Segmentation Quality Research Articles

Panoptic-PartFormer++: A Unified and Decoupled View for Panoptic Part Segmentation.

Panoptic Part Segmentation (PPS) unifies panoptic and part segmentation into one task. Previous works utilize separate approaches to handle things, stuff, and part predictions without shared computation and task association. We aim to unify these tasks at the architectural level, designing the first end-to-end unified framework, Panoptic-PartFormer. Moreover, we find the previous metric PartPQ biases to PQ. To handle both issues, we first design a meta-architecture that decouples part features and things/stuff features, respectively. We model things, stuff, and parts as object queries and directly learn to optimize all three forms of prediction as a unified mask prediction and classification problem. We term our model as Panoptic-PartFormer. Second, we propose a new metric Part-Whole Quality (PWQ), better to measure this task from pixel-region and part-whole perspectives. It also decouples the errors for part segmentation and panoptic segmentation. Third, inspired by Mask2Former, based on our meta-architecture, we propose Panoptic-PartFormer++ and design a new part-whole cross-attention scheme to boost part segmentation qualities further. We design a new part-whole interaction method using masked cross attention. Finally, extensive ablation studies and analysis demonstrate the effectiveness of both Panoptic-PartFormer and Panoptic-PartFormer++. Compared with previous Panoptic-PartFormer, our Panoptic-PartFormer++ achieves 2% PartPQ and 3% PWQ improvements on the Cityscapes PPS dataset and 5% PartPQ on the Pascal Context PPS dataset. On both datasets, Panoptic-PartFormer++ achieves new state-of-the-art results. Our models can serve as a strong baseline and aid future research in PPS.

Analysis of the effectiveness of using U-net architecture for classification and segmentation of glioma in MRI images

The purpose of the research is to analyze the efficiency of the U-net neural network architecture in decision support systems for glioma diagnostics and segmentation of brain areas affected by it on MRI images.Methods. To conduct experimental studies, a training dataset was generated and the data was normalized. A software implementation of the U-Net neural network architecture was performed using the Keras framework in the Python programming language. The neural network model was trained.Results. A series of experiments were conducted, during which error and classification matrices were obtained, the efficiency of classification of the trained neural network model for the "Tumor" and "No tumor" classes was assessed using metrics such as Recall, Precision and F1-measure, and the quality of segmentation of glioma-affected areas on the test data set was assessed. The quality of segmentation was assessed using the IoU metric, which reflects the ratio of the areas of the bounding boxes and is used to assess the accuracy of the spatial correspondence of the predicted segmented areas highlighted on the masks. Based on the results of testing the neural network model in solving the problem of segmenting brain areas affected by glioma, the average value of the IoU metric was 0.812, which is an acceptable result.Conclusion. The testing results showed that the neural network model based on the U-net architecture is able to effectively diagnose the presence of glioma with acceptable values of the classification and segmentation quality metrics, which indicates the possibility of using this neural network model in medical decision support systems for glioma diagnostics, as well as its segmentation on MRI images. However, it is advisable to refine this neural network model to reduce the number of false negative classification results, which is critically important in medical diagnostics.

Automated lung segmentation on chest MRI in children with cystic fibrosis

IntroductionSegmentation of lung structures in medical imaging is crucial for the application of automated post-processing steps on lung diseases like cystic fibrosis (CF). Recently, machine learning methods, particularly neural networks, have demonstrated remarkable improvements, often outperforming conventional segmentation methods. Nonetheless, challenges still remain when attempting to segment various imaging modalities and diseases, especially when the visual characteristics of pathologic findings significantly deviate from healthy tissue.MethodsOur study focuses on imaging of pediatric CF patients [mean age, standard deviation (7.50 ± 4.6)], utilizing deep learning-based methods for automated lung segmentation from chest magnetic resonance imaging (MRI). A total of 165 standardized annual surveillance MRI scans from 84 patients with CF were segmented using the nnU-Net framework. Patient cases represented a range of disease severities and ages. The nnU-Net was trained and evaluated on three MRI sequences (BLADE, VIBE, and HASTE), which are highly relevant for the evaluation of CF induced lung changes. We utilized 40 cases for training per sequence, and tested with 15 cases per sequence, using the Sørensen-Dice-Score, Pearson’s correlation coefficient (r), a segmentation questionnaire, and slice-based analysis.ResultsThe results demonstrated a high level of segmentation performance across all sequences, with only minor differences observed in the mean Dice coefficient: BLADE (0.96 ± 0.05), VIBE (0.96 ± 0.04), and HASTE (0.95 ± 0.05). Additionally, the segmentation quality was consistent across different disease severities, patient ages, and sizes. Manual evaluation identified specific challenges, such as incomplete segmentations near the diaphragm and dorsal regions. Validation on a separate, external dataset of nine toddlers (2–24 months) demonstrated generalizability of the trained model achieving a Dice coefficient of 0.85 ± 0.03.Discussion and conclusionOverall, our study demonstrates the feasibility and effectiveness of using nnU-Net for automated segmentation of lung halves in pediatric CF patients, showing promising directions for advanced image analysis techniques to assist in clinical decision-making and monitoring of CF lung disease progression. Despite these achievements, further improvements are needed to address specific segmentation challenges and enhance generalizability.

Scaling Terrain-Aware Spatial Machine Learning for Flood Mapping on Large Scale Earth Imagery Data

The accurate and prompt mapping of flood-affected regions is important for effective disaster management, including damage assessment and relief efforts. While high-resolution optical imagery from satellites during disasters presents an opportunity for automated flood inundation mapping, existing segmentation models face challenges due to noises like cloud cover and tree canopies. Thanks to the digital elevation model (DEM) data readily available from sources such as United States Geological Survey (USGS), terrain guidance was utilized by recent graphical models such as hidden Markov trees (HMTs) to improve segmentation quality. Unfortunately, these methods either can only handle a small area where water levels at different locations are assumed to be consistent, or require restricted assumptions such as there is only one river channel. This paper presents an algorithm for flood extent mapping on large-scale Earth imagery, applicable to a large geographic area with multiple river channels. Since water level can vary a lot from upstream to downstream, we propose to detect river pixels in order to partition the remaining pixels into localized zones, each with a unique water level. In each zone, water at all locations flow to the same river entry point. Pixels in each zone are organized by an HMT to capture water flow directions guided by elevations. Moreover, a novel regularization scheme is designed to enforce inter-zone consistency by penalizing pixel-pairs of adjacent zones that violate terrain guidance. Efficient parallelization is made possible by coloring the zone adjacency graph to identify zones and zone-pairs that have no dependency and hence can be processed in parallel, and incremental one-pass terrain-guided scanning is conducted wherever applicable to reuse computations. Experiments demonstrate that our solution is more accurate than existing solutions, and can efficiently and accurately map out flooding pixels in a giant area of size 24,805 × 40,129. Despite the imbalanced workloads caused by a few large zonal HMTs dominating the serial computing time, our parallelization approach is effective and manages to achieve up to 14.3 × speedup on a machine with Intel Xeon Gold 6126 CPU @ 2.60GHz (24 cores, 48 threads) using 32 threads.

Chimeric U-Net – Modifying the standard U-Net towards explainability

Healthcare guided by semantic segmentation has the potential to improve our quality of life through early and accurate disease detection. Convolutional Neural Networks, especially the U-Net-based architectures, are currently the state-of-the-art learning-based segmentation methods and have given unprecedented performances. However, their decision-making processes are still an active field of research. In order to reliably utilize such methods in healthcare, explainability of how the segmentation was performed is mandated. To date, explainability is studied and applied heavily in classification tasks. In this work, we propose the Chimeric U-Net, a U-Net architecture with an invertible decoder unit, that inherently brings explainability into semantic segmentation tasks. We find that having the restriction of an invertible decoder does not hinder the performance of the segmentation task. However, the invertible decoder helps to disentangle the class information in the latent space embedding and to construct meaningful saliency maps. Furthermore, we found that with a simple k-Nearest-Neighbours classifier, we could predict the Intersection over Union scores of unseen data, demonstrating that the latent space, constructed by the Chimeric U-Net, encodes an interpretable representation of the segmentation quality. Explainability is an emerging field, and in this work, we propose an alternative approach, that is, rather than building tools for explaining a generic architecture, we propose constraints on the architecture which induce explainability. With this approach, we could peer into the architecture to reveal its class correlations and local contextual dependencies, taking an insightful step towards trustworthy and reliable AI. Code to build and utilize the Chimeric U-Net is made available under:https://github.com/kenrickschulze/Chimeric-UNet---Half-invertible-UNet-in-Pytorch

The correlation between left atrium voltage mapping and 3D late gadolinium enhancement cardiac magnetic resonance imaging findings in naive AF ablation candidate

Abstract Introduction The extent of the efficacy of late gadolinium enhancement cardiac MR (LGE-CMR) in depicting left atrial (LA) fibrosis in atrial fibrillation (AF) patients, compared to invasive Electroanatomical Mapping (EAM), is ongoing research field, while the highest degree of accuracy found when using ADAS program. Earlier studies focused mainly on post-ablation patients, employing segmentation (quality) and point-by-point (PBP) comparison methods. The non-invasive LGE-CMR-based fibrosis map shows promise for preprocedural planning and real-time guidance. Aim Evaluate LGE-CMR-based LA fibrosis precision in ablation-naïve AF patients using ADAS software, comparing it to the gold standard EAM using CARTO-3, in both qualitative and quantitative methods. Methods Patients planned for first CARTO-based AF ablation, undergoing pre-ablation LGE-CMR. 3D substrate atrial maps were generated via ADAS based on LGE-CMR (CMR-fib) and via CARTO as EAM. Normal range LA voltage was defined by Image Intensity Ratio of 1.2-1.32 on ADAS and by 0.2-0.5mV on Carto. Comparison methods, qualitative segmentation and PBP quantitative, with EAM serving as the gold standard. Results 24 naive patients included in the study (mean age 63.5 ± 10.9 yrs; 7 females, 15 paroxysmal AF, 17 maps during sinus). Segmental qualitative analysis (384 segments) showed 77% scarred segments, predominantly small scars (up to ⅓ segment). Compared with EAM, the CMR had 68.4% agreement, with 80.4% sensitivity, 78% PPV, 36.6% specificity, 32.6% NPV , kappa 0.139 AUC 0.567. Better agreement was found for dense scar. PBP quantitative comparison (41,345 points) demonstrated 50% scarred points, with 89% accuracy between EAM and CMR, with 49.6% sensitivity, 99.7% specificity, 99.5% PPV, and 73.5% NPV kappa 0.53, AUC 0.747. In both methods, agreement was worse for LA roof and best for AP and septal wall. Conclusion LGE-CMR tends to underestimate scar presence in naive patients compared to EAM, potentially due to baseline voltage scars, which are less dense compared to iatrogenic scar, the voltage criteria for scar in literature. Sensitivity is higher in segmental comparison, while specificity is superior in PBP, consistent with expectations of minimizing the size of comparison (from a segment to point). Hence, qualitative segmental method is better when assessing overall scarring, while PBP for assessing smaller specific areas. PBP comparisons exhibit less bias than segmental comparisons, which are overly sensitive to small scars. The diminished reliability of the LA roof is probably due to the margin of the CMR scan. Our study result regarding low sensitivity and high specificity of PBP aligns with a prior study investigating naïve patients through PBP comparisons, highlighting the need for a larger-scale study to assess LGE-CMR accuracy, considering various wall thicknesses and establishing scar thresholds on ADAS and CARTO software, for ablation-naïve patients with LA body scars.PBP projection on LA bodyAUC for PBP comprasion

Improved algorithm for fracture-dissolution pore detection in resistivity imaging logging based on dung beetle optimization

Abstract Resistivity imaging logging has become a direct and precise method for visualizing the structural complexities of reservoir fractures and dissolution pores. The current use of Otsu's thresholding for segmentation results in poor segmentation quality and significant noise. Accurate segmentation of sub-images containing fracture and dissolution pore targets is essential for automated structure identification and subsequent parameter calculation. This study leverages the rapid convergence and robust global optimization capabilities of the Dung Beetle Optimizer to develop enhanced image segmentation approaches. Specifically, it introduces a refined K-Means algorithm for multi-category image segmentation and an Otsu algorithm for multi-threshold image segmentation, both optimized by the Dung Beetle Optimizer. Compared to conventional binary segmentation algorithms, this new algorithm effectively isolates noise and extracts multi-category information. Using the segmented sub-images, this paper integrates mathematical morphology techniques to compute parameters such as area, perimeter, tortuosity length, and pore shape factor for identified targets. Additionally, Principal Component Analysis is used to derive recognition factors for fractures and dissolution pores. Applications show that this factor can identify matrix, fracture, and dissolution pore targets in complex background images. By combining parameter information of the target area, the method effectively removes false information in resistivity imaging and segments sub-images of fractures and dissolution pores, calculating fracture area ratio, dissolution pore area ratio, and total area ratio.

RSPS-SAM: A Remote Sensing Image Panoptic Segmentation Method Based on SAM

Satellite remote sensing images contain complex and diverse ground object information and the images exhibit spatial multi-scale characteristics, making the panoptic segmentation of satellite remote sensing images a highly challenging task. Due to the lack of large-scale annotated datasets for panoramic segmentation, existing methods still suffer from weak model generalization capabilities. To mitigate this issue, this paper leverages the advantages of the Segment Anything Model (SAM), which can segment any object in remote sensing images without requiring any annotations and proposes a high-resolution remote sensing image panoptic segmentation method called Remote Sensing Panoptic Segmentation SAM (RSPS-SAM). Firstly, to address the problem of global information loss caused by cropping large remote sensing images for training, a Batch Attention Pyramid was designed to extract multi-scale features from remote sensing images and capture long-range contextual information between cropped patches, thereby enhancing the semantic understanding of remote sensing images. Secondly, we constructed a Mask Decoder to address the limitation of SAM requiring manual input prompts and its inability to output category information. This decoder utilized mask-based attention for mask segmentation, enabling automatic prompt generation and category prediction of segmented objects. Finally, the effectiveness of the proposed method was validated on the high-resolution remote sensing image airport scene dataset RSAPS-ASD. The results demonstrate that the proposed method achieves segmentation and recognition of foreground instances and background regions in high-resolution remote sensing images without the need for prompt input, while providing smooth segmentation boundaries with a panoptic segmentation quality (PQ) of 57.2, outperforming current mainstream methods.

A simulation study towards sparse-view spectral X-ray CT segmentation using spectrum-aware deep neural networks

ABSTRACT Spectral X-ray computed tomography (SXCT) is able to provide richer spectral information than conventional X-ray CT, thereby improving the capability of discriminating materials in segmentation tasks. However, segmentation algorithms for conventional CT images often lose their performance in material segmentation with SXCT images due to the spectrum-dependent attenuation properties of X-rays. Recently, deep neural networks (DNNs) have demonstrated their effectiveness in segmentation tasks. However, processing SXCT with DNNs requires significant memory resources and may take significant computation time. To address these issues, this paper introduces a simulation study towards realising deep-learning-based SXCT image segmentation. The proposed method employs a convolutional neural network (CNN) that selectively utilises the significant spectrum channels for accurate material segmentation. The proposed method builds upon a prior study on adaptive CT reconstruction, but to further enhance the performance, we introduce a self-attention-based spectrum-selection module. During training, the selection module assigns importance weights to spectrum channels based on the attention scores. Once the network has been trained, the proposed method can obtain satisfactory segmentation results from projection images at one or a combination of several spectrum channels recommended by the selection module. The experimental results demonstrate that the proposed method achieves a significant reduction in computation time while maintaining the segmentation quality.

CellSAM: Advancing Pathologic Image Cell Segmentation via Asymmetric Large-Scale Vision Model Feature Distillation Aggregation Network.

Segment anything model (SAM) has attracted extensive interest as a potent large-scale image segmentation model, with prior efforts adapting it for use in medical imaging. However, the precise segmentation of cell nucleus instances remains a formidable challenge in computational pathology, given substantial morphological variations and the dense clustering of nuclei with unclear boundaries. This study presents an innovative cell segmentation algorithm named CellSAM. CellSAM has the potential to improve the effectiveness and precision of disease identification and therapy planning. As a variant of SAM, CellSAM integrates dual-image encoders and employs techniques such as knowledge distillation and mask fusion. This innovative model exhibits promising capabilities in capturing intricate cell structures and ensuring adaptability in resource-constrained scenarios. The experimental results indicate that this structure effectively enhances the quality and precision of cell segmentation. Remarkably, CellSAM demonstrates outstanding results even with minimal training data. In the evaluation of particular cell segmentation tasks, extensive comparative analyzes show that CellSAM outperforms both general fundamental models and state-of-the-art (SOTA) task-specific models. Comprehensive evaluation metrics yield scores of 0.884, 0.876, and 0.768 for mean accuracy, recall, and precision respectively. Extensive experiments show that CellSAM excels in capturing subtle details and complex structures and is capable of segmenting cells in images accurately. Additionally, CellSAM demonstrates excellent performance on clinical data, indicating its potential for robust applications in treatment planning and disease diagnosis, thereby further improving the efficiency of computer-aided medicine.

A Deep Learning-based Pipeline for Segmenting the Cerebral Cortex Laminar Structure in Histology Images.

Characterizing the anatomical structure and connectivity between cortical regions is a critical step towards understanding the information processing properties of the brain and will help provide insight into the nature of neurological disorders. A key feature of the mammalian cerebral cortex is its laminar structure. Identifying these layers in neuroimaging data is important for understanding their global structure and to help understand the connectivity patterns of neurons in the brain. We studied Nissl-stained and myelin-stained slice images of the brain of the common marmoset (Callithrix jacchus), which is a new world monkey that is becoming increasingly popular in the neuroscience community as an object of study. We present a novel computational framework that first acquired the cortical labels using AI-based tools followed by a trained deep learning model to segment cerebral cortical layers. We obtained a Euclidean distance of for the cortical labels acquisition, which was in the acceptable range by computing the half Euclidean distance of the average cortex thickness ( ). We compared our cortical layer segmentation pipeline with the pipeline proposed by Wagstyl et al. (PLoS biology, 18(4), e3000678 2020) adapted to 2D data. We obtained a better mean percentile Hausdorff distance (95HD) of . Whereas a mean 95HD of was obtained from Wagstyl et al. We also compared our pipeline's performance against theirs using their dataset (the BigBrain dataset). The results also showed better segmentation quality, Jaccard Index acquired from our pipeline, while was stated in their paper.

Evolutionary Grid Optimization and Deep Learning for Improved In Vitro Cellular Spheroid Localization

The recently developed high-throughput system for cell spheroid generation (SpheroWell) is a promising technology for cost- and time-efficient in vitro analysis of, for example, chondrogenic differentiation. It is a compartmental growth surface where spheroids develop from a cell monolayer by self-assembling and aggregation. In order to automatize the analysis of spheroids, we aimed to develop imaging software and improve the localization of cell compartments and fully formed spheroids. Our workflow provides automated detection and localization of spheroids in different formation stages within Petri dishes based on images created with a low-budget camera imaging setup. This automated detection enables a fast and inexpensive analysis workflow by processing a stack of images within a short period of time, which is essential for the extraction of early readout parameters. Our workflow combines image processing algorithms and deep learning-based image localization/segmentation methods like Mask R-CNN and Unet++. These methods are refined by an evolution strategy for automated grid detection, which is able to improve the overall segmentation and classification quality. Besides the already pre-trained neural networks and predefined image processing parameters, our evolution-based post-processing provides the required adaptability for our workflow to deliver a consistent and reproducible quality. This is especially important due to the use of a low-budget imaging setup with various light conditions. The to-be-detected objects of the three different stages show improved results using our evolutionary post-processing for monolayer and starting aggregation with Dice coefficients of 0.7301 and 0.8562, respectively, compared with the raw scores of 0.2879 and 0.8187. The Dice coefficient of the fully formed spheroids in both cases is 0.8829. With our algorithm, we provide automated analyses of cell spheroid by self-assembling in SpheroWell dishes, even if the images are created using a low-budget camera setup.

Market segments and pricing of fine wines over their lifecycle

Is the fine wine market segmented? The literature shows that this is the case for ordinary wines, and this has significant implications for hedonic analysis. We apply a finite mixture model to a 20-year dataset of auction prices. Our results reveal the existence of five segments in the fine wine market that differ in both their composition and the effects of scores, age, and market conditions on prices. This segmentation helps reconcile the divergent findings in existing research: some wines are highly sensitive to market conditions and have yielded significant returns (investment wines), while prices of other wines are primarily influenced by age (collectible wines) or quality (drinking wines). Our findings enhance the understanding of fine wine pricing and offer new insights into the age-price relationship, which is generally linear but varies by segment and vintage quality.

A machine learning algorithm for creating isotropic 3D aortic segmentations from routine cardiac MR localizers

BackgroundThe identification and measurement of aortic aneurysms is an important clinical problem. While specialized high-resolution 3D CMR sequences allow detailed aortic assessment, they are time-consuming which limits their use in screening routine cardiac scans and in population studies. MethodsA 3D U-Net, U-NetLR, was used to create 3D isotropic segmentations of the aorta from standard anisotropic 2D trans-axial localizers with low through-plane resolution. Training data was generated from high-resolution 3D isotropic whole heart images by simulating anisotropic images that resemble the low-resolution 2D localizers (the inputs). These inputs were paired with 3D isotropic ‘ground truth’ segmentation masks (the targets) created by a clinician from the high-resolution isotropic images. Segmentation quality was evaluated using an external dataset from UK Biobank. Segmentation accuracy was measured against ground-truth segmentations from concurrently acquired cardiac-triggered, respiratory-gated, high-resolution 3D isotropic whole heart images. Finally, the proposed method was compared to U-NetHR, a 3D U-Net variant trained directly on high-resolution 3D isotropic images. A second observer was recruited to investigate the interobserver variability. ResultsQualitative validation on an external dataset (UK Biobank) of 180 subjects showed that 93 % of 3D segmentations with the proposed model (U-NetLR) were considered suitable for clinical use. In quantitative analysis, the proposed method (U-NetLR) showed good agreement with ground-truth segmentations from isotropic 3D images with a mean DICE score of 0.9, which is no difference from automated segmentations made directly on the high-resolution 3D isotropic aorta images (U-NetHR). When comparing measurements, there is no significant difference between U-NetLR, U-NetHR and two clinical observers in the diameter measurements at the mid ascending aorta, mid aortic arch, and descending aorta. ConclusionsA new method of producing isotropic 3D aortic segmentations from routine CMR 2D anisotropic localizers shows good agreement with segmentation made directly from 3D isotropic images. The method has the potential to be used as a simple screening method for aortic aneurysms without the need for additional sequences.

Hyperparameter selection for dataset-constrained semantic segmentation: Practical machine learning optimization.

This paper provides a pedagogical example for systematic machine learning optimization in small dataset image segmentation, emphasizing hyperparameter selections. A simple process is presented for medical physicists to examine hyperparameter optimization. This is also applied to a case-study, demonstrating the benefit of the method. An unrestricted public Computed Tomography (CT) dataset, with binary organ segmentation, was used to develop a multiclass segmentation model. To start the optimization process, a preliminary manual search of hyperparameters was conducted and from there a grid search identified the most influential result metrics. A total of 658 different models were trained in 2100h, using 13160 effective patients. The quantity of results was analyzed using random forest regression, identifying relative hyperparameter impact. Metric implied segmentation quality (accuracy 96.8%, precision 95.1%) and visual inspection were found to be mismatched. In this work batch normalization was most important, but performance varied with hyperparameters and metrics selected. Targeted grid-search optimization and random forest analysis of relative hyperparameter importance, was an easily implementable sensitivity analysis approach. The proposed optimization method gives a systematic and quantitative approach to something intuitively understood, that hyperparameters change model performance. Even just grid search optimization with random forest analysis presented here can be informative within hardware and data quality/availability limitations, adding confidence to model validity and minimize decision-making risks. By providing a guided methodology, this work helps medical physicists to improve their model optimization, irrespective of specific challenges posed by datasets and model design.

Multi-label semantic segmentation of magnetic resonance images of the prostate gland

Prostate segmentation is a substantial factor in the diagnostic pathway of suspicious prostate lesions. Medical doctors are assisted by computer-aided detection and diagnosis systems systems and methods derived from artificial intelligence deep learning-based systems have to be trained on existing data. Especially freely available labeled prostate magnetic resonance imaging data is rare. With regard to this problem, we show a method to combine two existing small datasets to form a bigger one. We present a data processing pipeline consisting of a cascaded network architecture that is able to perform multi-label semantic segmentation, hence, capable of classifying each pixel in a T2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ extrm{T}_{2}$$\\end{document}-weighted image not only to one class but to a subset of the prostate zones and classes of interest. This delivers richer information such as overlapping zones that are key to medical radiological examination. Additionally, we describe how to integrate expert knowledge in our deep learning system. To increase data variety for training and evaluation we use image augmentation on our two datasets—a freely available dataset and our new open-source dataset. To combine the datasets we denoise the contourings in our dataset by using an effective yet simple algorithm based on standard computer vision methods only. The performance of the presented methodology is compared and evaluated using the dice score metric and 5-fold cross-validation on all datasets. Although we trained on tiny datasets our method achieves excellent segmentation quality and is even able to detect prostate cancer. Our method to combine the two datasets reduces segmentation errors and increases data variety. The proposed architecture significantly improves performance by including expert knowledge via feature-map concatenation. On the initiative for collaborative computer vision benchmarking dataset we achieve on average dice scores of approximately 91% for the whole prostate gland, 67% for the peripheral zone and 75% for the prostate central gland. We find that image augmentation except contrast limited adaptive histogram equalisation did not have much influence on the segmentation quality. Derived and enhanced from existing methods we present an approach that is able to deliver multi-label semantic segmentation results for prostate magnetic resonance imaging. This approach is simple and could be applied to other applications of deep learning as well. It improves the segmentation results by a large margin. Once tweaked to the data, our denoising and combination algorithm delivers robust and accurate results even on data with segmentation errors.

A novel metastatic tumor segmentation method with a new evaluation metric in clinic study.

Brain metastases are the most common brain malignancies. Automatic detection and segmentation of brain metastases provide significant assistance for radiologists in discovering the location of the lesion and making accurate clinical decisions on brain tumor type for precise treatment. However, due to the small size of the brain metastases, existing brain metastases segmentation produces unsatisfactory results and has not been evaluated on clinic datasets. In this work, we propose a new metastasis segmentation method DRAU-Net, which integrates a new attention mechanism multi-branch weighted attention module and DResConv module, making the extraction of tumor boundaries more complete. To enhance the evaluation of both the segmentation quality and the number of targets, we propose a novel medical image segmentation evaluation metric: multi-objective segmentation integrity metric, which effectively improves the evaluation results on multiple brain metastases with small size. Experimental results evaluated on the BraTS2023 dataset and collected clinical data show that the proposed method has achieved excellent performance with an average dice coefficient of 0.6858 and multi-objective segmentation integrity metric of 0.5582. Compared with other methods, our proposed method achieved the best performance in the task of segmenting metastatic tumors.

Impact of an institutional process change adopting end-systolic coronary CTA acquisition and automated dose selection on patient throughput and image quality

IntroductionGuidelines recommend prospective ECG-triggered mid-diastolic coronary computed tomographic angiography (CCTA) acquisition after achieving optimal heart rate (HR) control in order to optimize scan image quality. With dual-source CCTA, prospective end-systolic acquisition has been shown to be less prone to motion artifacts at higher heart rates and may improve scan and CT laboratory efficiency by allowing CCTA without routine pre-scan beta-blocker (BB) administration. MethodsWe implemented an institutional process change in CCTA performance effective January 2023, comprising a transition from prospective ECG-triggered mid-diastolic acquisitions individually supervised by a physician at the scanner to an algorithmic approach predominately utilizing prospective end-systolic acquisition (200–400 ​ms after R peak), employing an automated dose selection algorithm, without BB administration. All scans were performed on a third-generation 192-slice dual-source scanner. We reviewed 300 consecutive CCTAs done pre- and post-process change in Jan 2022 (phase 0), Jan 2023 (phase 1), and in May 2023 (phase 2) after implementation of a process improvement involving more selective utilization of automated tube potential/current algorithms (CARE kV) to optimize image quality. Coronary segmental image quality was assessed by two experienced CCTA readers by consensus using an 18-segment SCCT model on a 5-point Likert scale (1 ​= ​non-interpretable; 2 ​= ​poor; 3 ​= ​acceptable; 4 ​= ​good; 5 ​= ​excellent). Measures of radiation dose, medication administration, and time required for patient scanning were compared. Logistic regression was used to determine factors associated with patient-level reduction in image quality (IQ) and with repeat scans. ResultsPost-process change, there was a significant reduction in the median overall patient appointment [phase 0: 95 (75–125) min vs. phase 1: 68 (52–88) min and phase 2: 72 (59–90) min; P ​< ​0.001] and scan times [phase 0: 13 (10–16) min vs. phase 1: 8 (6–13) min and phase 2: 9 (7–13) min; P ​< ​0.001]. Median IQ score in both post-process change phases was 4 (4–5) compared to a median score of 5 (4–5) pre-process change (P for comparison <0.001). The majority of segments post-process change had “good” IQ (Phase 1 segmental IQ scores: 5 ​= ​36.7 ​%, 4 ​= ​46.8 ​%, 3 ​= ​13 ​%, 2 ​= ​2.6 ​%, 1 ​= ​0.9 ​%; Phase 2 segmental IQ scores: 5 ​= ​26 ​%, 4 ​= ​49.7 ​%, 3 ​= ​16.3 ​%, 2 ​= ​6.1 ​%, 1 ​= ​1.9 ​%), whereas pre-process change, the majority of segments had “excellent” IQ (Phase 0 segmental IQ scores: 5 ​= ​56 ​%, 4 ​= ​34.3 ​%, 3 ​= ​7.5 ​%, 2 ​= ​1.8 ​%, 1 ​= ​0.4 ​%) There was no significant increase in non-interpretable scans at the patient level. The 22 ​% re-scan rate in phase 1 (vs. 6 ​% in phase 0, P ​= ​.002) improved to 15 ​% in phase 2. While patient related factors of body mass index [adjusted OR obese 2.64, 95 ​% CI 1.12–6.51, P ​= ​0.03; aOR morbidly obese 6.94, 95 ​% CI 2.21–23.52, P ​= ​0.001] and average HR [aOR (per 10 bpm increase) 1.51, 95 ​% CI 1.21–1.9, P ​< ​0.001] were associated with the scoring of any segment as ​≤ ​3 ​at the patient level in a fully adjusted model, the improved phase 2 of the process change was not [aOR 1.61, 95 ​% CI 0.78–3.32]. ConclusionImplementation of an institutional process change utilizing prospective ECG-triggered dual-source end-systolic acquisition avoided the use of beta-blockers, significantly reduced patient appointment and scan times with acceptable diagnostic performance.

Adaptive pre-processing methods for increasing the accuracy of segmentation of dental X-RAY images

The subject of research in the article is the effectiveness of adaptive methods of preprocessing of medical images, in particular adaptive bilateral filter and modified CLAHE, in the tasks of segmentation of dental X-ray images. These methods make it possible to preserve important image details and effectively reduce noise, even in cases of high variability of images coming from different sources. The goal of the work is to study the impact of adaptive preprocessing methods on increasing the accuracy of segmentation of medical images and to determine the optimal combination of methods that provide the best results in segmentation tasks. The article addresses the following tasks: experimental comparison of adaptive preprocessing methods with traditional approaches, evaluation of segmentation efficiency using metrics such as Dice Score, Jacquard Coefficient (IoU Score), Precision and Sensitivity/Completeness (Recall)), as well as analysis of the effect of pre-processing on the quality of segmentation. The following methods are used: mathematical modeling, neural network training based on the U-Net model with a pre-trained timm-resnest101e encoder, image scaling to 512x512 pixels, training with a fixed learning rate of 0.001. The following results were obtained: the combined use of the adaptive bilateral filter and the modified CLAHE provided the highest segmentation quality indicators (Dice Score 0.9603 and Jacquard Coefficient (IoU Score) 0.94501), surpassing traditional methods. This proves the advantage of adaptive approaches in preserving the contours of objects and reducing noise. Conclusions: the application of adaptive preprocessing methods significantly improves the accuracy of segmentation of medical images. The combined approach including the adaptive bilateral filter and the modified CLAHE is the most effective for medical imaging tasks, which increases the accuracy of diagnosis and the reliability of automated decision support systems.

Improved genetic algorithm for multi-threshold optimization in digital pathology image segmentation

This paper presents an improved genetic algorithm focused on multi-threshold optimization for image segmentation in digital pathology. By innovatively enhancing the selection mechanism and crossover operation, the limitations of traditional genetic algorithms are effectively addressed, significantly improving both segmentation accuracy and computational efficiency. Experimental results demonstrate that the improved genetic algorithm achieves the best balance between precision and recall within the threshold range of 0.02 to 0.05, and it significantly outperforms traditional methods in terms of segmentation performance. Segmentation quality is quantified using metrics such as precision, recall, and F1 score, and statistical tests confirm the superior performance of the algorithm, especially in its global search capabilities for complex optimization problems. Although the algorithm’s computation time is relatively long, its notable advantages in segmentation quality, particularly in handling high-precision segmentation tasks for complex images, are highly pronounced. The experiments also show that the algorithm exhibits strong robustness and stability, maintaining reliable performance under different initial conditions. Compared to general segmentation models, this algorithm demonstrates significant advantages in specialized tasks, such as pathology image segmentation, especially in resource-constrained environments. Therefore, this improved genetic algorithm offers an efficient and precise multi-threshold optimization solution for image segmentation, providing valuable reference for practical applications.