaches.

Deep-MSIM: Fast Image Reconstruction with Deep Learning in Multifocal Structured Illumination Microscopy

Fast and precise reconstruction algorithm is desired for for multifocal structured illumination microscopy (MSIM) to obtain the super-resolution image. This work proposes a deep convolutional neural network (CNN) to learn a direct mapping from raw MSIM images to super-resolution image, which takes advantage of the computational advances of deep learning to accelerate the reconstruction. The method is validated on diverse biological structures and in vivo imaging of zebrafish at a depth of 100 µm. The results show that high-quality, super-resolution images can be reconstructed in one-third of the runtime consumed by conventional MSIM method, without compromising spatial resolution. Last but not least, a fourfold reduction in the number of raw images required for reconstruction is achieved by using the same network architecture, yet with different training data.

Artificial Intelligence framework with traditional computer vision and deep learning approaches for optimal automatic segmentation of left ventricle with scar

Automatic segmentation of the cardiac left ventricle with scars remains a challenging and clinically significant task, as it is essential for patient diagnosis and treatment pathways. This study aimed to develop a novel framework and cost function to achieve optimal automatic segmentation of the left ventricle with scars using LGE-MRI images. To ensure the generalization of the framework, an unbiased validation protocol was established using out-of-distribution (OOD) internal and external validation cohorts, and intra-observation and inter-observer variability ground truths. The framework employs a combination of traditional computer vision techniques and deep learning, to achieve optimal segmentation results. The traditional approach uses multi-atlas techniques, active contours, and k-means methods, while the deep learning approach utilizes various deep learning techniques and networks. The study found that the traditional computer vision technique delivered more accurate results than deep learning, except in cases where there was breath misalignment error. The optimal solution of the framework achieved robust and generalized results with Dice scores of 82.8 ± 6.4% and 72.1 ± 4.6% in the internal and external OOD cohorts, respectively. The developed framework offers a high-performance solution for automatic segmentation of the left ventricle with scars using LGE-MRI. Unlike existing state-of-the-art approaches, it achieves unbiased results across different hospitals and vendors without the need for training or tuning in hospital cohorts. This framework offers a valuable tool for experts to accomplish the task of fully automatic segmentation of the left ventricle with scars based on a single-modality cardiac scan.

An efficient convolutional neural network-based diagnosis system for citrus fruit diseases

Introduction: Fruit diseases have a serious impact on fruit production, causing a significant drop in economic returns from agricultural products. Due to its excellent performance, deep learning is widely used for disease identification and severity diagnosis of crops. This paper focuses on leveraging the high-latitude feature extraction capability of deep convolutional neural networks to improve classification performance. Methods: The proposed neural network is formed by combining the Inception module with the current state-of-the-art EfficientNetV2 for better multi-scale feature extraction and disease identification of citrus fruits. The VGG is used to replace the U-Net backbone to enhance the segmentation performance of the network. Results: Compared to existing networks, the proposed method achieved recognition accuracy of over 95%. In addition, the accuracies of the segmentation models were compared. VGG-U-Net, a network generated by replacing the backbone of U-Net with VGG, is found to have the best segmentation performance with an accuracy of 87.66%. This method is most suitable for diagnosing the severity level of citrus fruit diseases. In the meantime, transfer learning is applied to improve the training cycle of the network model, both in the detection and severity diagnosis phases of the disease. Discussion: The results of the comparison experiments reveal that the proposed method is effective in identifying and diagnosing the severity of citrus fruit diseases identification.

A new dental CBCT metal artifact reduction method based on a dual-domain processing framework

Objective.Cone beam computed tomography (CBCT) has been wildly used in clinical treatment of dental diseases. However, patients often have metallic implants in mouth, which will lead to severe metal artifacts in the reconstructed images. To reduce metal artifacts in dental CBCT images, which have a larger amount of data and a limited field of view compared to computed tomography images, a new dental CBCT metal artifact reduction method based on a projection correction and a convolutional neural network (CNN) based image post-processing model is proposed in this paper. Approach.The proposed method consists of three stages: (1) volume reconstruction and metal segmentation in the image domain, using the forward projection to get the metal masks in the projection domain; (2) linear interpolation in the projection domain and reconstruction to build a linear interpolation (LI) corrected volume; (3) take the LI corrected volume as prior and perform the prior based beam hardening correction in the projection domain, and (4) combine the constructed projection corrected volume and LI-volume slice-by-slice in the image domain by two concatenated U-Net based models (CNN1 and CNN2). Simulated and clinical dental CBCT cases are used to evaluate the proposed method. The normalized root means square difference (NRMSD) and the structural similarity index (SSIM) are used for the quantitative evaluation of the method.Main results.The proposed method outperforms the frequency domain fusion method (FS-MAR) and a state-of-art CNN based method on the simulated dataset and yields the best NRMSD and SSIM of 4.0196 and 0.9924, respectively. Visual results on both simulated and clinical images also illustrate that the proposed method can effectively reduce metal artifacts.Significance. This study demonstrated that the proposed dual-domain processing framework is suitable for metal artifact reduction in dental CBCT images.

Deep learning for the rapid automatic segmentation of forearm muscle boundaries from ultrasound datasets

Ultrasound (US) is widely used in the clinical diagnosis and treatment of musculoskeletal diseases. However, the low efficiency and non-uniformity of artificial recognition hinder the application and popularization of US for this purpose. Herein, we developed an automatic muscle boundary segmentation tool for US image recognition and tested its accuracy and clinical applicability. Our dataset was constructed from a total of 465 US images of the flexor digitorum superficialis (FDS) from 19 participants (10 men and 9 women, age 27.4 ± 6.3 years). We used the U-net model for US image segmentation. The U-net output often includes several disconnected regions. Anatomically, the target muscle usually only has one connected region. Based on this principle, we designed an algorithm written in C++ to eliminate redundantly connected regions of outputs. The muscle boundary images generated by the tool were compared with those obtained by professionals and junior physicians to analyze their accuracy and clinical applicability. The dataset was divided into five groups for experimentation, and the average Dice coefficient, recall, and accuracy, as well as the intersection over union (IoU) of the prediction set in each group were all about 90%. Furthermore, we propose a new standard to judge the segmentation results. Under this standard, 99% of the total 150 predicted images by U-net are excellent, which is very close to the segmentation result obtained by professional doctors. In this study, we developed an automatic muscle segmentation tool for US-guided muscle injections. The accuracy of the recognition of the muscle boundary was similar to that of manual labeling by a specialist sonographer, providing a reliable auxiliary tool for clinicians to shorten the US learning cycle, reduce the clinical workload, and improve injection safety.

Funcmasker-flex: An Automated BIDS-App for Brain Segmentation of Human Fetal Functional MRI data

Fetal functional magnetic resonance imaging (fMRI) offers critical insight into the developing brain and could aid in predicting developmental outcomes. As the fetal brain is surrounded by heterogeneous tissue, it is not possible to use adult- or child-based segmentation toolboxes. Manually-segmented masks can be used to extract the fetal brain; however, this comes at significant time costs. Here, we present a new BIDS App for masking fetal fMRI, funcmasker-flex, that overcomes these issues with a robust 3D convolutional neural network (U-net) architecture implemented in an extensible and transparent Snakemake workflow. Open-access fetal fMRI data with manual brain masks from 159 fetuses (1103 total volumes) were used for training and testing the U-net model. We also tested generalizability of the model using 82 locally acquired functional scans from 19 fetuses, which included over 2300 manually segmented volumes. Dice metrics were used to compare performance of funcmasker-flex to the ground truth manually segmented volumes, and segmentations were consistently robust (all Dice metrics ≥ 0.74). The tool is freely available and can be applied to any BIDS dataset containing fetal bold sequences. Funcmasker-flex reduces the need for manual segmentation, even when applied to novel fetal functional datasets, resulting in significant time-cost savings for performing fetal fMRI analysis.

Suitability evaluation and potential estimation of photovoltaic power generation and carbon emission reduction in the Qinghai-Tibet Plateau

The expansion of power development industry is facing enormous pressure to reduce carbon emissions in the context of global decarbonization. Using solar energy instead of traditional fossil energy to adjust energy structure is one of the important means for reducing carbon emissions. Existing research focuses on the evaluation of the generation potential of centralized or distributed photovoltaic power plants, rather than the comprehensive evaluation of multi-type power plants. Based on multi-source remote sensing data for information extraction and suitability evaluation, this paper develops a method to comprehensively evaluate the construction potential of multi-type photovoltaic power stations and determine the potential of photovoltaic power generation and carbon emission reduction on the Qinghai-Tibet Plateau (QTP). The results showed that estimating the power generation potential of only single-type photovoltaic power stations cannot accurately reflect the photovoltaic power generation potential of QTP. It is also demonstrated that the emission reduction effect of the photovoltaic power generation in all prefecture-level cities of QTP can meet national emission reduction targets, showing high annual power generation potential, of which 86.59% is concentrated in Qinghai province's Guoluo, Yushu, and Haixi. An accurate estimation of the photovoltaic power generation potential in QTP can provide a useful theoretical basis for developing carbon-saving and emission reduction strategies for clean energy in China.

Research of segmentation recognition of small disease spots on apple leaves based on hybrid loss function and CBAM

Identification technology of apple diseases is of great significance in improving production efficiency and quality. This paper has used apple Alternaria blotch and brown spot disease leaves as the research object and proposes a disease spot segmentation and disease identification method based on DFL-UNet+CBAM to address the problems of low recognition accuracy and poor performance of small spot segmentation in apple leaf disease recognition. The goal of this paper is to accurately prevent and control apple diseases, avoid fruit quality degradation and yield reduction, and reduce the resulting economic losses. DFL-UNet+CBAM model has employed a hybrid loss function of Dice Loss and Focal Loss as the loss function and added CBAM attention mechanism to both effective feature layers extracted by the backbone network and the results of the first upsampling, enhancing the model to rescale the inter-feature weighting relationships, enhance the channel features of leaf disease spots and suppressing the channel features of healthy parts of the leaf, and improving the network's ability to extract disease features while also increasing model robustness. In general, after training, the average loss rate of the improved model decreases from 0.063 to 0.008 under the premise of ensuring the accuracy of image segmentation. The smaller the loss value is, the better the model is. In the lesion segmentation and disease identification test, MIoU was 91.07%, MPA was 95.58%, F1 Score was 95.16%, MIoU index increased by 1.96%, predicted disease area and actual disease area overlap increased, MPA increased by 1.06%, predicted category correctness increased, F1 Score increased by 1.14%, the number of correctly identified lesion pixels increased, and the segmentation result was more accurate. Specifically, compared to the original U-Net model, the segmentation of Alternaria blotch disease, the MIoU value increased by 4.41%, the MPA value increased by 4.13%, the Precision increased by 1.49%, the Recall increased by 4.13%, and the F1 Score increased by 2.81%; in the segmentation of brown spots, MIoU values increased by 1.18%, MPA values by 0.6%, Precision by 0.78%, Recall by 0.6%, and F1 Score by 0.69%. The spot diameter of the Alternaria blotch disease is 0.2-0.3cm in the early stage, 0.5-0.6cm in the middle and late stages, and the spot diameter of the brown spot disease is 0.3-3cm. Obviously, brown spot spots are larger than Alternaria blotch spots. The segmentation performance of smaller disease spots has increased more noticeably, according to the quantitative analysis results, proving that the model's capacity to segment smaller disease spots has greatly improved. The findings demonstrate that for the detection of apple leaf diseases, the method suggested in this research has a greater recognition accuracy and better segmentation performance. The model in this paper can obtain more sophisticated semantic information in comparison to the traditional U-Net, further enhance the recognition accuracy and segmentation performance of apple leaf spots, and address the issues of low accuracy and low efficiency of conventional disease recognition methods as well as the challenging convergence of conventional deep convolutional networks.

Attention-guided cascaded network with pixel-importance-balance loss for retinal vessel segmentation

Accurate retinal vessel segmentation from fundus images is essential for eye disease diagnosis. Many deep learning methods have shown great performance in this task but still struggle with limited annotated data. To alleviate this issue, we propose an Attention-Guided Cascaded Network (AGC-Net) that learns more valuable vessel features from a few fundus images. Attention-guided cascaded network consists of two stages: the coarse stage produces a rough vessel prediction map from the fundus image, and the fine stage refines the missing vessel details from this map. In attention-guided cascaded network, we incorporate an inter-stage attention module (ISAM) to cascade the backbone of these two stages, which helps the fine stage focus on vessel regions for better refinement. We also propose Pixel-Importance-Balance Loss (PIB Loss) to train the model, which avoids gradient domination by non-vascular pixels during backpropagation. We evaluate our methods on two mainstream fundus image datasets (i.e., DRIVE and CHASE-DB1) and achieve AUCs of 0.9882 and 0.9914, respectively. Experimental results show that our method outperforms other state-of-the-art methods in performance.

Improved U-net-based leukocyte segmentation method

Significance

Leukocytes are mainly composed of neutrophils, basophils, eosinophils, monocytes, and lymphocytes. The number and proportion of different types of leukocytes correspond to different diseases, so an accurate segmentation of each type of leukocyte is important for the diagnosis of disease. However, the acquisition of blood cell images can be affected by external environmental factors, which can lead to variable light and darkness, complex backgrounds, and poorly characterized leukocytes.

Aim

To address the problem of complex blood cell images collected under different environments and the lack of obvious leukocyte features, a leukocyte segmentation method based on improved U-net is proposed.

Approach

First, adaptive histogram equalization-retinex correction was introduced for data enhancement to make the leukocyte features in the blood cell images clearer. Then, to address the problem of similarity between different types of leukocytes, convolutional block attention module is added to the four skip connections of U-net to focus the features from spatial and channel aspects, so that the network can quickly locate the high-value information of features in different channels and spaces. It avoids the problem of large amount of repeated computation of low-value information, prevents overfitting, and improves the training efficiency and generalization ability of the network. Finally, to solve the problem of class imbalance in blood cell images and to better segment the cytoplasm of leukocytes, a loss function combining focal loss and Dice loss is proposed.

Results

We use the BCISC public dataset to verify the effectiveness of the proposed method. The segmentation of multiple leukocytes using the method of this paper can achieve 99.53% accuracy and 91.89% mIoU.

Conclusions

The experimental results show that the method achieves good segmentation results for lymphocytes, basophils, neutrophils, eosinophils, and monocytes.

RSU-Net: U-net based on residual and self-attention mechanism in the segmentation of cardiac magnetic resonance images

BACKGROUND

Automated segmentation techniques for cardiac magnetic resonance imaging (MRI) are beneficial for evaluating cardiac functional parameters in clinical diagnosis. However, due to the characteristics of unclear image boundaries and anisotropic resolution anisotropy produced by cardiac magnetic resonance imaging technology, most of the existing methods still have the problems of intra-class uncertainty and inter-class uncertainty. However, due to the irregularity of the anatomical shape of the heart and the inhomogeneity of tissue density, the boundaries of its anatomical structures become uncertain and discontinuous. Therefore, fast and accurate segmentation of cardiac tissue remains a challenging problem in medical image processing.

METHODOLOGY

We collected cardiac MRI data from 195 patients as training set and 35patients from different medical centers as external validation set. Our research proposed a U-net network architecture with residual connections and a self-attentive mechanism (Residual Self-Attention U-net, RSU-Net). The network relies on the classic U-net network, adopts the U-shaped symmetric architecture of the encoding and decoding mode, improves the convolution module in the network, introduces skip connections, and improves the network's capacity for feature extraction. Then for solving locality defects of ordinary convolutional networks. To achieve a global receptive field, a self-attention mechanism is introduced at the bottom of the model. The loss function employs a combination of Cross Entropy Loss and Dice Loss to jointly guide network training, resulting in more stable network training.

RESULTS

In our study, we employ the Hausdorff distance (HD) and the Dice similarity coefficient (DSC) as metrics for assessing segmentation outcomes. Comparsion was made with the segmentation frameworks of other papers, and the comparison results prove that our RSU-Net network performs better and can make accurate segmentation of the heart. New ideas for scientific research.

CONCLUSION

Our proposed RSU-Net network combines the advantages of residual connections and self-attention. This paper uses the residual links to facilitate the training of the network. In this paper, a self-attention mechanism is introduced, and a bottom self-attention block (BSA Block) is used to aggregate global information. Self-attention aggregates global information, and has achieved good segmentation results on the cardiac segmentation dataset. It facilitates the diagnosis of cardiovascular patients in the future.

Hepatic vessels segmentation using deep learning and preprocessing enhancement

PURPOSE

Liver hepatic vessels segmentation is a crucial step for the diagnosis process in patients with hepatic diseases. Segmentation of liver vessels helps to study the liver internal segmental anatomy that helps in the preoperative planning of surgical treatment.

METHODS

Recently, the convolutional neural networks (CNN) have been proved to be efficient for the task of medical image segmentation. The paper proposes an automatic deep learning-based system for liver hepatic vessels segmentation of Computed Tomography (CT) datasets from different sources. The proposed work focuses on the combination of different steps; it starts by a preprocessing step to improve the vessels appearance within the liver region of interest in the CT scans. Coherence enhancing diffusion filtering (CED) and vesselness filtering methods are used to improve vessels contrast and intensity homogeneity. The proposed U-net based network architecture is implemented with modified residual block to include concatenation skip connection. The effect of enhancement using filtering step was studied. Also, the effect of data mismatch used in training and validation is studied.

RESULTS

The proposed method is evaluated using many CT datasets. Dice similarity coefficient (DSC) is used to evaluate the method. The average DSC score achieved a score 79%.

CONCLUSIONS

The proposed approach succeeded to segment liver vasculature from the liver envelope accurately, which makes it as potential tool for clinical preoperative planning.

Composite Attention Residual U-Net for Rib Fracture Detection

Computed tomography (CT) images play a vital role in diagnosing rib fractures and determining the severity of chest trauma. However, quickly and accurately identifying rib fractures in a large number of CT images is an arduous task for radiologists. We propose a U-net-based detection method designed to extract rib fracture features at the pixel level to find rib fractures rapidly and precisely. Two modules are applied to the segmentation network-a combined attention module (CAM) and a hybrid dense dilated convolution module (HDDC). The features of the same layer of the encoder and the decoder are fused through CAM, strengthening the local features of the subtle fracture area and increasing the edge features. HDDC is used between the encoder and decoder to obtain sufficient semantic information. Experiments show that on the public dataset, the model test brings the effects of Recall (81.71%), F1 (81.86%), and Dice (53.28%). Experienced radiologists reach lower false positives for each scan, whereas they have underperforming neural network models in terms of detection sensitivities with a long time diagnosis. With the aid of our model, radiologists can achieve higher detection sensitivities than computer-only or human-only diagnosis.

A convolutional neural network model for EPID-based non-transit dosimetry

PURPOSE

To develop an alternative computational approach for EPID-based non-transit dosimetry using a convolutional neural network model.

METHOD

A U-net followed by a non-trainable layer named True Dose Modulation recovering the spatialized information was developed. The model was trained on 186 Intensity-Modulated Radiation Therapy Step & Shot beams from 36 treatment plans of different tumor locations to convert grayscale portal images into planar absolute dose distributions. Input data were acquired from an amorphous-Silicon Electronic Portal Image Device and a 6 MV X-ray beam. Ground truths were computed from a conventional kernel-based dose algorithm. The model was trained by a two-step learning process and validated through a five-fold cross-validation procedure with sets of training and validation of 80% and 20%, respectively. A study regarding the dependance of the amount of training data was conducted. The performance of the model was evaluated from a quantitative analysis based the ϒ-index, absolute and relative errors computed between the inferred dose distributions and ground truths for six square and 29 clinical beams from seven treatment plans. These results were also compared to those of an existing portal image-to-dose conversion algorithm.

RESULTS

For the clinical beams, averages of ϒ-index and ϒ-passing rate (2%-2mm > 10% D_max ) of 0.24 (±0.04) and 99.29 (±0.70)% were obtained. For the same metrics and criteria, averages of 0.31 (±0.16) and 98.83 (±2.40)% were obtained with the six square beams. Overall, the developed model performed better than the existing analytical method. The study also showed that sufficient model accuracy can be achieved with the amount of training samples used.

CONCLUSION

A deep learning-based model was developed to convert portal images into absolute dose distributions. The accuracy obtained shows that this method has great potential for EPID-based non-transit dosimetry.

High precision tracking analysis of cell position and motion fields using 3D U-net network models

Cells are the basic units of biological organization, and the quantitative analysis of cellular states is an important topic in medicine and is valuable in revealing the complex mechanisms of microscopic world organisms. In order to better understand cell cycle changes as well as drug actions, we need to track cell migration and division. In this paper, we propose a novel engineering model for tracking cells using cell position and motion fields (CPMF). The training sample does not need to be manually annotated, and we modify and edit it against the ground truth using auxiliary tools. The core idea of the project is to combine detection and correlation, and the cell sequence samples are trained by a U-Net network model composed of 3D CNNs, which can track the migration, division, and entry and exit of cells in the field of view with high accuracy in all directions. The average detection accuracy of the cell coordinates is 98.38% and the average tracking accuracy is 98.70%.

Alternaria spore exposure in Bavaria, Germany, measured using artificial intelligence algorithms in a network of BAA500 automatic pollen monitors

Although Alternaria spores are well-known allergenic fungal spores, automatic bioaerosol recognition systems have not been trained to recognize these particles until now. Here we report the development of a new algorithm able to classify Alternaria spores with BAA500 automatic bioaerosol monitors. The best validation score was obtained when the model was trained on both data from the original dataset and artificially generated images, with a validation unweighted mean Intersection over Union (IoU), also called Jaccard Index, of 0.95. Data augmentation techniques were applied to the training set. While some particles were not recognized (false negatives), false positives were few. The results correlated well with manual counts (mean of four Hirst-type traps), with R² = 0.78. Counts from BAA500 were 1.92 times lower than with Hirst-type traps. The algorithm was then used to re-analyze the historical automatic pollen monitoring network (ePIN) dataset (2018-2022), which lacked Alternaria spore counts. Re-analysis of past data showed that Alternaria spore exposure in Bavaria was very variable, with the highest counts in the North (Marktheidenfeld, 154 m a.s.l.), and the lowest values close to the mountains in the South (Garmisch-Partenkirchen, 735 m a.s.l.). This approach shows that in our network future algorithms can be run on past datasets. Over time, the use of different algorithms could lead to misinterpretations as stemming from climate change or other phenological causes. Our approach enables consistent, homogeneous treatment of long-term series, thus preventing variability in particle counts owing to changes in the algorithms.

Deep-Learning-Based Automatic Segmentation of Parotid Gland on Computed Tomography Images

This study aims to develop an algorithm for the automatic segmentation of the parotid gland on CT images of the head and neck using U-Net architecture and to evaluate the model's performance. In this retrospective study, a total of 30 anonymized CT volumes of the head and neck were sliced into 931 axial images of the parotid glands. Ground truth labeling was performed with the CranioCatch Annotation Tool (CranioCatch, Eskisehir, Turkey) by two oral and maxillofacial radiologists. The images were resized to 512 × 512 and split into training (80%), validation (10%), and testing (10%) subgroups. A deep convolutional neural network model was developed using U-net architecture. The automatic segmentation performance was evaluated in terms of the F1-score, precision, sensitivity, and the Area Under Curve (AUC) statistics. The threshold for a successful segmentation was determined by the intersection of over 50% of the pixels with the ground truth. The F1-score, precision, and sensitivity of the AI model in segmenting the parotid glands in the axial CT slices were found to be 1. The AUC value was 0.96. This study has shown that it is possible to use AI models based on deep learning to automatically segment the parotid gland on axial CT images.

Human induced pluripotent stem cell formation and morphology prediction during reprogramming with time-lapse bright-field microscopy images using deep learning methods

BACKGROUND AND OBJECTIVE

Human induced pluripotent stem cells (hiPSCs) represent an ideal source for patient specific cell-based regenerative medicine; however, efficiency of hiPSC formation from reprogramming cells is low. We use several deep-learning results from time-lapse brightfield microscopy images during culture, to early detect the cells potentially reprogramming into hiPSCs and predict the colony morphology of these cells for improving efficiency of culturing a new hiPSC line.

METHODS

Sets of time-lapse bright-field images are taken to track reprogramming process of CD34+ cells biologically identified as just beginning reprogramming. Prior the experiment, 9 classes of templates with distinct cell features clipped from microscopy images at various reprogramming stages are used to train a CNN model. The CNN is then used to classify a microscopy image as probability images of these classes. Probability images of some class are used to train a densely connected convolutional network for extracting regions of this class on a microscopy image. A U-net is trained to segment cells on the time-lapse images in early reprogramming stage during culture. The segmented cells are classified by the extracted regions to count various types of cells appearing in the early reprogramming stage for predicting the identified cells potentially forming hiPSCs. The probability images of hiPSC classes are also used to train a spatiotemporal RNN for predicting the future hiPSC colony morphology of the potential cells.

RESULTS

Experimental results show the prediction (before 7 days after of beginning of the reprogramming) achieved 0.8 accuracy, and 66% of the identified cells under different culture conditions, predicted as forming, finally formed hiPSCs. The predicted hiPSC images and extracted colonies on the images show the prediction for future 1.5 days achieved high accuracy of hiPSC colony areas and image similarity.

CONCLUSIONS

Our study proposes a method using several deep learning models to efficiently select the reprogramming cells possibly forming hiPSCs and to predict the shapes of growing hiPSC colonies. The proposed method is expected to improve the efficiency when establishing a new hiPSC line culture.

Segmentation ability map: Interpret deep features for medical image segmentation

Deep convolutional neural networks (CNNs) have been widely used for medical image segmentation. In most studies, only the output layer is exploited to compute the final segmentation results and the hidden representations of the deep learned features have not been well understood. In this paper, we propose a prototype segmentation (ProtoSeg) method to compute a binary segmentation map based on deep features. We measure the segmentation abilities of the features by computing the Dice between the feature segmentation map and ground-truth, named as the segmentation ability score (SA score for short). The corresponding SA score can quantify the segmentation abilities of deep features in different layers and units to understand the deep neural networks for segmentation. In addition, our method can provide a mean SA score which can give a performance estimation of the output on the test images without ground-truth. Finally, we use the proposed ProtoSeg method to compute the segmentation map directly on input images to further understand the segmentation ability of each input image. Results are presented on segmenting tumors in brain MRI, lesions in skin images, COVID-related abnormality in CT images, prostate segmentation in abdominal MRI, and pancreatic mass segmentation in CT images. Our method can provide new insights for interpreting and explainable AI systems for medical image segmentation. Our code is available on: https://github.com/shengfly/ProtoSeg.

U-Patch GAN: A Medical Image Fusion Method Based on GAN

Although medical imaging is frequently used to diagnose diseases, in complex diagnostic situations, specialists typically need to look at different modalities of image information. Creating a composite multimodal medical image can aid professionals in making quick and accurate diagnoses of diseases. The fused images of many medical image fusion algorithms, however, are frequently unable to precisely retain the functional and structural information of the source image. This work develops an end-to-end model based on GAN (U-Patch GAN) to implement the self-supervised fusion of multimodal brain images in order to enhance the fusion quality. The model uses the classical network U-net as the generator, and it uses the dual adversarial mechanism based on the Markovian discriminator (PatchGAN) to enhance the generator's attention to high-frequency information. To ensure that the network satisfies the Lipschitz continuity, we apply the spectral norm to each layer of the network. We also propose better adversarial loss and feature loss (feature matching loss and VGG-16 perceptual loss) based on the F-norm, which significantly enhance the quality of fused images. On public data sets, we performed a lot of tests. First, we studied how clinically useful the fused image was. The model's performance in single-slice images and continuous-slice images was then confirmed by comparison with other six most popular mainstream fusion approaches. Finally, we verify the effectiveness of the adversarial loss and feature loss.

A dense residual U-net for multiple sclerosis lesions segmentation from multi-sequence 3D MR images

Multiple Sclerosis (MS) is an autoimmune disease that causes brain and spinal cord lesions, which magnetic resonance imaging (MRI) can detect and characterize. Recently, deep learning methods have achieved remarkable results in the automated segmentation of MS lesions from MRI data. Hence, this study proposes a novel dense residual U-Net model that combines attention gate (AG), efficient channel attention (ECA), and Atrous Spatial Pyramid Pooling (ASPP) to enhance the performance of the automatic MS lesion segmentation using 3D MRI sequences. First, convolution layers in each block of the U-Net architecture are replaced by residual blocks and connected densely. Then, AGs are exploited to capture salient features passed through the skip connections. The ECA module is appended at the end of each residual block and each downsampling block of U-Net. Later, the bottleneck of U-Net is replaced with the ASSP module to extract multi-scale contextual information. Furthermore, 3D MR images of Fluid Attenuated Inversion Recovery (FLAIR), T1-weighted (T1-w), and T2-weighted (T2-w) are exploited jointly to perform better MS lesion segmentation. The proposed model is validated on the publicly available ISBI2015 and MSSEG2016 challenge datasets. This model produced an ISBI score of 92.75, a mean Dice score of 66.88%, a mean positive predictive value (PPV) of 86.50%, and a mean lesion-wise true positive rate (LTPR) of 60.64% on the ISBI2015 testing set. Also, it achieved a mean Dice score of 67.27%, a mean PPV of 65.19%, and a mean sensitivity of 74.40% on the MSSEG2016 testing set. The results show that the proposed model performs better than the results of some experts and some of the other state-of-the-art methods realized related to this particular subject. Specifically, the best Dice score and the best LTPR are obtained on the ISBI2015 testing set by using the proposed model to segment MS lesions.

Self-mentoring: A new deep learning pipeline to train a self-supervised U-net for few-shot learning of bio-artificial capsule segmentation

BACKGROUND

Accurate segmentation of microscopic structures such as bio-artificial capsules in microscopy imaging is a prerequisite to the computer-aided understanding of important biomechanical phenomenons. State-of-the-art segmentation performances are achieved by deep neural networks and related data-driven approaches. Training these networks from only a few annotated examples is challenging while producing manually annotated images that provide supervision is tedious.

METHOD

Recently, self-supervision, i.e. designing a neural pipeline providing synthetic or indirect supervision, has proved to significantly increase generalization performances of models trained on few shots. The objective of this paper is to introduce one such neural pipeline in the context of micro-capsule image segmentation. Our method leverages the rather simple content of these images so that a trainee network can be mentored by a referee network which has been previously trained on synthetically generated pairs of corrupted/correct region masks.

RESULTS

Challenging experimental setups are investigated. They involve from only 3 to 10 annotated images along with moderately large amounts of unannotated images. In a bio-artificial capsule dataset, our approach consistently and drastically improves accuracy. We also show that the learnt referee network is transferable to another Glioblastoma cell dataset and that it can be efficiently coupled with data augmentation strategies.

CONCLUSIONS

Experimental results show that very significant accuracy increments are obtained by the proposed pipeline, leading to the conclusion that the self-supervision mechanism introduced in this paper has the potential to replace human annotations.

A novel denoising method for CT images based on U-net and multi-attention

Reducing the radiation dose may lead to increased noise in medical computed tomography (CT), which can adversely affect the radiologists' judgment. Many efforts have been devoted to the denoising of low-dose CT (LDCT) images. However, it is often observed that denoised medical images usually lose some important clinical lesion edge information and may affect doctors' clinical diagnosis. For a denoising neural network, it is expected that the neural network can well retain the detailed features and make the network more anthropomorphic, and to simulate the attention mechanism of observation, being a valuable feature of the thinking process of human brain. Based on U-network (U-Net) and multi-attention mechanism, a novel denoising method for medical CT images is proposed in this study. To obtain different feature information in CT images, three attention modules are proposed in our method. The local attention module is developed to localize the surrounding information of the feature map and calculate each pixel from the context extracted from the feature map. The multi-feature channel attention module can automatically learn and extract features, suppress some invalid information and add different weights to each channel in the feature map according to different tasks. The hierarchical attention module allows the deep neural network to extract a large amount of feature information. This study also introduces an enhanced learning module to learn and retain the detail information in the image by stacking multi-layer convolution layer, batch normalization (BN) layer and activation function layer to increase the network depth. Experimental studies are conducted, and comparisons with the state-of-the-art networks are made, and the results demonstrate that the developed method can effectively remove the noise in CT images and improve the image quality in the evaluation metrics of peak signal to noise ratio (PSNR) and structural similarity (SSIM). Our method achieved 34.7329 of PSNR and 0.9293 of SSIM for σ = 10 on the QIN_LUNG_CT dataset, and achieved 28.9163 of PSNR and 0.8602 of SSIM on the Mayo Clinic LDCT Grand Challenge dataset.

Semantic segmentation in medical images through transfused convolution and transformer networks

Recent decades have witnessed rapid development in the field of medical image segmentation. Deep learning-based fully convolution neural networks have played a significant role in the development of automated medical image segmentation models. Though immensely effective, such networks only take into account localized features and are unable to capitalize on the global context of medical image. In this paper, two deep learning based models have been proposed namely USegTransformer-P and USegTransformer-S. The proposed models capitalize upon local features and global features by amalgamating the transformer-based encoders and convolution-based encoders to segment medical images with high precision. Both the proposed models deliver promising results, performing better than the previous state of the art models in various segmentation tasks such as Brain tumor, Lung nodules, Skin lesion and Nuclei segmentation. The authors believe that the ability of USegTransformer-P and USegTransformer-S to perform segmentation with high precision could remarkably benefit medical practitioners and radiologists around the world.

RUC-Net: A Residual-Unet-Based Convolutional Neural Network for Pixel-Level Pavement Crack Segmentation

Automatic crack detection is always a challenging task due to the inherent complex backgrounds, uneven illumination, irregular patterns, and various types of noise interference. In this paper, we proposed a U-shaped encoder-decoder semantic segmentation network combining Unet and Resnet for pixel-level pavement crack image segmentation, which is called RUC-Net. We introduced the spatial-channel squeeze and excitation (scSE) attention module to improve the detection effect and used the focal loss function to deal with the class imbalance problem in the pavement crack segmentation task. We evaluated our methods using three public datasets, CFD, Crack500, and DeepCrack, and all achieved superior results to those of FCN, Unet, and SegNet. In addition, taking the CFD dataset as an example, we performed ablation studies and compared the differences of various scSE modules and their combinations in improving the performance of crack detection.

Fully Automated Segmentation Models of Supratentorial Meningiomas Assisted by Inclusion of Normal Brain Images

To train an automatic brain tumor segmentation model, a large amount of data is required. In this paper, we proposed a strategy to overcome the limited amount of clinically collected magnetic resonance image (MRI) data regarding meningiomas by pre-training a model using a larger public dataset of MRIs of gliomas and augmenting our meningioma training set with normal brain MRIs. Pre-operative MRIs of 91 meningioma patients (171 MRIs) and 10 non-meningioma patients (normal brains) were collected between 2016 and 2019. Three-dimensional (3D) U-Net was used as the base architecture. The model was pre-trained with BraTS 2019 data, then fine-tuned with our datasets consisting of 154 meningioma MRIs and 10 normal brain MRIs. To increase the utility of the normal brain MRIs, a novel balanced Dice loss (BDL) function was used instead of the conventional soft Dice loss function. The model performance was evaluated using the Dice scores across the remaining 17 meningioma MRIs. The segmentation performance of the model was sequentially improved via the pre-training and inclusion of normal brain images. The Dice scores improved from 0.72 to 0.76 when the model was pre-trained. The inclusion of normal brain MRIs to fine-tune the model improved the Dice score; it increased to 0.79. When employing BDL as the loss function, the Dice score reached 0.84. The proposed learning strategy for U-net showed potential for use in segmenting meningioma lesions.

Intelligent image analysis recognizes important orchid viral diseases

Phalaenopsis orchids are one of the most important exporting commodities for Taiwan. Most orchids are planted and grown in greenhouses. Early detection of orchid diseases is crucially valuable to orchid farmers during orchid cultivation. At present, orchid viral diseases are generally identified with manual observation and the judgment of the grower's experience. The most commonly used assays for virus identification are nucleic acid amplification and serology. However, it is neither time nor cost efficient. Therefore, this study aimed to create a system for automatically identifying the common viral diseases in orchids using the orchid image. Our methods include the following steps: the image preprocessing by color space transformation and gamma correction, detection of leaves by a U-net model, removal of non-leaf fragment areas by connected component labeling, feature acquisition of leaf texture, and disease identification by the two-stage model with the integration of a random forest model and an inception network (deep learning) model. Thereby, the proposed system achieved the excellent accuracy of 0.9707 and 0.9180 for the image segmentation of orchid leaves and disease identification, respectively. Furthermore, this system outperformed the naked-eye identification for the easily misidentified categories [cymbidium mosaic virus (CymMV) and odontoglossum ringspot virus (ORSV)] with the accuracy of 0.842 using two-stage model and 0.667 by naked-eye identification. This system would benefit the orchid disease recognition for Phalaenopsis cultivation.

A vegetation classification method based on improved dual-way branch feature fusion U-net

Aiming at the problems of complex structure parameters and low feature extraction ability of U-Net used in vegetation classification, a deep network with improved U-Net and dual-way branch input is proposed. Firstly, The principal component analysis (PCA) is used to reduce the dimension of hyperspectral remote sensing images, and the effective bands are obtained. Secondly, the depthwise separable convolution and residual connections are combined to replace the common convolution layers of U-Net for depth feature extraction to ensure classification accuracy and reduce the complexity of network parameters. Finally, normalized difference vegetation index (NDVI), gray level co-occurrence matrix (GLCM) and edge features of hyperspectral remote sensing images are extracted respectively. The above three artificial features are fused as one input, and PCA dimension reduction features are used as another input. Based on the improved U-net, a dual-way vegetation classification model is generated. Taking the hyperspectral remote sensing image of Matiwan Village, Xiong'an, Beijing as the experimental object, the experimental results show that the precision and recall of the improved U-Net are significantly improved with the residual structure and depthwise separable convolution, reaching 97.13% and 92.36% respectively. In addition, in order to verify the effectiveness of artificial features and dual-way branch design, the accuracy of single channel and the dual-way branch are compared. The experimental results show that artificial features in single channel network interfere with the original hyperspectral data, resulting in reduction of the recognition accuracy. However, the accuracy of the dual-way branch network has been improved, reaching 98.67%. It shows that artificial features are effective complements of network features.

Objective evaluation of chondrocyte density & cloning after joint injury using convolutional neural networks

Variations in chondrocyte density and organization in cartilage histology sections are associated with osteoarthritis progression. Rapid, accurate quantification of these two features can facilitate the evaluation of cartilage health and advance the understanding of their significance. The goal of this work was to adapt deep-learning-based methods to detect articular chondrocytes and chondrocyte clones from safranin-O-stained cartilage to evaluate chondrocyte cellularity and organization. The U-net and "you-only-look-once" (YOLO) models were trained and validated for identifying chondrocytes and chondrocyte clones, respectively. Validated models were then used to quantify chondrocyte and clone density in talar cartilage from Yucatan minipigs sacrificed 1 week, 3, 6, and 12 months after fixation of an intra-articular fracture of the hock joint. There was excellent/good agreement between expert researchers and the developed models in identifying chondrocytes/clones (U-net: R²  = 0.93, y = 0.90x-0.69; median F1 score: 0.87/YOLO: R²  = 0.79, y = 0.95x; median F1 score: 0.67). Average chondrocyte density increased 1 week after fracture (from 774 to 856 cells/mm² ), decreased substantially 3 months after fracture (610 cells/mm² ), and slowly increased 6 and 12 months after fracture (638 and 683 cells/mm² , respectively). Average detected clone density 3, 6, and 12 months after fracture (11, 11, 9 clones/mm² ) was higher than the 4-5 clones/mm² detected in normal tissue or 1 week after fracture and show local increases in clone density that varied across the joint surface with time. The accurate evaluation of cartilage cellularity and organization provided by this deep learning approach will increase objectivity of cartilage injury and regeneration assessments.