Subsequent Image Research Articles

Automated decision support for Hallux Valgus treatment options using anteroposterior foot radiographs

BACKGROUND Assessment of the potential utility of deep learning with subsequent image analysis to automate the measurement of hallux valgus and intermetatarsal angles from radiographs to serve as a preoperative aid in establishing hallux valgus severity for clinical decision-making. AIM To investigate the accuracy of automated measurements of angles of hallux valgus from radiographs for further integration with the preoperative planning process. METHODS The data comprises 265 consecutive digital anteroposterior weightbearing foot radiographs. 181 radiographs were utilized for training (161) and validating (20) a U-Net neural network to achieve a mean Sørensen–Dice index > 97% on bone segmentation. 84 test radiographs were used for manual (computer assisted) and automated measurements of hallux valgus severity determined by hallux valgus (HVA) and intermetatarsal angles (IMA). The reliability of manual and computer-based measurements was calculated using the interclass correlation coefficient (ICC) and standard error of measurement (SEM). Inter- and intraobserver reliability coefficients were also compared. An operative treatment recommendation was then applied to compare results between automated and manual angle measurements. RESULTS Very high reliability was achieved for HVA and IMA between the manual measurements of three independent clinicians. For HVA, the ICC between manual measurements was 0.96-0.99. For IMA, ICC was 0.78-0.95. Comparing manual against automated computer measurement, the reliability was high as well. For HVA, absolute agreement ICC and consistency ICC were 0.97, and SEM was 0.32. For IMA, absolute agreement ICC was 0.75, consistency ICC was 0.89, and SEM was 0.21. Additionally, a strong correlation (0.80) was observed between our approach and traditional clinical adjudication for preoperative planning of hallux valgus, according to an operative treatment algorithm proposed by EFORT. CONCLUSION The proposed automated, artificial intelligence assisted determination of hallux valgus angles based on deep learning holds great potential as an accurate and efficient tool, with comparable accuracy to manual measurements by expert clinicians. Our approach can be effectively implemented in clinical practice to determine the angles of hallux valgus from radiographs, classify the deformity severity, streamline preoperative decision-making prior to corrective surgery.

CGP-Uformer: A low-dose CT image denoising Uformer based on channel graph perception.

An effective method for achieving low-dose CT is to keep the number of projection angles constant while reducing radiation dose at each angle. However, this leads to high-intensity noise in the reconstructed image, adversely affecting subsequent image processing, analysis, and diagnosis. This paper proposes a novel Channel Graph Perception based U-shaped Transformer (CGP-Uformer) network, aiming to achieve high-performance denoising of low-dose CT images. The network consists of convolutional feed-forward Transformer (ConvF-Transformer) blocks, a channel graph perception block (CGPB), and spatial cross-attention (SC-Attention) blocks. The ConvF-Transformer blocks enhance the ability of feature representation and information transmission through the CNN-based feed-forward network. The CGPB introduces Graph Convolutional Network (GCN) for Channel-to-Channel feature extraction, promoting the propagation of information across distinct channels and enabling inter-channel information interchange. The SC-Attention blocks reduce the semantic difference in feature fusion between the encoder and decoder by computing spatial cross-attention. By applying CGP-Uformer to process the 2016 NIH AAPM-Mayo LDCT challenge dataset, experiments show that the peak signal-to-noise ratio value is 35.56 and the structural similarity value is 0.9221. Compared to the other four representative denoising networks currently, this new network demonstrates superior denoising performance and better preservation of image details.

Development and evaluation of a vision driven sensor for estimating fuel feeding rates in combustion and gasification processes

A machine vision driven sensor for estimating the instantaneous feeding rate of pelletized fuels was developed and tested experimentally in combustion and gasification processes. The feeding rate was determined from images of the pellets sliding on a transfer chute into the reactor. From the images the apparent area and velocity of the pellets were extracted. Area was determined by a segmentation model created using a machine learning framework and velocities by image registration of two subsequent images. The measured weight of the pelletized fuel passed through the feeding system was in good agreement with the weight estimated by the sensor. The observed variations in the fuel feeding correlated with the variations in the gaseous species concentrations measured in the reactor core and in the exhaust. Since the developed sensor measures the ingoing fuel feeding rate prior to the reactor, its signal could therefore help improve process control.

Motion detail feature extraction of sports injury based on three-dimensional sensor tracking

In response to the issues of low extraction accuracy and prolonged extraction time associated with traditional methods of motion detail feature extraction in sports injury analysis, we propose a novel approach based on three-dimensional sensor tracking. This method leverages the principles of sports injury action image analysis, utilizing three-dimensional sensors to track and collect sports injury action images. These images are then subjected to wavelet transform for noise reduction and detail enhancement. Subsequent image preprocessing identifies the sports injury area, and a wavelet analysis extracts detailed characteristics from the sports injury action. Simulation experiments demonstrate the high accuracy and efficiency of our proposed method in motion detail feature extraction for sports injury.

Deep learning network for parallel self-denoising and segmentation in visible light optical coherence tomography of the human retina.

Visible light optical coherence tomography (VIS-OCT) of the human retina is an emerging imaging modality that uses shorter wavelengths in visible light range than conventional near-infrared (NIR) light. It provides one-micron level axial resolution to better separate stratified retinal layers, as well as microvascular oximetry. However, due to the practical limitation of laser safety and comfort, the permissible illumination power is much lower than NIR OCT, which can be challenging to obtain high-quality VIS-OCT images and subsequent image analysis. Therefore, improving VIS-OCT image quality by denoising is an essential step in the overall workflow in VIS-OCT clinical applications. In this paper, we provide the first VIS-OCT retinal image dataset from normal eyes, including retinal layer annotation and "noisy-clean" image pairs. We propose an efficient co-learning deep learning framework for parallel self-denoising and segmentation simultaneously. Both tasks synergize within the same network and improve each other's performance. The significant improvement of segmentation (2% higher Dice coefficient compared to segmentation-only process) for ganglion cell layer (GCL), inner plexiform layer (IPL) and inner nuclear layer (INL) is observed when available annotation drops to 25%, suggesting an annotation-efficient training. We also showed that the denoising model trained on our dataset generalizes well for a different scanning protocol.

Development of novel parameters for characterising scale morphology of wool fibre and its correlation with dye diffusion coefficient of acid dye

This paper reports the development of novel surface parameters which can be used to characterise the scale structure of wool fibres obtained from different breeds. Scanning electron microscopy and subsequent image analysis technique were used to study wool fibres from Leicester, Dartmoor, Ryeland and Herdwick breeds of sheep. Novel scale parameters related to wool fibre’s effective chemical diffusion pathway were developed. Namely, the total scale perimeter per 100 µm fibre length and scale perimeter index, which is the total scale perimeter per 100 µm length divided by the fibre diameter. Wool fibres obtained from different breeds showed significant differences in their scale pattern with the change in fibre diameter. The scale perimeter per 100 µm length increased with the fibre diameter and showed a polynomial correlation. It was also demonstrated that an increase in the diameter of the wool fibre resulted in an increase in the apparent dye diffusion coefficient, which contrasts the established theory that finer fibres are associated with a higher dyeing rate. The increase in effective diffusion pathway (total scale perimeter per 100 µm) for the wool fibres (among different breeds) resulted in a higher dye diffusion rate at the initial phase of dyeing (liquor to surface).

ULBPNet: Low-light image enhancement using U-shaped lightening back-projection

Images captured under low-light conditions often have various issues, such as low brightness and contrast, and also color distortion. The aim of low-light image enhancement is to improve the visual quality of such images to facilitate subsequent image processing and computer vision tasks. However, this task faces major challenges from the inherent limitations of the low-light environment. This paper proposes a novel low-light image enhancement network using a U-shaped enhanced lightening back-projection. The network architecture primarily comprises a series of lightening blocks and a color balance block. We construct the lightening block, which incorporates a U-shaped enhanced lightening back-projection module and a saturation-guided fusion module, by assuming that the saturation levels of the low-light and the normal images are similar. Additionally, a color balance block is included to address issues such as color fading or over-enhancement. The effectiveness of the proposed network is assessed via extensive experiments conducted on widely-used benchmarks. The simulation results confirm that the proposed low-light enhancement network outperforms other state-of-the-art approaches. https://github.com/JYP-1317/.

Image cascade matching method based on an improved random sampling consistency algorithm

In the process of image matching, image quality will have a certain impact on the extraction of image features, directly affecting the matching accuracy of subsequent images. Therefore, an image cascade matching method based on improved random sampling consistency algorithm is proposed. This method first needs to improve the quality of the matched target image through preprocessing results such as image denoising and enhancement, and prepare for feature point extraction during subsequent image cascade matching; Then extract the feature points of the target to be matched in the image, and use the improved random sampling consistency method to quickly estimate and simplify the registration model based on the extraction results. Use similar feature triangles to perform model pre testing, and use the maximum Euclidean distance method to improve the accuracy of image cascade registration, achieving image cascade matching. The experimental results show that the proposed method can achieve image enhancement while removing image noise, improve the quality of the image, comprehensively extract the feature points to be matched in the image, and achieve accurate matching of image feature points.

Underwater image quality assessment method based on color space multi-feature fusion

The complexity and challenging underwater environment leading to degradation in underwater image. Measuring the quality of underwater image is a significant step for the subsequent image processing step. Existing Image Quality Assessment (IQA) methods do not fully consider the characteristics of degradation in underwater images, which limits their performance in underwater image assessment. To address this problem, an Underwater IQA (UIQA) method based on color space multi-feature fusion is proposed to focus on underwater image. The proposed method converts underwater images from RGB color space to CIELab color space, which has a higher correlation to human subjective perception of underwater visual quality. The proposed method extract histogram features, morphological features, and moment statistics from luminance and color components and concatenate the features to obtain fusion features to better quantify the degradation in underwater image quality. After features extraction, support vector regression(SVR) is employed to learn the relationship between fusion features and image quality scores, and gain the quality prediction model. Experimental results on the SAUD dataset and UIED dataset show that our proposed method can perform well in underwater image quality assessment. The performance comparisons on LIVE dataset, TID2013 dataset,LIVEMD dataset,LIVEC dataset and SIQAD dataset demonstrate the applicability of the proposed method.

Colorimetric Sensing of Ascorbic Acid Using Cu-Phen MOFs and Subsequent Digital Image Analysis with Smartphone

Vitamin C, also known as ascorbic acid, is a crucial nutrient involved in a number of enzymatic processes for tissue healing. Additionally serving as an antioxidant, ascorbic acid is crucial for keeping the immune system strong. It is essential to create a quantitative analytical approach to ascertain ascorbic acid concentration in many samples since its consumption from the daily food should be in the proper quantity. Here, we present a colorimetric probe made of synthetic metal-organic frameworks (MOFs) that changes its color only when samples containing ascorbic acid are present. In this study, phenanthroline served as an organic ligand or linker while copper served as the core metal that formed bonds with it. Cu-Phen MOF color fluctuations at 410 nm (from 408 nm to 412 nm) are linearly related to variations in light absorption. Moreover, as a substitute for spectrophotometry UV-visible, we have created a digital image-based colorimetry. The Cu-Phen MOFs’ color change has the maximum slope and linearity when the blue color intensity is used. The detection limit with high precision of 4.2% under ideal circumstances was 0.1 ppm. The established approach allowed for the exact and accurate assessment of ascorbic acid in genuine samples of star fruit (Averrhoa carambola L.).

Arterial spin labeling image findings in the acute phase in paediatric patients with acute encephalopathy with biphasic seizures and late reduced diffusion.

Diagnosing acute encephalopathy with biphasic seizures and late reduced diffusion (AESD) after the first seizure (early seizure/seizures, ES/ESs) is challenging because a reduced apparent diffusion coefficient (ADC) in the cortical or subcortical white matter, often described as having a "bright-tree appearance (BTA)," is usually not observed until secondary seizures (late seizures, LSs) occur. Previous studies have reported hypoperfusion on arterial spin labeling (ASL) within 24 h after ES/ESs in patients with AESD and hyperperfusion within 24 h after LS onset. This study aimed to investigate cerebral blood flow in the hyperacute phase (between ES/ESs and LSs) using ASL in patients with AESD. Eight ASL images were acquired in six patients with AESD admitted to our hospital from October 2021 to October 2022. ASL findings in the hyperacute phase were investigated and video-electroencephalogram findings obtained around ASL image acquisition in the hyperacute phase were evaluated. Four ASL images were obtained for three patients before LS onset, with three images showing hyperperfusion areas and one image showing hypoperfusion areas. These hyperperfuion regions coincided with BTA on subsequent images of these patients.In one patient, the first ASL image was obtained in the late hyperacute phase and revealed hyperperfusion areas with a slightly abnormal change on diffusion-weighted image (DWI), which were not accompanied by ADC abnormalities. The second ASL image obtained 51 h after the first ASL, and before LS onset revealed more prominent hyperperfusion areas than the first ASL image, which were accompanied by BTA. In another patient, the ASL image obtained 82 h after ES revealed hyperperfusion areas without abnormal change on DWI or ADC. This study revealed that two patients exhibited hyperperfusion regions and another patient exhibited hypoperfusion regions among three patients who underwent ASL imaging during the period from 24 h after ES/ESs to LSs in patients with LSs or cooling initiation in patients without LSs due to early anaesthesia induction (late hyperacute phase). Further prospective studies on cerebral blood flow are required to explore the relationship among the timing of image acquisition, the presence of electrographic seizures, and ASL findings in patients with AESD.

An Improved Visual SLAM Method with Adaptive Feature Extraction

The feature point method is the mainstream method to accomplish inter-frame estimation in visual Simultaneous Localization and Mapping (SLAM) methods, among which the Oriented FAST and Rotated BRIEF (ORB) feature-based method provides an equilibrium of accuracy as well as efficiency. However, the ORB algorithm is prone to clustering phenomena, and its unequal distribution of extracted feature points is not conducive to the subsequent camera tracking. To solve the above problems, this paper suggests an adaptive feature extraction algorithm that first constructs multiple-scale images using an adaptive Gaussian pyramid algorithm, calculates adaptive thresholds, and uses an adaptive meshing method for regional feature point detection to adapt to different scenes. The method uses Adaptive and Generic Accelerated Segment Test (AGAST) to speed up feature detection and the non-maximum suppression method to filter feature points. The feature points are then divided equally by a quadtree technique, and the orientation of those points is determined by an intensity centroid approach. Experiments were conducted on publicly available datasets, and the outcomes demonstrate the algorithm has good adaptivity and solves the problem of a large number of corner point clusters that may result from using manually set detection thresholds. The RMSE of the absolute trajectory error of SLAM applying this method on four sequences of TUM RGB-D datasets is decreased by 13.88% when compared with ORB-SLAM3. It is demonstrated that the algorithm provides high-quality feature points for subsequent image alignment, and the application to SLAM improves the reliability and accuracy of SLAM.

MI-Common Data Model: Extending Observational Medical Outcomes Partnership-Common Data Model (OMOP-CDM) for Registering Medical Imaging Metadata and Subsequent Curation Processes.

The explosion of big data and artificial intelligence has rapidly increased the need for integrated, homogenized, and harmonized health data. Many common data models (CDMs) and standard vocabularies have appeared in an attempt to offer harmonized access to the available information, with Observational Medical Outcomes Partnership (OMOP)-CDM being one of the most prominent ones, allowing the standardization and harmonization of health care information. However, despite its flexibility, still capturing imaging metadata along with the corresponding clinical data continues to pose a challenge. This challenge arises from the absence of a comprehensive standard representation for image-related information and subsequent image curation processes and their interlinkage with the respective clinical information. Successful resolution of this challenge holds the potential to enable imaging and clinical data to become harmonized, quality-checked, annotated, and ready to be used in conjunction, in the development of artificial intelligence models and other data-dependent use cases. To address this challenge, we introduce medical imaging (MI)-CDM-an extension of the OMOP-CDM specifically designed for registering medical imaging data and curation-related processes. Our modeling choices were the result of iterative numerous discussions among clinical and AI experts to enable the integration of imaging and clinical data in the context of the ProCAncer-I project, for answering a set of clinical questions across the prostate cancer's continuum. Our MI-CDM extension has been successfully implemented for the use case of prostate cancer for integrating imaging and curation metadata along with clinical information by using the OMOP-CDM and its oncology extension. By using our proposed terminologies and standardized attributes, we demonstrate how diverse imaging modalities can be seamlessly integrated in the future.

Decoupled learning for brain image registration.

Image registration is one of the important parts in medical image processing and intelligent analysis. The accuracy of image registration will greatly affect the subsequent image processing and analysis. This paper focuses on the problem of brain image registration based on deep learning, and proposes the unsupervised deep learning methods based on model decoupling and regularization learning. Specifically, we first decompose the highly ill-conditioned inverse problem of brain image registration into two simpler sub-problems, to reduce the model complexity. Further, two light neural networks are constructed to approximate the solution of the two sub-problems and the training strategy of alternating iteration is used to solve the problem. The performance of algorithms utilizing model decoupling is evaluated through experiments conducted on brain MRI images from the LPBA40 dataset. The obtained experimental results demonstrate the superiority of the proposed algorithm over conventional learning methods in the context of brain image registration tasks.

The use of aromatic aldehydes for the test determination of cefotaxime

Indicator papers based on immobilized reagents with their subsequent digital image processing are widely used for the quick and cheap screening study of drugs when assessing the content of the main substance in drugs for the presence of counterfeits. Indicator papers based on reagents immobilized on a cellulose matrix 4-dimethylaminocinnamic aldehyde (DMACA), 4-dimethylaminobenzaldehyde (DMAB), and 3-methoxy-4-hydroxybenzaldehyde (vanillin) were proposed for the determination of β-lactam antibiotic cefotaxime. To specify optimal conditions for the immobilization of reagents on a solid matrix, the intensities (changes in intensities) of color parameters were recorded at a different concentration of reagents, type of surfactants, and paper porosity. The detection limits of visual-colorimetric determination of cefotaxime were 1.0 mg/ml (DMACA), 0.12 mg/ml (DMAB), and 0.30 mg/ml (vanillin). Colorimetric processing of images of the test agents was carried out in the GIMP (GNU Image Manipulation Program) 2.10 program by averaging the images of indicator papers using the blur-sharp function, determining the intensity of color parameters. The photos were taken using a 48-megapixel camera of a POCO X3 pro smartphone in a specialized box measuring 22 × 23 × 24 cm with two LED lighting strips (5 V, 1 A). Linear dependences of the intensities of color parameters of the RGB, HSV, and CMYK models on the cefotaxime concentration (log c) were obtained: y = –15x + 45 (r2 = 0. 99, color parameter H for DMACA), y = –51x + 130 (r2 = = 0.99, parameter B for DMAB), y = 36x + 30 (r2 = 0. 99, S parameter for vanillin). Linear dependences of the areas and perimeters of the petal diagrams of indicator papers on the cefotaxime concentration have been plotted for the studied systems of cefotaxime — immobilized reagent. The correctness of the determination results was assessed using the analysis of spiked samples, the relative error did not exceed 25%. The developed test tools based on immobilized aromatic aldehydes are recommended for the semi-quantitative determination of cefotaxime.

Underwater Image Enhancement Based on Hybrid Enhanced Generative Adversarial Network

In recent years, underwater image processing has played an essential role in ocean exploration. The complexity of seawater leads to the phenomena of light absorption and scattering, which in turn cause serious image degradation problems, making it difficult to capture high-quality underwater images. A novel underwater image enhancement model based on Hybrid Enhanced Generative Adversarial Network (HEGAN) is proposed in this paper. By designing a Hybrid Underwater Image Synthesis Model (HUISM) based on a physical model and a deep learning method, many richly varied paired underwater images are acquired to compensate for the missing problem of underwater image enhancement dataset training. Meanwhile, the Detection Perception Enhancement Model (DPEM) with Perceptual Loss is designed to transfer the coding knowledge in the form of the gradient to the enhancement model through the perceptual loss, which leads to the generation of visually better and detection-friendly underwater images. Then, the synthesized and enhanced models are integrated into an adversarial network to generate high-quality underwater clear images through game learning. Experiments show that the proposed method significantly outperforms several state-of-the-art methods both qualitatively and quantitatively. Furthermore, it is also demonstrated that the method can improve target detection performance in underwater environments, which has specific application value for subsequent image processing.

Influence of ultrasound machine settings on quantitative measures derived from spatial frequency analysis of muscle tissue

BackgroundUltrasound is a powerful tool for diagnostic purposes and provides insight into both normal and pathologic tissue structure. Spatial frequency analysis (SFA) methods characterize musculoskeletal tissue organization from ultrasound images. Both sonographers in clinical imaging and researchers may alter a minimized range of ultrasound settings to optimize image quality, and it is important to know how these small adjustments of these settings affect SFA parameters. The purpose of this study was to investigate the effects of making small adjustments in a typical default ultrasound machine setting on extracted spatial frequency parameters (peak spatial frequency radius (PSFR), Mmax, Mmax%, and Sum) in the biceps femoris muscle.MethodsLongitudinal B-mode images were collected from the biceps femoris muscle in 36 participants. The window depth, foci locations, and gain were systematically adjusted consistent with clinical imaging procedures for a total of 27 images per participant. Images were analyzed by identifying a region of interest (ROI) in the middle portion of the muscle belly in a template image and using a normalized two-dimensional cross-correlation technique between the template image and subsequent images. The ROI was analyzed in the frequency domain using conventional SFA methods. Separate linear mixed effects models were run for each extracted parameter.ResultsPSFR was affected by modifications in focus location only (p < 0.001) with differences noted between all locations. Mmax% was influenced by the interaction of gain and focus location (p < 0.001) but was also independently affected by increasing window depth (p < 0.001). Both Mmax and Sum parameters were sensitive to small changes in machine settings with the interaction of focus location and window depth (p < 0.001 for both parameters) as well as window depth and gain (p < 0.001 for both) influencing the extracted values.ConclusionsFrequently adjusted imaging settings influence some SFA statistics. PSFR and Mmax% appear to be most robust to small changes in image settings, making them best suited for comparison across individuals and between studies, which is appealing for the clinical utility of the SFA method.

Recognition and Extraction of Power Transmission Lines Based on Infrared Image Processing for Line-following Robots

To further improve the real-time performance and accuracy of power transmission line maintenance, this paper primarily focuses on the preliminary line recognition and extraction method based on thermal image processing of infrared images collected by line-following robots for thermal fault detection. Firstly, filtering and noise reduction techniques along with enhanced image processing are applied to preprocess the collected infrared images. This effectively addresses the noise and interference from background objects, which can affect the extraction of overheated areas on the lines, while also reducing the computational memory required for subsequent image processing. Subsequently, an improved Canny edge detection algorithm is employed to extract the edges of foreground objects in the images. Additionally, a region-growing algorithm combined with the geometric features of the lines is employed to filter out unwanted thermal sources, enabling the accurate segmentation and extraction of power transmission lines. This forms a solid foundation for subsequent detection and identification of abnormal hotspots on the extracted lines, and holds significance for the inspection and maintenance of existing thermal faults and potential hotspots in power transmission lines.

Estimation-free spatial-domain image reconstruction of structured illumination microscopy

Structured illumination microscopy (SIM) achieves super-resolution (SR) by modulating the high-frequency information of the sample into the passband of the optical system and subsequent image reconstruction. The traditional Wiener-filtering-based reconstruction algorithm operates in the Fourier domain, it requires prior knowledge of the sinusoidal illumination patterns which makes the time-consuming procedure of parameter estimation to raw datasets necessary, besides, the parameter estimation is sensitive to noise or aberration-induced pattern distortion which leads to reconstruction artifacts. Here, we propose a spatial-domain image reconstruction method that does not require parameter estimation but calculates patterns from raw datasets, and a reconstructed image can be obtained just by calculating the spatial covariance of differential calculated patterns and differential filtered datasets (the notch filtering operation is performed to the raw datasets for attenuating and compensating the optical transfer function (OTF)). Experiments on reconstructing raw datasets including nonbiological, biological, and simulated samples demonstrate that our method has SR capability, high reconstruction speed, and high robustness to aberration and noise.

A U-Net with Statistical Shape Restrictions Applied to the Segmentation of the Left Ventricle in Echocardiographic Images

This paper aims to introduce an innovative approach to semantic segmentation by leveraging a convolutional neural network (CNN) for predicting the shape and pose parameters of the left ventricle (LV). Our approach involves a modified U-Net architecture with a regression layer as the final stage, as opposed to the traditional classification layer. This modification allows us to predict all the shape and pose parameters of a statistical shape model, including rotation, translation, scale, and deformation. The adapted U-Net is trained using data from a point distribution model (PDM) of the LV. The experimental results demonstrate a mean Dice coefficient of 0.82 on good quality images, and 0.66 including mean and low-quality images. Our approach successfully overcomes a common issue encountered in CNN-based semantic segmentation. Unlike the inaccurate pixel classification that often leads to unwanted blobs, our CNN generates statistically valid shapes. These shapes hold significant potential in initializing other methods, such as active shape models (ASMs). Our novel CNN-based approach provides a novel solution for semantic segmentation, offering shapes and pose parameters that can enhance the accuracy and reliability of subsequent medical image analysis methods.