3D Surface Models Research Articles

A feasibility study of three-dimensional ultrasound imaging of the vagina under distension.

The distension properties of the vagina are critical to its function including support of surrounding organs, childbirth, and intercourse. It could be altered by many pathophysiological processes like pregnancy, radiotherapy, and reconstruction surgery. However, there are no clinically available diagnostic tools capable of quantifying the distension properties of the vagina. A proof-of-concept study was designed to assess the feasibility of a novel three-dimensional (3D) ultrasound imaging technique that allows quantitative evaluation of the vagina under distension. Patients with symptomatic pelvic organ prolapse (POP) were recruited for the study. An ultrathin, oversized bag was inserted into the vagina and filled with water using a modified urodynamics system. The instilled water volume and intravaginal pressure were continuously recorded. At maximum vaginal capacity, 3D transintroital ultrasound of the distended vagina and surrounding pelvic structures was performed. Exams were performed in duplicate for each patient, two hours apart (round A and round B). Following the development of a 3D surface model of the distended vagina from each scan, several measurements were obtained, including cross-sectional area, anteroposterior (AP) length and lateral width in the plane of minimum hiatal dimensions (PMHD), AP and lateral diameter at the pubic symphysis (PS) level, maximum and minimum diameter, and maximum vertical length. To assess repeatability between measurements in two rounds, the coefficient of variation (CV) and the intraclass correlation coefficient (ICC) were calculated for each measurement. Correlations between physical measurements including the pelvic organ prolapse quantification (POP-Q) system and vaginal diameter measurements, and obtained metrics were also assessed. Sixteen patients with POP (average age 69 years) completed both rounds of imaging. There was sufficient echogenicity on 3D transintroital ultrasound of the distended vaginal wall to establish boundaries for 3D surface models of the vagina. Overall, all metrics had good or excellent reliability (ICC=0.77-0.93, p<0.05; CV=3%-18%) except maximum diameter, which demonstrated only moderate reliability (ICC=0.67, p=0.092). Strong correlations were found between physical exam measurements including D point of POP-Q, introitus diameter and lateral diameter at apex, and maximum vaginal capacity, maximum vertical length, lateral diameter at PS, minimum diameter, and distended PMHD measurements. The results demonstrated that this system could generate 3D models of the shape of the distended vagina and provide multiple metrics that could be reliably calculated from automated analyses of the models. A novel system for evaluation of the distension properties of the vagina was developed and preliminary evaluation was performed. This system may represent a technique for evaluation of the biomechanical and structural properties of the vagina.

Asymmetry in Class II subdivision malocclusion: Assessment based on 3D surface models.

There is currently no consensus in the literature whether the aetiology of a Class II subdivision is dental, skeletal or both. The aim of this study was to identify and quantify skeletal and dental asymmetries in Class II subdivision malocclusions. CBCTs from 33 Class II subdivision malocclusion patients were used to construct 3D volumetric label maps. Eighteen landmarks were identified. The original scan and associated 3D volumetric label map were mirrored. Registration of the original and mirrored images relative to the anterior cranial base, maxilla and mandible were performed. Surface models were generated, and 3D differences were quantified. Statistical analysis was performed. Anterior cranial base registration showed significant differences for fossa vertical difference, fossa roll, mandibular yaw, mandibular lateral displacement and lower midline displacement. Regional registrations showed significant differences for antero-posterior (A-P) mandibular length, maxillary roll, A-P maxillary first molar position, maxillary first molar yaw and maxillary first molar roll. Class II subdivision patients also show an asymmetric mandibular length as well as an asymmetric gonial angle. Moderate correlations were found between the A-P molar relationship and fossa A-P difference, mandibular first molar A-P difference, maxillary first molar A-P difference and maxillary first molar yaw. This study suggests that Class II subdivisions can result from both significant skeletal and dental factors. Skeletal factors include a shorter mandible as well as posterior and higher displacement of the fossa on the Class II side, resulting in mandibular yaw. Dental factors include maxillary and mandibular first molar antero-posterior asymmetry.

Low-cost, smartphone-based instant three-dimensional registration system for infant functional near-infrared spectroscopy applications.

To effectively apply functional near-infrared spectroscopy (fNIRS)/diffuse optical tomography (DOT) devices, a three-dimensional (3D) model of the position of each optode on a subject's scalp and the positions of that subject's cranial landmarks are critical. Obtaining this information accurately in infants, who rarely stop moving, is an ongoing challenge. We propose a smartphone-based registration system that can potentially achieve a full-head 3D scan of a 6-month-old infant instantly. The proposed system is remotely controlled by a custom-designed Bluetooth controller. The scanned images can either be manually or automatically aligned to generate a 3D head surface model. A full-head 3D scan of a 6-month-old infant can be achieved within 2s via this system. In testing on a realistic but static infant head model, the average Euclidean error of optode position using this device was 1.8mm. This low-cost 3D registration system therefore has the potential to permit accurate and near-instant fNIRS/DOT spatial registration.

A novel method for predicting the emergence of toxicity interaction in ternary mixtures

The environment is teeming with a wide variety of pollutants, but the complexity and diversity of their combinations make it difficult to fully assess their toxicity interaction. A novel toxicity interaction prediction method (TIPM) based on the three-dimensional (3D) surface form of the concentration addition (CA) deviation model (dCA) was proposed to predict the emergence of toxicity interaction in ternary mixtures. Doxycycline hyclate (DH), bromoacetic acid (BAA) and iodoacetic acid (IAA) were used as target pollutants. The toxicity of binary and ternary mixtures designed by the direct equipartition ray design method (EquRay) and the uniform design ray method (UD-Ray) against Escherichia coli (E. coli) was determined by using a time-dependent microplate toxicity analysis (t-MTA) method. The toxicity interaction within mixtures was characterized qualitatively and quantitatively using dCA 3D surface modeling and the emergence of DH-MAA-IAA toxicity interaction was predicted by TIPM. The results showed that the dCA 3D surface model could well characterize the toxicity interactions of the mixtures, and toxicity interaction was closely related to the components’ concentration ratio (pi). TIPM could predict the emergence of DH-MAA-IAA toxicity interactions well based on the relationship. Due the model is only related to the toxicity interactions and pi value of a mixture, so it can be suggested to predict toxicity interaction within the more complex multicomponent mixtures, which provides a novel approach for the environmental risk assessment and prediction of hazardous substances.

Three-Dimensional Printing of Models of Patellofemoral Joint Articular Cartilage in Patients With Patella Instability for Observing Joint Congruity

Three-dimensional (3D) modeling and printing are increasingly used in the field of orthopaedic surgery for both research and patient care. One area where they are particularly helpful is in improving our understanding of the patellofemoral (PF) joint. Heretofore, morphological studies that use 3D models of the PF joint have primarily been based on computed tomography imaging data and thus do not incorporate articular cartilage. Here, we describe a method for creating 3D models of the articular surfaces of the PF joint based on magnetic resonance imaging. Models created using this technique can be used to improve our understanding of the morphology of the articular surfaces of the PF joint and its relationship to joint pathologies. Of particular interest is our finding of articular congruity in printed articular cartilage surfaces of dysplastic PF joints of recurrent patella dislocators.

Interpretable artificial intelligence for classification of alveolar bone defect in patients with cleft lip and palate

Cleft lip and/or palate (CLP) is the most common congenital craniofacial anomaly and requires bone grafting of the alveolar cleft. This study aimed to develop a novel classification algorithm to assess the severity of alveolar bone defects in patients with CLP using three-dimensional (3D) surface models and to demonstrate through an interpretable artificial intelligence (AI)-based algorithm the decisions provided by the classifier. Cone-beam computed tomography scans of 194 patients with CLP were used to train and test the performance of an automatic classification of the severity of alveolar bone defect. The shape, height, and width of the alveolar bone defect were assessed in automatically segmented maxillary 3D surface models to determine the ground truth classification index of its severity. The novel classifier algorithm renders the 3D surface models from different viewpoints and captures 2D image snapshots fed into a 2D Convolutional Neural Network. An interpretable AI algorithm was developed that uses features from each view and aggregated via Attention Layers to explain the classification. The precision, recall and F-1 score were 0.823, 0.816, and 0.817, respectively, with agreement ranging from 97.4 to 100% on the severity index within 1 group difference. The new classifier and interpretable AI algorithm presented satisfactory accuracy to classify the severity of alveolar bone defect morphology using 3D surface models of patients with CLP and graphically displaying the features that were considered during the deep learning model's classification decision.

Automated Determination of the Volume of Loose Engineering Deposits Using Terrestrial Laser Scanning

A new method for automatic volume determination of loose engineering deposits (LEDs) using point clouds collected by terrestrial laser scanning is proposed. The method starts with a problem of lacking bottom surface point clouds when scanning LEDs, assuming that the bottom surface is flat, a spatial plane is generated based on the plane fitting to the bottom surface. Then the sample point cloud is projected on this fitted plane to construct the subface point clouds, which, together with the initial point cloud data, are stitched into complete point cloud data with good closure, integrity, and accuracy. The 3D surface model established by introducing the alpha shape algorithm calculates the volume, and the validation experiments on three soil mound models and one field experiment on LEDs with a minimum average error of only 1.69% fully validate the effectiveness and accuracy of the method. The encouraging results show that effective volume calculation can be performed on realistic LEDs using the proposed method.

Evaluating The Approaches of the 3D Models Generation of a Soil Surface: A Review Study

There are various approaches used to create 3D models of soil surfaces, including field surveying, photogrammetry and laser scanning methods. However, most of these methods are high cost, requiring many employees and long processing time. Thus, this study will evaluate these approaches and will show the merits and the significances of each approach used for 3D soil generations. The review study shows that the videogrammetric method is considered the efficient approach for 3D soil surface generation because it is economic, precise, fast, and needs few workers.

ROCK MASS DISCONTINUITY DETERMINATION WITH TRANSFER LEARNING

Abstract. Rock mass discontinuity and orientation are among the important rock mass features. They are conventionally determined with scan-line surveys by engineering geologists in field, which can be difficult or impossible depending on site accessibility. Photogrammetry and computer vision techniques can aid to automatically perform these measurements, although variations in size, shape and appearance of rock masses make the task challenging. Here we propose an automated approach for the detection of rock mass discontinuities using deep learning and photogrammetric image processing methods. Two deep convolutional neural network (DCNNs) were implemented for this purpose and applied to basalts in Kizilcahamam Guvem Geosite near Ankara, Türkiye. Red-green-blue (RGB) band images of the site were taken from an off-the-shelf camera with 1.7 mm resolution and a 3D digital surface model and orthophotos were produced by using photogrammetric software. The discontinuities were delineated manually on the orthophoto and converted to masks. The first DCNN model was based on the open-source crack dataset consisting of a total of 11,298 road and pavement images, which were used to train the Resnet-18 model (Model-1). The second model (Model-2) was based on fine-tuning of Model-1 using the study data from Kizilcahamam. After fine-tuning, Model-2 was able to achieve high performance with a Jaccard Score of 88% on the test data. The results show high potential of the methodology for transfer learning with fine-tuning of a small amount of data that can be applied to other sites and rock mass types as well.

Integrated Information System for 3D Interactive Reconstruction of an Archaeological Site

Archaeological recording is intended to preserve as much information as possible about the finds. However, once the pieces are removed from the site, there is information regarding the original positioning of these pieces that may be lost or not accurately recorded and can be relevant for further studies. This spatial arrangement can also be crucial for subsequent piece restoration or to understand certain aspects of ancient cultures. In this article, we describe a software prototype and a methodology to virtually reconstruct an archaeological site for posterity, once it has been excavated. The system is implemented with a client-server architecture. In the server, a spatial database stores and manages the three-dimensional (3D) models of the finds, as well as several 3D site ground surface models acquired at different times during the excavation process. On the client side, a graphical interface allows the user to manipulate the find models to re-create and virtually reconstruct the original spatial arrangement of the archaeological site. Topological relationships among the finds are stored in the database to provide further spatial analysis. The result is an integrated information system that goes beyond 3D visualization, making the site last for posterity after its excavation and allowing further spatial analysis.

Faulting, sediment loading, and flow of underlying ductile units: A case study from the Western Ionian Basin Offshore Eastern Sicily

Tectono-stratigraphic analysis coupled with digital 3D surface modelling derived from high-resolution seismic profiles is performed along a narrow turbidite basin offshore E-Sicily in order to increase understanding on the processes that contributed to the shaping of the Western Ionian Basin. Seismic-reflector patterns of the identified Pliocene-Quaternary sequence point to syn-depositional deformation during the Pliocene associated with the simultaneous activity of regional faults and underlying ductile units. Long-wavelength sediment fanning results from the extensional activity of the Malta Escarpment faults. Conversely, internal reflector architecture and lateral terminations indicate localized subsidence associated with the growth of uprising structures in the easternmost part of the basin. Lateral shifting of basin depocenters is in line with withdrawal effects observed in basins floored by ductile units (salt or shale). 3D modelling of time-reference surfaces highlights sub-circular depressions associated with nearby structural culminations. This pattern is similar to salt-withdrawal minibasins commonly reported in evaporite-floored basins. Accordingly, salt migration/flow triggered by sediment loading, locally enhanced by fault activity, is proposed as the process controlling basin evolution during the Pliocene in the Western Ionian Domain. Nevertheless, the possibility of shale/mud tectonics as the ductile source of deformation cannot be discounted.

Modeling of the Three-Dimensional Point Cloud Based on the Binocular Stereovision and Structure Identification

Abstract The existence of weak planes affects the stability of rock engineering, especially the failure modes of the surrounding rocks. Therefore, investigation of the structural planes of rock mass and the extraction of discontinuities is important for rock mechanics. In this work, the LenaCV binocular stereovision system was studied and applied to establish three-dimensional (3D) surface models. Calibration of the binocular camera parameters, optimal distance discussion, and 3D modeling optimization were conducted. The optimal measuring distance and the intrinsic parameters suitable for the LenaCV stereovision were discussed. Finally, the 3D reconstruction of two case studies was carried out and the 3D point cloud modeling was established. The structural discontinuity mapping in rock faces was also carried out. The feasibility of the binocular stereovision system, as well as the optimum baseline distance, was discussed and verified. The study results can provide a new platform for the economic investigation of 3D rock mass structures and provide a new approach for the efficient extraction of rock mass discontinuities.

Cardiac Pulsatile Helical Deformation of the Thoracic Aorta Before and After Thoracic Endovascular Aortic Repair of Type B Dissections.

Type B aortic dissections propagate with either achiral (nonspiraling) or right-handed chiral (spiraling) morphology, have mobile dissection flaps, and are often treated with thoracic endovascular aortic repair (TEVAR). We aim to quantify cardiac-induced helical deformation of the true lumen of type B aortic dissections before and after TEVAR. Retrospective cardiac-gated computed tomography (CT) images before and after TEVAR of type B aortic dissections were used to construct systolic and diastolic 3-dimensional (3D) surface models, including true lumen, whole lumen (true+false lumens), and branch vessels. This was followed by extraction of true lumen helicity (helical angle, twist, and radius) and cross-sectional (area, circumference, and minor/major diameter ratio) metrics. Deformations between systole and diastole were quantified, and deformations between pre- and post-TEVAR were compared. Eleven TEVAR patients (59.9±4.6 years) were included in this study. Pre-TEVAR, there were no significant cardiac-induced deformations of helical metrics; however, post-TEVAR, significant deformation was observed for the true lumen proximal angular position. Pre-TEVAR, cardiac-induced deformations of all cross-sectional metrics were significant; however, only area and circumference deformations remained significant post-TEVAR. There were no significant differences of pulsatile deformation from pre- to post-TEVAR. Variance of proximal angular position and cross-sectional circumference deformation decreased after TEVAR. Pre-TEVAR, type B aortic dissections did not exhibit significant helical cardiac-induced deformation, indicating that the true and false lumens move in unison (do not move with respect to each other). Post-TEVAR, true lumens exhibited significant cardiac-induced deformation of proximal angular position, suggesting that exclusion of the false lumen leads to greater rotational deformations of the true lumen and lack of true lumen major/minor deformation post-TEVAR means that the endograft promotes static circularity. Population variance of deformations is muted after TEVAR, and dissection acuity influences pulsatile deformation while pre-TEVAR chirality does not. Description of thoracic aortic dissection helical morphology and dynamics, and understanding the impact of thoracic endovascular aortic repair (TEVAR) on dissection helicity, are important for improving endovascular treatment. These findings provide nuance to the complex shape and motion of the true and false lumens, enabling clinicians to better stratify dissection disease. The impact of TEVAR on dissection helicity provides a description of how treatment alters morphology and motion, and may provide clues for treatment durability. Finally, the helical component to endograft deformation is important to form comprehensive boundary conditions for testing and developing new endovascular devices.

3D analysis of maxillomandibular morphology in hyperdivergent and hypodivergent individuals: A cross-sectional study.

The aim of this cross-sectional study was to investigate maxillomandibular morphology in hyperdivergent and hypodivergent individuals, using 3D surface models generated by cone-beam computed tomography (CBCT). The sample consisted of 60 CBCTs (30 males, 30 females) patients aged 12-30 years, divided into two groups comprising hyperdivergent (≥35°) and hypodivergent (≤30°) individuals, according to the mandibular plane (MP) angle. Multiplanar reconstructions were used to mark the landmarks, and 3D surface models were created to evaluate structures of the maxillomandibular complex, including condyle, ramus, symphysis and palatal height. Intergroup comparisons were performed by independent t-test. Pearson's correlation test was used (P < .05) to evaluate the correlation of the MP angle with the angles and linear measurements of other structures. Significant differences were found between the groups regarding condylar width, ramus height, condylar plus ramus height, mandibular length, gonial angle, palatal plane angle and palatal-mandibular angle. No differences (P > .05) were found for the condylar height, symphysis inclination angle or palatal height. Correlations (P < .05) were found between the MP angle and structures of the maxillomandibular complex. Hyperdivergent (MP ≥ 35°) and hypodivergent (MP ≤ 30°) individuals present different skeletal morphology regarding condylar width, ramus height, condylar plus ramus height, mandibular length, gonial angle, palatal plane angle and palatal-mandibular angle. There is a significant correlation between MP angle and morphological structures such as condyle, ramus, symphysis, palatal plane angle and palatal-mandibular angle.

Automatic segmentation and quantification of the optic nerve on MRI using a 3D U-Net.

Pathological conditions associated with the optic nerve (ON) can cause structural changes in the nerve. Quantifying these changes could provide further understanding of disease mechanisms. We aim to develop a framework that automatically segments the ON separately from its surrounding cerebrospinal fluid (CSF) on magnetic resonance imaging (MRI) and quantifies the diameter and cross-sectional area along the entire length of the nerve. Multicenter data were obtained from retinoblastoma referral centers, providing a heterogeneous dataset of 40 high-resolution 3D T2-weighted MRI scans with manual ground truth delineations of both ONs. A 3D U-Net was used for ON segmentation, and performance was assessed in a tenfold cross-validation () and on a separate test-set () by measuring spatial, volumetric, and distance agreement with manual ground truths. Segmentations were used to quantify diameter and cross-sectional area along the length of the ON, using centerline extraction of tubular 3D surface models. Absolute agreement between automated and manual measurements was assessed by the intraclass correlation coefficient (ICC). The segmentation network achieved high performance, with a mean Dice similarity coefficient score of 0.84, median Hausdorff distance of 0.64mm, and ICC of 0.95 on the test-set. The quantification method obtained acceptable correspondence to manual reference measurements with mean ICC values of 0.76 for the diameter and 0.71 for the cross-sectional area. Compared with other methods, our method precisely identifies the ON from surrounding CSF and accurately estimates its diameter along the nerve's centerline. Our automated framework provides an objective method for ON assessment in vivo.

Artificial Intelligence Applications in Orthodontics

Objective This manuscript describes strategies for assessment of precision of several diagnostic artificial intelligence (AI) tools in orthodontics, available open-source image analysis platforms, as well as the use of three-dimensional (3D) surface models and superimpositions.Results The advances described in this manuscript present perspectives on the controversies of whether AI is smarter than clinicians and may replace human clinical decisions. A thorough orthodontic diagnosis requires comprehensive 3D analysis of the interrelationships among the dentition, craniofacial skeleton and soft tissues. Forecasts have indicated that 3D printing technology will provide more than 60% of all dental treatment needs by 2025, and orthodontic companies as well as remote monitoring companies are already using AI technology, being it essential that the clinicians are prepared and knowledgeable with the technology advances now available.Conclusions The AI applications in orthodontics rely on the implementation into diagnostic image records, data analysis for clinical practice and research applications. Continuous training and validation of the AI orthodontic image tools are essential for improving the performance and generalizability of these methods.

LatentPCN: latent space-constrained point cloud network for reconstruction of 3D patient-specific bone surface models from calibrated biplanar X-ray images.

Accurate three-dimensional (3D) models play crucial roles in computer assisted planning and interventions. MR or CT images are frequently used to derive 3D models but have the disadvantages that they are expensive or involving ionizing radiation (e.g., CT acquisition). An alternative method based on calibrated 2D biplanar X-ray images is highly desired. A point cloud network, referred as LatentPCN, is developed for reconstruction of 3D surface models from calibrated biplanar X-ray images. LatentPCN consists of three components: an encoder, a predictor, and a decoder. During training, a latent space is learned to represent shape features. After training, LatentPCN maps sparse silhouettes generated from 2D images to a latent representation, which is taken as the input to the decoder to derive a 3D bone surface model. Additionally, LatentPCN allows for estimation of a patient-specific reconstruction uncertainty. We designed and conducted comprehensive experiments on datasets of 25 simulated cases and 10 cadaveric cases to evaluate the performance of LatentLCN. On these two datasets, the mean reconstruction errors achieved by LatentLCN were 0.83mm and 0.92mm, respectively. A correlation between large reconstruction errors and high uncertainty in the reconstruction results was observed. LatentPCN can reconstruct patient-specific 3D surface models from calibrated 2D biplanar X-ray images with high accuracy and uncertainty estimation. The sub-millimeter reconstruction accuracy on cadaveric cases demonstrates its potential for surgical navigation applications.

UAV-based RGB and TIR imaging for geothermal monitoring: a case study at Kestanbol geothermal field, Northwestern Turkey.

Kestanbol is one of the most important geothermal fields in NW Turkey. This study conducted the first-ever surveys over a 10ha reach of the Kestanbol geothermal field using an unmanned aerial vehicle (UAV) equipped with visible (RGB) and thermal infrared (TIR) cameras. Low-altitude flights below 40m above the ground were operated above the Kestanbol geothermal field. Approximately 3500 RGB and TIR images were captured using the UAV. We recorded high-resolution RGB and TIR data of the Kestanbol geothermal field and applied the structure from motion (SfM) algorithm to identify the distribution of geothermal springs and seeps. The Kestanbol geothermal field was monitored to create a georeferenced RGB orthophoto, RGB 3D surface model, thermal anomaly map, and digital surface model (DSM) of the area with centimeter-level accuracy. In the TIR orthophoto, the surface temperature in the geothermal field was found to be between 15 and 75°C. All the thermal anomalies revealed by the survey were verified by field observations. The geothermal springs and seeps were parallel to the NE-SW regional tectonic trends. The results of this study demonstrate an effective technique for monitoring and assessing geothermal water using UAV-based RGB and TIR imaging and provide an accurate basis for geothermal development projects. RGB and TIR imaging using UAVs are considered promising methods for improving the assessment of the effects of geothermal water on the environment.

3D printed plates based on generative design biomechanically outperform manual digital fitting and conventional systems printed in photopolymers in bridging mandibular bone defects of critical size in dogs.

Conventional plate osteosynthesis of critical-sized bone defects in canine mandibles can fail to restore former functionality and stability due to adaption limits. Three-dimensional (3D) printed patient-specific implants are becoming increasingly popular as these can be customized to avoid critical structures, achieve perfect alignment to individual bone contours, and may provide better stability. Using a 3D surface model for the mandible, four plate designs were created and evaluated for their properties to stabilize a defined 30 mm critical-size bone defect. Design-1 was manually designed, and further shape optimized using Autodesk ® Fusion 360 (ADF360) and finite element analysis (FE) to generate Design-2. Design-4 was created with the generative design (GD) function from ADF360 using preplaced screw terminals and loading conditions as boundaries. A 12-hole reconstruction titanium locking plate (LP) (2.4/3.0 mm) was also tested, which was scanned, converted to a STL file and 3D printed (Design-3). Each design was 3D printed from a photopolymer resin (VPW) and a photopolymer resin in combination with a thermoplastic elastomer (VPWT) and loaded in cantilever bending using a customized servo-hydraulic mechanical testing system; n = 5 repetitions each. No material defects pre- or post-failure testing were found in the printed mandibles and screws. Plate fractures were most often observed in similar locations, depending on the design. Design-4 has 2.8-3.6 times ultimate strength compared to other plates, even though only 40% more volume was used. Maximum load capacities did not differ significantly from those of the other three designs. All plate types, except D3, were 35% stronger when made of VPW, compared to VPWT. VPWT D3 plates were only 6% stronger. Generative design is faster and easier to handle than optimizing manually designed plates using FE to create customized implants with maximum load-bearing capacity and minimum material requirements. Although guidelines for selecting appropriate outcomes and subsequent refinements to the optimized design are still needed, this may represent a straightforward approach to implementing additive manufacturing in individualized surgical care. The aim of this work is to analyze different design techniques, which can later be used for the development of implants made of biocompatible materials.

Automated pothole condition assessment in pavement using photogrammetry-assisted convolutional neural network

ABSTRACT Automated detection of pavement distress can prevent deterioration of premature surface disintegration in pavements. Potholes that are a common sight in harsh and cold terrains are a severe threat to road safety and a major contributing factor to pavement distress. To facilitate timely detection and repair of potholes, a computationally light and feasible, intelligent pavement pothole detection system is proposed by developing a novel workflow for image-based detection and severity assessment. A single-stage CNN architecture, RetinaNet is modified and optimised to best detect potholes and used in combination with a novel pothole depth estimation algorithm. A comparative evaluation of the model’s performance against the existing state-of-the-art model on the benchmark dataset establishes the proposed model’s high performance and applicability in real-time scenarios. The depth estimation algorithm is based on a 3D road surface model generated by employing the photogrammetric process of structure from motion (SfM). The point cloud data obtained thereafter, is used for accurate measurement of pothole depth. The comparison of the derived depth with the onsite depth measurement of the pothole reveals a mean error below 5%. This method leads to a practical and intelligent solution to be implemented as part of a potential pavement health assessment system for future practice.