Digital Image Research Articles

Machine learning-based estimation of patient body weight from radiation dose metrics in computed tomography.

Currently, precise patient body weight (BW) at the time of diagnostic imaging cannot always be used for radiation dose management. Various methods have been explored to address this issue, including the application of deep learning to medical imaging and BW estimation using scan parameters. This study develops and evaluates machine learning-based BW prediction models using 11 features related to radiation dose obtained from computed tomography (CT) scans. A dataset was obtained from 3996 patients who underwent positron emission tomography CT scans, and training and test sets were established. Dose metrics and descriptive data were automatically calculated from the CT images or obtained from Digital Imaging and Communications in Medicine metadata. Seven machine-learning models and three simple regression models were employed to predict BW using features such as effective diameter (ED), water equivalent diameter (WED), and mean milliampere-seconds. The mean absolute error (MAE) and correlation coefficient between the estimated BW and the actual BW obtained from each BW prediction model were calculated. Our results found that the highest accuracy was obtained using a light gradient-boosting machine model, which had an MAE of 1.99kg and a strong positive correlation between estimated and actual BW (ρ=0.972). The model demonstrated significant predictive power, with 73% of patients falling within a±5% error range. WED emerged as the most relevant dose metric for BW estimation, followed by ED and sex. The proposed machine-learning approach is superior to existing methods, with high accuracy and applicability to radiation dose management. The model's reliance on universal dose metrics that are accessible through radiation dose management software enhances its practicality. In conclusion, this study presents a robust approach for BW estimation based on CT imaging that can potentially improve radiation dose management practices in clinical settings.

Retinal Image Dataset of Infants and Retinopathy of Prematurity

Retinopathy of prematurity (ROP) represents a vasoproliferative disease, especially in newborns and infants, which can potentially affect and damage the vision. Despite recent advances in neonatal care and medical guidelines, ROP still remains one of the leading causes of worldwide childhood blindness. The paper presents a unique dataset of 6,004 retinal images of 188 newborns, most of whom are premature infants. The dataset is accompanied by the anonymized patients’ information from the ROP screening acquired at the University Hospital Ostrava, Czech Republic. Three digital retinal imaging camera systems are used in the study: Clarity RetCam 3, Natus RetCam Envision, and Phoenix ICON. The study is enriched by the software tool ReLeSeT which is aimed at automatic retinal lesion segmentation and extraction from retinal images. Consequently, this tool enables computing geometric and intensity features of retinal lesions. Also, we publish a set of pre-processing tools for feature boosting of retinal lesions and retinal blood vessels for building classification and segmentation models in ROP analysis.

The Precision of All-on-Four Implant Position Recorded from Three Different CBCT Machines.

To investigate the dimensional discrepancy and degree of deviation of All-on-Four implant position between different cone-beam computed tomography (CBCT) machines. Four implants (4.5 × 10 mm Superline II, Dentium, South Korea) were placed in an All-on-Four style in an artificial mandible. The jaw was radiated 30 times using three different CBCT machines (Rainbow CT, Dentium; Veraview X800, Morita, Japan; Planmeca Viso G3, Planmeca OY, Finland). A total of 30 Digital Imaging and Communications in Medicine (DICOM) files were exported, n = 10. All-on-Four implants from each DICOM file were segmented and exported as an STL file (three-dimensional image) using Blue Sky Plan software (version 4.12.13/Blue Sky Bio, United States). All-on-Four implant zone dimensions (X, Y, and Z axes) and the total degree of deviation between All-on-Four implants per CBCT machine were measured using Autodesk Meshmixer software (version 3.5.474/California, United States). The data distribution's normality and variances' equality were tested with Shapiro-Wilk's and Levene's tests, respectively (p-value < 0.05). Data were analyzed using Brown-Forsythe one-way analysis of variance and Tamhane's post hoc tests to compare the differences between the groups (p-value <0.05). The respective X, Y, and Z mean dimensions of the All-on-Four implant zone were: Dentium (34.95, 14.71, and 9.97); Morita (34.88, 14.74, and 10.56); and Planmeca (34.73, 15.15, and 12.33). Significant differences between CBCT machines were found in all axes (p-value < 0.05); however, the Z-axis had the most differences. Notably, Planmeca exhibited the highest standard deviation (SD) in all axes (0.16-0.35), exhibiting the lowest consistency in the CBCT machines' readings. The Dentium exhibited the lowest deviation in the implant position, with the lowest SD (0.61). A significant difference in the total degree of deviation was spotted when only Morita was included in the comparison (p-value < 0.05). This study's findings are of significant importance as they reveal that the implant position recorded from the CBCT machines was most discrepant in the buccolingual dimension (Z-axis). Planmeca exhibited the least implant-dimensional accuracy of the CBCT machines, while Dentium exhibited the highest implant position accuracy. These results could significantly impact the choice of CBCT machine for implant placement, especially since an accurate CBCT image is crucial for digital implant planning.

Damage evolution and ductile fracture of commercially-pure titanium sheets subjected to simple tension and cyclic bending under tension

This paper describes a study into the evolution of damage in commercial-purity titanium (CP-Ti) sheets subjected to cyclic bending under tension (CBT) and uniaxial tension (simple tension, ST). Sections were taken from sheets that were strained to various levels under both methods and were imaged using X-ray computed tomography (XCT) to reveal insights into the size, density, and distribution of microvoids in the sheets. Digital image correlation (DIC) was also used to observe the surface strain of CBT and ST sheets during testing. These results showed that elongation to failure (ETF) for CBT deformation is about 2.5x higher than for ST, local longitudinal strains in the bulk of the CBT sample are around 1.3x higher than the peak strain in the ST necked region, and 1.7x higher in the localized region of CBT failure. It was found that the volume-density of voids in both sheets followed a similar exponential increase with strain, reaching nearly 45x higher in the CBT than the ST sheets before the onset of failure, mainly due to the higher strain levels achieved. A higher volume density of voids developed in the center of sheets at high levels of strain, for both processes, with the density in the sheet center becoming approximately double that found near the edges. Scanning electron microscopy (SEM) was used to examine the fracture surface of CBT and ST sheets after failure. The observations are presented and discussed highlighting the CBT process as effective to delay ductile fracture of the CP-Ti sheets.

Identification, uncertainty quantification and validation of orthotropic material properties for additively manufactured polymers

The mechanical behavior of polymeric parts manufactured by fused filament fabrication is often characterized as orthotropic because of the layer-wise extrusion of the material. To perform reliable simulations of the mechanical response of these parts subjected to certain loads, the material parameters have to be known. This leads to the task of material parameter identification from suitable experimental data. In this work, the identification of the material parameters for orthotropic material behavior is studied in detail with a focus on the specific characteristics of additively manufactured parts. This comprises working out the differences that arise compared to the parameter identification for classical orthotropic composite materials. Moreover, the material parameters are identified from tensile and shear test information together with full-field displacement data from digital image correlation measurements. Then, the obtained parameters are compared for analytical and numerical approaches, where a non-linear least-squares method with finite elements is drawn on. The parameters are not only identified, but also an uncertainty analysis is carried out and, subsequently, the different methods are validated. Further, a clear derivation of commonly applied analytical equations is provided with a special focus on the underlying assumptions. It turns out that the analytical parameter identification procedures can not be easily transferred to numerical procedures, especially when using full-field data, because of correlations between the parameters. Moreover, the identified parameters are employed to prove the reasonability of orthotropic material behavior for specimens manufactured by fused filament fabrication. Finally, during validation it turns out that all parameter identification methods provide a good agreement with experimental data, showing only small differences between the results of the different calibration strategies.

Dynamic bending analysis of metallized silicon nitride substrate during thermal cycling via digital image correlation

Dynamic bending analysis of metallized silicon nitride substrate during thermal cycling via digital image correlation

Evaluation method of tunnel cracking disease under biased pressure based on enhanced image fractal features

This paper focuses on the tunnel cracking disease under biased pressure. Through model tests on three lining conditions, the study investigates the expansion process of surface cracks on the lining under biased pressure. Proposing an improved method for calculating fractal characteristics of lining crack disease based on box counting dimension. Quantifying crack disease through digital image processing to obtain two parameters: total width and fractal dimension. Discussed the relationship between quantification parameters of surface crack disease in linings and the service performance of the lining. 29 groups of crack disease quantization parameters obtained by model test were selected to establish the disease sample space, and the spatial data were automatically learned and grouped based on clustering method. Finally, a new diagnostic Index of Tunnel Lining Defect Index-Crack & Dimension (TLDI-C&D) and its disease classification criteria for the whole lining ring are proposed. The results indicate that under the biased pressure, regardless of the presence of pre-existing defects, the biased pressure region is the most severely diseased area in terms of cracking. Compared to the intact lining, the initial cracking loads for arch crown defects and arch waist defects are advanced by 33.1 % and 73.0 %, respectively. After the apparent cracks in the lining are fully formed, pre-existing defects no longer affect the speed and distribution of crack propagation.The larger the fractal dimension, the more complex the morphology of the corresponding crack disease. Under the same image area, the ratio of crack quantity to width has a greater impact on the fractal dimension. There is a corresponding relationship between quantification parameters of crack disease and the bearing performance of the lining structure, making them indicators for measuring the load-bearing capacity of the lining. The most reasonable grading level for lining crack disease under the biased pressure is four levels. The diagnostic index for tunnel lining cracks, TLDI-C&D is 1.0～2.2 (level I), 2.2～3.2 (level II), 3.2～4.3 (level III), and 4.3～5.5 (level IV). On-site measurements have proven the accuracy of this crack disease diagnostic evaluation method.

Digital image correlation approach for low-cycle fatigue life monitoring of 13HMF power engineering steel

In this paper, strain evolution of 13HMF power engineering steel was analysed during force-controlled low cycle fatigue (LCF). The material performance under cyclic loading was monitored by using Digital Image Correlation (DIC) technique for different values of stress amplitude exceeding the yield strength of the material significantly. Data collected from DIC was compared to the extensometer ones to confirm the effectiveness of the proposed approach. Finally, the cloud data reflecting the fatigue performance of 13HMF was generated in the form of strain-stress amplitude - service life distribution map.

Source Separation and Classification using Generative Adversarial Networks and Weak Class Supervision

In this paper, we propose a decomposition-based weakly-supervised model that utilizes the class labels of the sources present in mixtures. We apply this weak class supervision approach to superimposed handwritten digit images using both non-negative matrix factorization (NMF) and generative adversarial networks (GANs). In this way, we can learn non-linear representations of the sources. The results of our experiments demonstrate that the proposed weakly-supervised methods are viable and mostly on par with the fully supervised baselines. The proposed joint classification and separation approach achieves a weakly-supervised source classification performance of 90.3 in terms of F1 score and outperforms the multi-label source classifier baseline of 68.2 when there are two sources. The separation performance of the proposed method is measured in terms of peak-signal-to-noise-ratio (PSNR) as 16 dB, outperforming the class-informed sparse NMF which achieves separation of two sources with a PSNR value of 13.9 dB. We show that it is possible to replace supervised training with weakly-supervised methods without performance penalty.

Multi-class Breast Cancer Classification Using CNN Features Hybridization

Breast cancer has become the leading cause of cancer mortality among women worldwide. The timely diagnosis of such cancer is always in demand among researchers. This research pours light on improving the design of computer-aided detection (CAD) for earlier breast cancer classification. Meanwhile, the design of CAD tools using deep learning is becoming popular and robust in biomedical classification systems. However, deep learning gives inadequate performance when used for multilabel classification problems, especially if the dataset has an uneven distribution of output targets. And this problem is prevalent in publicly available breast cancer datasets. To overcome this, the paper integrates the learning and discrimination ability of multiple convolution neural networks such as VGG16, VGG19, ResNet50, and DenseNet121 architectures for breast cancer classification. Accordingly, the approach of fusion of hybrid deep features (FHDF) is proposed to capture more potential information and attain improved classification performance. This way, the research utilizes digital mammogram images for earlier breast tumor detection. The proposed approach is evaluated on three public breast cancer datasets: mammographic image analysis society (MIAS), curated breast imaging subset of digital database for screening mammography (CBIS-DDSM), and INbreast databases. The attained results are then compared with base convolutional neural networks (CNN) architectures and the late fusion approach. For MIAS, CBIS-DDSM, and INbreast datasets, the proposed FHDF approach provides maximum performance of 98.706%, 97.734%, and 98.834% of accuracy in classifying three classes of breast cancer severities.

Interfacial cohesion model of ultra‐high performance concrete wet joints under the influence of multiple factors

AbstractThe interfacial cohesion between precast normal concrete (NC) and cast‐in‐place ultra‐high performance concrete (UHPC) is an important index to evaluate their interfacial bond strength, which is of great importance for the application of UHPC as a connection material for precast structural bridges. Interfacial cohesion is related to several influencing factors. However, there needs to be more research on the interrelationship model between multiple influencing factors and interfacial cohesion. This study took the UHPC wet joint in a precast concrete structure as the object. First, the relationship between the interface cohesion of UHPC wet joint and substrate strength, UHPC age, interfacial roughness, interfacial moisture content, and curing method was studied; The result shows that the key factors affecting interfacial cohesion include UHPC age, interface roughness, and substrate strength, with interfacial moisture content potentially playing a secondary role. Second, the failure types of the interface zone surface by using digital image correlation (DIC) are divided into three categories. Finally, the quantitative mathematical model of interfacial cohesion under the coupling effect of multiple factors was established based on the above four factors. The model is in good agreement with the experimental data.

Excellent accuracy of trained neonatal nurses in the detection of referral-warranted retinopathy of prematurity within an established telemedicine screening programme.

To evaluate level of agreement of specialist trained retinopathy of prematurity (ROP) nurses compared with an experienced paediatric ophthalmologist in detection of referral-warranted ROP (RWROP) using wide-field digital retinal imaging. This is a prospective, observational, blinded study of neonates in a level III neonatal intensive care unit, from July 2020 to November 2022. Image capture using wide-field digital retinal imaging followed by ROP grading and staging was completed by trained ROP nurses. This was then compared with findings by an experienced paediatric ophthalmologist. The primary outcome was presence of RWROP in either eye. One hundred and ninety-five neonates (55% male) with a total of 768 screening visits were included. At the initial screen, nurse and ophthalmologist agreed about presence of RWROP for 191 of 195 neonates (98%, kappa = 0.79, P < 0.0001), with 100% sensitivity for RWROP detection. Including all 768 screening episodes, agreement was 98% for RWROP. There was disagreement in 16 screenings (2%) for 11 (6%) neonates. Of the five screenings (0.7%) that the ophthalmologist thought were RWROP and the nurse did not, three were disagreements about whether the zone was posterior zone 2 or zone 1. We found excellent levels of agreement and add evidence that interpretations by specialist trained nurses could be safely integrated into a 'hybrid ROP screening system'.

Extraction and reduction of parameters from hand-drawn Archimedes spirals for clinical tremor assessment

BackgroundPatients’ hand-drawn Archimedes spirals are widely used in the neurological community to grade tremors. These spirals are either drawn on paper and Xeroxed/scanned into digital images or digitizing tablets are used for the drawings. This process introduces artifacts such as variable widths of the drawn lines with varying pixel grey scale values. Xeroxing introduces additional artifacts resulting from paper misalignments. These artifacts and the presence of the reference spiral in the image complicate an automatic extraction of a mathematical spiral signal from the image. New methodsWe introduce a mathematical mapping that transforms the image pixels of the patient's hand-drawn spiral into a one-dimensional discrete signal that can be used for mathematical analysis. ResultsA cohort of 18 hand-drawn spirals with various artifacts is used to validate our method.We extract the parameters of the discrete signals and show that the signals can be represented by truncating to as few as 150 parameters with a truncation RMS error of 6.26 % across the cohort. Using only 150 features makes machine learning a viable option for future applications. Furthermore, our method can be used to evaluate the frequency and the amplitude of the tremor. Comparison with existing methodsIn existing methods, the patient draws the spiral on a digitizing tablet, and features are extracted from this data for machine learning. We recognize that a vast majority of hospitals are still using the pencil-and-paper approach, and there is an abundance of ready-to-be-mined tremor-related data already stored as paper or digitized drawings. Our procedure is equally applicable to Xeroxed documents as well as files generated from digital tablets. ConclusionsWe have validated a new procedure requiring minimal user intervention to automatically extract a patient's hand-drawn spiral as a discrete mathematical one-dimensional signal from a scanned image or a file from a digital tablet.

Pre-Operative Planning of a DCR Surgery Using Virtual Reality

ABSTRACT Purpose To assess a novel Virtual Reality (VR) tool designed to enhance understanding of the nasal anatomy in patients eligible for DCR surgery. Methods Preoperative Computed Tomography (CT) scans of the orbit were obtained and loaded as DICOM (Digital Imaging and Communications in Medicine) files onto the D2P software (3D Systems Inc. Littleton, CO) for tissue segmentation and 3D model preparation. Segmentation was performed on several anatomical structures, including the skull, lacrimal sac, nasal septum, inferior and middle turbinate. The resulting 3D model was visualized using a VR headset. After completing the segmentation procedure, ten cases were evaluated by a panel of six surgeons, including both senior and resident physicians from ENT and oculoplastic specialties. Results The dataset under examination comprised images from 10 preoperative CT scans of the orbits of patients eligible for Endo-DCR. When evaluating the CT using the VR tool, in 73.3% of the cases ENT surgeons were right about the side of pathology, while only 43.3% ophthalmologists were right (chi-square, p = .018). In 72.8% of the cases ENT surgeons were evaluated right that there is a septum deviation, while only in 47.2% of the cases the ophthalmologists were right (chi-square, p = .094). When evaluating the CT using the VR tool, in 60% of the cases consultants were right about the pathology, while 57.7% of the residents were right (chi-square, p = .853). In 81.7% of the cases consultants were evaluated right that there is a septum deviation, while only in 58.3% of the cases the ophthalmologists were right (chi-square, p = .198). Discussion ENT surgeons, as well as consultants, interpreted the CT better than the ophthalmologists and residents. Surprisingly, the VR system did not help them to interpret the CT better. Further, more extensive studies should be done to build a VR system that assists in the correct interpretation of the preoperative CT before DCR surgery as well as during DCR surgery.

Experimental study of the fatigue failure behavior of aluminum alloy 2024-T351 under multiaxial loading

Multiaxial fatigue has been drawing much attention because it is more applicable to engineering practice. In this paper, an experimental study of aluminum alloy (AA) 2024-T351 was carried out under multiaxial loading with an aim to assess the damage evolution process and failure mechanism under in-phase and out-of-phase loadings. Firstly, fatigue life and stress response under multiaxial cyclic loads were obtained, and it found that although there is non-proportional hardening, the fatigue life subjected to proportional loading is significantly shorter than that of under a nonproportional loading, which was tried to be explained by the ductility of the material. Secondly, a detailed analysis of the damage evolution process based on the degradation of the elastic modulus and the fatigue failure process based on the digital image correlation (DIC) method was provided. Next, a micro-analysis of the specimens’ fracture appearance was conducted to obtain the fracture characteristics and found that AA2024-T351 presents a dominant shear fracture mode under proportional loading and a mixed mode of tensile and shear fracture under non-proportional loading. Last but not least, The Fatemi-Socie criterion was modified by considering the material’s ductility and the interaction between normal stress and shear stress acting on the critical plane. The multiaxial life prediction results of the modified FS model for AA2024-T351 in this paper were all within the scatter bands of 3 on life.

Characterization on fibre kinking fracture of laminated composites under combined compression and shear at high loading rate

This paper presented a novel method to characterize the fracture toughness and cohesive law of laminated composites under combined compression and shear at high loading rate. Compact compression specimens with off-axis fibres which introduce the in-plane shear stresses were conducted on the uniaxial bidirectional electromagnetic Hopkinson bar system and the displacement field and strain field were recorded by the high-speed camera for the digital image correlation analysis and J-integral calculation. The results reveal that the in-plane shear stresses bring the increase of fracture toughness at crack initiation and propagation and advance the damage initiation of the specimens. The fracture surfaces indicate that the shear stresses cause the fibre bundles' shear failure during the formation of the kink band, accompanied by more energy dissipation with the increase of off-axis angle.

Computer-aided diagnosis system for grading brain tumor using histopathology images based on color and texture features

BackgroundCancer pathology shows disease development and associated molecular features. It provides extensive phenotypic information that is cancer-predictive and has potential implications for planning treatment. Based on the exceptional performance of computational approaches in the field of digital pathogenic, the use of rich phenotypic information in digital pathology images has enabled us to identify low-level gliomas (LGG) from high-grade gliomas (HGG). Because the differences between the textures are so slight, utilizing just one feature or a small number of features produces poor categorization results.MethodsIn this work, multiple feature extraction methods that can extract distinct features from the texture of histopathology image data are used to compare the classification outcomes. The successful feature extraction algorithms GLCM, LBP, multi-LBGLCM, GLRLM, color moment features, and RSHD have been chosen in this paper. LBP and GLCM algorithms are combined to create LBGLCM. The LBGLCM feature extraction approach is extended in this study to multiple scales using an image pyramid, which is defined by sampling the image both in space and scale. The preprocessing stage is first used to enhance the contrast of the images and remove noise and illumination effects. The feature extraction stage is then carried out to extract several important features (texture and color) from histopathology images. Third, the feature fusion and reduction step is put into practice to decrease the number of features that are processed, reducing the computation time of the suggested system. The classification stage is created at the end to categorize various brain cancer grades. We performed our analysis on the 821 whole-slide pathology images from glioma patients in the Cancer Genome Atlas (TCGA) dataset. Two types of brain cancer are included in the dataset: GBM and LGG (grades II and III). 506 GBM images and 315 LGG images are included in our analysis, guaranteeing representation of various tumor grades and histopathological features.ResultsThe fusion of textural and color characteristics was validated in the glioma patients using the 10-fold cross-validation technique with an accuracy equals to 95.8%, sensitivity equals to 96.4%, DSC equals to 96.7%, and specificity equals to 97.1%. The combination of the color and texture characteristics produced significantly better accuracy, which supported their synergistic significance in the predictive model. The result indicates that the textural characteristics can be an objective, accurate, and comprehensive glioma prediction when paired with conventional imagery.ConclusionThe results outperform current approaches for identifying LGG from HGG and provide competitive performance in classifying four categories of glioma in the literature. The proposed model can help stratify patients in clinical studies, choose patients for targeted therapy, and customize specific treatment schedules.

Spot test and quantitative analysis of NTO explosive by smartphone-based digital image colorimetry

Spot test and quantitative analysis of NTO explosive by smartphone-based digital image colorimetry

Ballistic impact wave and bulge profile propagation characteristics and blunt injury assessment of UHMWPE laminate composite

Amid escalating localized conflicts worldwide, the elucidation of the mechanisms and quantification blunt trauma the emergency services and military personnel confronted becomes imperatively. However, the understanding of bullet-induced bulge formation on the laminate's back face, impact wave propagation, and consequent trauma assessment methods remains incomplete. This study applies 3-dimension digital image correlation (3D-DIC) method to reveals these phenomena by utilizing a lead-core bullet traveling at 334.93 m/s to impact an ultra-high molecular weight polyethylene (UHMWPE) laminate. The results show that the initial transverse propagation velocity of bulge profile is 24.79% smaller than the theoretical calculation, which indicates the reduction caused by cohesive matrix. The transverse wave demonstrates a double exponential attenuation pattern, which could be decomposed into two single exponential attenuation curves. The bulge profile shows hyperbolic pattern, which reveals the stability of the bulge shape and the temporal evolution pattern. Increasing the laminate-body gap leads to a decline in Blunt Criterion (BC) and the Abbreviated Injury Scale (AIS) values reduce from 2 to 0 correspondingly. The numerical model confirms the compression wave velocity as 1447.77 ± 123.10 m/s in the thickness direction and 9542.86 ± 3087.75 m/s in the fiber direction. These findings reveals and quantified the propagation patterns of back bulge and compressive wave, and thus could provide data-driven insights to improve body armor performance and blunt trauma assessment for individual soldiers.

Role of slip in hydrogen-assisted crack initiation in Ni-based alloy 725.

We conduct in situ tensile straining experiments to investigate the role of hydrogen and slip in crack initiation in nickel-based alloy 725. Our experiments reveal no tendency for hydrogen to enhance localized slip and no necessity of slip for crack initiation. We use electrochemical charging to introduce hydrogen into samples, melt extraction to measure hydrogen content, and digital image correlation to analyze localized plastic strains during in situ tensile tests in a scanning electron microscope. Cracks initiate both in regions with and without nearby localized slip. Moreover, the fraction of cracks initiating with no nearby slip is greater at higher hydrogen content. Slip-assisted crack initiation generally occurs at locations where intergranular slip is arrested, especially at intersections of slipping coherent twin boundaries with thin twin lamellae. Cracks that initiate without nearby slip occur at a wider variety of microstructural features, including inclusions, triple junctions, and surface flaws.