Skeletal Feature Research Articles

Age estimation using medial clavicle by histomorphometry method with artificial intelligence: A review.

This review research critically assesses the evolving landscape of age estimation methodologies, with a particular focus on the innovative integration of histomorphometry and artificial intelligence (AI) in the analysis of the medial clavicle. The medial clavicle emerges as a crucial skeletal feature for predicting age, offering valuable insights into the morphological changes occurring throughout an individual's lifespan. Through an in-depth exploration of histological complexities, including variations in osteons, trabecular structures, and cortical thickness, this review elucidates their utility as viable indicators for age-related evaluations. This framework is augmented by the incorporation of AI technology, which enables automatic picture identification, feature extraction, and complicated pattern analysis. Our review of previous research highlights the promise of AI in improving prediction models for nuanced age estimates, highlighting the importance of large-scale, diversified datasets and thorough cross-validation. This thorough study, which addresses ethical concerns as well as the influence of population-specific characteristics, moves the debate around age estimate ahead, presenting insights with consequences for forensic anthropology, clinical diagnoses, and future research avenues.

Skeletal muscle feature of different populations in large yellow croaker (Larimichthys crocea): from an epigenetic point of view.

Fish skeletal muscle is composed of well-defined fiber types. In order to identify potential candidate genes affecting muscle growth and development under epigenetic regulation. Bisulfite sequencing was utilized to analyze and compare the muscle DNA methylation profiles of Larimichthys crocea inhabiting different environments. The results revealed that DNA methylation in L. crocea was predominantly CG methylation, with 2,396 differentially methylated regions (DMRs) identified through comparisons among different populations. The largest difference in methylation was observed between the ZhouShan and JinMen wild populations, suggesting that L. crocea may have undergone selection and domestication. Additionally, GO and KEGG enrichment analysis of differentially methylated genes (DMGs) revealed 626 enriched GO functional categories, including various muscle-related genes such as myh10, myf5, myf6, ndufv1, klhl31, map3k4, syn2b, sostdc1a, bag4, and hsp90ab. However, significant enrichment in KEGG pathways was observed only in the JinMen and XiangShan populations of L. crocea. Therefore, this study provides a theoretical foundation for a better understanding of the epigenetic regulation of skeletal muscle growth and development in L. crocea under different environmental conditions.

An Overview of Behavioral Recognition

Human behavior recognition has become a popular research topic in the field of computer vision. With the introduction of deep learning and attention mechanisms, this field has been further promoted. However, issues such as dataset acquisition and preprocessing operations on multimodal datasets, modeling of long time information in videos, and fusion of temporal and spatial information still exist. In this paper, we first outline some video action recognition datasets and related preprocessing techniques, including frame extraction, optical flow extraction, and skeletal feature acquisition. Then, the relevant models are classified and parsed according to their characteristics and the types of input data modalities. In addition, we evaluate the performance of the models on several benchmark datasets to gain a deeper understanding of the model development process. Finally, we summarized the current challenges faced in the field of video behavior recognition, including model timeliness, data set subjectivity and effective fusion of multi-modal features, and proposed possible future improvement directions in order to provide more ideas and methods for subsequent research.

Artificial Intelligence Enabled Teaching Optimization for General Studies Elective Courses

Abstract Network general knowledge course is a meaningful way to cultivate comprehensive quality talents in Chinese higher education based on the modern information technology environment, and the optimization of network general knowledge course teaching is directly related to whether the network general knowledge course can be effectively implemented, which affects the overall effect of talent cultivation. In this paper, starting from the purpose and goal of humanities general education courses, the optimization direction of teaching general education courses using artificial intelligence is clarified. Skeletal sequence feature extraction is used to obtain students’ general education classroom behavior data, and a stackable multimodal attention module consisting of a self-attention layer and a multimodal co-attention layer is designed for feature fusion and action prediction based on multimodal characteristics. After data integration and processing, the HGCL-LG network model is proposed to complete the diagnosis of students’ general knowledge cognition. The trend of interest assessment and multimodal model calculation value is consistent. In contrast, the student’s interest attention in general knowledge classroom above 0.33 accounts for more than 40% of the overall information, and the interest attention is good. Students’ scores on the objective questions of general knowledge are good; the average score range is between 1.5 and 2 points, the degree of dispersion is small, and the mastery of basic knowledge of general knowledge is good.

Static and dynamic validation of kinect for ergonomic postural analysis using electro-goniometers as a gold standard:A preliminary study.

Evaluation of the working postures and development of new techniques are paramount in reducing the awkward postures and occurrence of musculoskeletal disorders (MSDs). The Kinect sensor, a portable and cost-effective device, appears to be a promising alternative to study work postures. The current study aimed to evaluate the validity of Kinect against the gold-standard instrument (electro-goniometers) for body joint angle measurements. A unique software application was developed to measure the critical body joint angles for postural evaluation by using the Kinect's skeletal tracking feature. The body joint angle data of ten volunteers were measured simultaneously by both Kinect and electro-goniometers. The validation analysis was conducted in both static and dynamic domains of application. Minimal variation was observed between the two techniques, and the Kinect correlated well for upper-arm joint angles of 45∘, 60∘ and 90∘; lower-arm joint angles of 30∘, 45∘, 60∘, and 90∘; straight neck position, neck joint angle at maximum possible flexion; straight trunk position, trunk bend angle at full flexion. In dynamic analysis, four out of five ICC values were > 0.75 except for the upper arm. Discrepancies in the results indicated the disapproval of Kinect for only wrist measurements. The results of the static and dynamic studies gave a sufficient basis to consider the Kinect tool as an alternative to contemporary posture-based ergonomic evaluation methods.

Worker Abnormal Behavior Recognition Based on Spatio-Temporal Graph Convolution and Attention Model

In response to the problem where many existing research models only consider acquiring the temporal information between sequences of continuous skeletons and in response to the lack of the ability to model spatial information, this study proposes a model for recognizing worker falls and lays out abnormal behaviors based on human skeletal key points and a spatio-temporal graph convolutional network (ST-GCN). Skeleton extraction of the human body in video sequences was performed using Alphapose. To resolve the problem of graph convolutional networks not being effective enough for skeletal key points feature aggregation, we propose an NAM-STGCN model that incorporates a normalized attention mechanism. By using the activation function PReLU to optimize the model structure, the improved ST-GCN model can more effectively extract skeletal key points action features in the spatio-temporal dimension for the purposes of abnormal behavior recognition. The experimental results show that our optimized model achieves a 96.72% accuracy for recognition on the self-built dataset, which is 4.92% better than the original model; the model loss value converges below 0.2. Tests were performed on the KTH and Le2i datasets, which are both better than typical classification recognition networks. The model can precisely identify abnormal human behaviors, facilitating the detection of abnormalities and rescue in a timely manner and offering novel ideas for smart site construction.

Skin and soft tissue modeling and its impact on apparel modeling

Rigid body avatars do not fully define the complex interaction between human and body-worn product (humanoid-to-coveroid). Skin and soft tissue modeling to create more realistic 3D humanoid body models are needed. We considered if humanoid split lines relevant to pattern-engineering practice can be related to biodynamic and fold lines of the skin. Changes in skin and tissue are expected, depending on the dermis, the effects of movement, and the effects of coveroid pressure. The physiological functions of the skin may be assigned mechanical parameters for dynamic study utilizing biodynamic excisional skin tension (BEST) lines, main folding lines (MFL) with Langer’s lines. Critical to such study is the connecting of the skin to the rig (humanoid virtual skeleton). The use of stable (skeletal feature points related to both the virtual skeleton and apparel block patterns) and morphological (skin feature points identifying areas of morphological variation and dynamic study) landmarks for connecting the skin to rig was analyzed. We utilized these landmarks to drive lines as BEST, MFL and Langer’s lines for the mapping of skin deformations. Initial findings suggest the use of stable and morphological landmarks could have profoundly positive effects throughout the entire digital product creation (DPC) production pipeline and should be further explored & are important in developing standard topology practice.

Multimodal spatiotemporal skeletal kinematic gait feature fusion for vision-based fall detection

Fall happens when a person's movement coordination is disturbed, forcing them to rest on the ground unintentionally causing serious health risks. The objective of this work is to develop a Multimodal SpatioTemporal Skeletal Kinematic Gait Feature Fusion (MSTSK-GFF) classifier for detecting fall using video data. The walking pattern of an individual is referred to as gait. The event of fall recorded in video shows discrepancies and irregularities in gait patterns. Analysis of these patterns plays a vital role in the identification of fall risk. However, assessment of the gait patterns from video data remains challenging due to its spatial and temporal feature dependencies. The proposed MSTSK-GFF framework presents a multimodal feature fusion process that overcomes these challenges and generates two sets of spatiotemporal kinematic gait features using SpatioTemporal Graph Convolution Network (STGCN) and 1D-CNN network model. These two generated feature sets are combined using concatenative feature fusion process and classification model is constucted for detecting fall. For optimizing the network weights, a bio-inspired spotted hyena optimizer is applied during training process. Finally, performance of the classification model is evaluated and compared to detect fall in videos. The proposed work is experimented with the two vision-based fall datasets namely, UR Fall Detection (URFD) dataset and self-build dataset. The experimental outcome proves the effectiveness of MSTSK-GFF in terms of its classification accuracy of 96.53% and 95.80% with two datasets when compared with existing state-of-the-art techniques.

Dysregulation of Tweak and Fn14 in skeletal muscle of spinal muscular atrophy mice

BackgroundSpinal muscular atrophy (SMA) is a childhood neuromuscular disorder caused by depletion of the survival motor neuron (SMN) protein. SMA is characterized by the selective death of spinal cord motor neurons, leading to progressive muscle wasting. Loss of skeletal muscle in SMA is a combination of denervation-induced muscle atrophy and intrinsic muscle pathologies. Elucidation of the pathways involved is essential to identify the key molecules that contribute to and sustain muscle pathology. The tumor necrosis factor-like weak inducer of apoptosis (TWEAK)/TNF receptor superfamily member fibroblast growth factor-inducible 14 (Fn14) pathway has been shown to play a critical role in the regulation of denervation-induced muscle atrophy as well as muscle proliferation, differentiation, and metabolism in adults. However, it is not clear whether this pathway would be important in highly dynamic and developing muscle.MethodsWe thus investigated the potential role of the TWEAK/Fn14 pathway in SMA muscle pathology, using the severe Taiwanese Smn−/−; SMN2 and the less severe Smn2B/− SMA mice, which undergo a progressive neuromuscular decline in the first three post-natal weeks. We also used experimental models of denervation and muscle injury in pre-weaned wild-type (WT) animals and siRNA-mediated knockdown in C2C12 muscle cells to conduct additional mechanistic investigations.ResultsHere, we report significantly dysregulated expression of Tweak, Fn14, and previously proposed downstream effectors during disease progression in skeletal muscle of the two SMA mouse models. In addition, siRNA-mediated Smn knockdown in C2C12 myoblasts suggests a genetic interaction between Smn and the TWEAK/Fn14 pathway. Further analyses of SMA, Tweak−/−, and Fn14−/− mice revealed dysregulated myopathy, myogenesis, and glucose metabolism pathways as a common skeletal muscle feature, providing further evidence in support of a relationship between the TWEAK/Fn14 pathway and Smn. Finally, administration of the TWEAK/Fn14 agonist Fc-TWEAK improved disease phenotypes in the two SMA mouse models.ConclusionsOur study provides mechanistic insights into potential molecular players that contribute to muscle pathology in SMA and into likely differential responses of the TWEAK/Fn14 pathway in developing muscle.

Tracking using Human Pose Matching with Deep Association Metric

This paper proposes a novel approach to track multiple people utilizing skeletal information combined with visual appearance features to improve the accuracy of tracking people across different frames of a video. We extracted the appearance feature vectors and skeletal feature vectors for each detected person in every frame. Each individual was tracked by considering the cosine distance between the skeletal feature vectors along with the euclidean distance between the appearance feature vectors across different frames of a video. This reduces the dependency of the tracker over appearances of people thus making it more consistent, especially in videos with people having similar appearances such as sports videos with players wearing similar jerseys. The stance of an individual in continuing frames is expected to be similar considering the high frame rate of modern camera devices. Therefore it is befitting to consider skeletal features along with appearance features for tracking. Our paper is an incremental paper demonstrating improvement over SORT with a deep association metric approach. Our approach utilizing skeletal information combined with visual appearance information returns better MOT results on the MOT17 dataset using the yolov3 detector.

Design and Implementation of Home Training System based on Personal Fitness Using Kinect Sensor

In this paper, we implemented and developed a home training program using sensors. The implemented program grasps the user's athletic ability and suggests personalized exercise program. In modern society, the aging population is growing rapidly, and most of them are working and living in the interior. For this reason, health management of the aging population is becoming an important social issue, and it is recognized that muscle strength is an important factor. In this paper, we designed and developed a home training program for aged people to exercise in a comfortable environment like natural environment. In the developed program, we extract the skeletal feature points of the user using Microsoft's Kinect sensor and generate feature vectors to check the user's motion state. The proposed algorithm measures the user 's athletic ability and suggests a suitable exercise based on the information. As a result of applying kinematic sensor measurements to the proposed algorithm, it showed up to 97% accuracy. We measured the user's athletic ability through the program using the proposed algorithm and suggested the exercise suitable for the individual ability based on this information.

Andersen-Tawil syndrome: deep phenotyping reveals significant cardiac and neuromuscular morbidity.

Andersen-Tawil syndrome is a neurological channelopathy caused by mutations in the KCNJ2 gene that encodes the ubiquitously expressed Kir2.1 potassium channel. The syndrome is characterized by episodic weakness, cardiac arrythmias and dysmorphic features. However, the full extent of the multisystem phenotype is not well described. In-depth, multisystem phenotyping is required to inform diagnosis and guide management. We report our findings following deep multimodal phenotyping across all systems in a large case series of 69 total patients, with comprehensive data for 52. As a national referral centre, we assessed point prevalence and showed it is higher than previously reported, at 0.105 per 100 000 population in England. While the classical phenotype of episodic weakness is recognized, we found that a quarter of our cohort have fixed myopathy and 13.5% required a wheelchair or gait aid. We identified frequent fat accumulation on MRI and tubular aggregates on muscle biopsy, emphasizing the active myopathic process underpinning the potential for severe neuromuscular disability. Long exercise testing was not reliable in predicting neuromuscular symptoms. A normal long exercise test was seen in five patients, of whom four had episodic weakness. Sixty-seven per cent of patients treated with acetazolamide reported a good neuromuscular response. Thirteen per cent of the cohort required cardiac defibrillator or pacemaker insertion. An additional 23% reported syncope. Baseline electrocardiograms were not helpful in stratifying cardiac risk, but Holter monitoring was. A subset of patients had no cardiac symptoms, but had abnormal Holter monitor recordings which prompted medication treatment. We describe the utility of loop recorders to guide management in two such asymptomatic patients. Micrognathia was the most commonly reported skeletal feature; however, 8% of patients did not have dysmorphic features and one-third of patients had only mild dysmorphic features. We describe novel phenotypic features including abnormal echocardiogram in nine patients, prominent pain, fatigue and fasciculations. Five patients exhibited executive dysfunction and slowed processing which may be linked to central expression of KCNJ2. We report eight new KCNJ2 variants with in vitro functional data. Our series illustrates that Andersen-Tawil syndrome is not benign. We report marked neuromuscular morbidity and cardiac risk with multisystem involvement. Our key recommendations include proactive genetic screening of all family members of a proband. This is required, given the risk of cardiac arrhythmias among asymptomatic individuals, and a significant subset of Andersen-Tawil syndrome patients have no (or few) dysmorphic features or negative long exercise test. We discuss recommendations for increased cardiac surveillance and neuropsychometry testing.

Robust, wireless gastric optogenetic implants for the study of peripheral pathways and applications in obesity.

Optogenetics has the potential to transform the study of organ functions in the peripheral nervous system via relatively easy access to the nerves and a direct link between the brain and organ systems. Implementation typically requires a static skeletal feature for the securement of a fiber. Unfortunately, the soft nature of peripheral nervous systems makes the wired fiber-optic approach less ideal for the study of the peripheral nervous system. Existing wireless approaches could bypass some constraints associated with optical fibers and thereby offer organ specificity. However, they suffer from durability loss due to considerable biological strains and unable to perform longitudinal experiments. Here, we propose a new class of wireless gastric optogenetic implant for identifying signaling pathways, in particular viscerosensory pathways, that can regulate food intake to treat obesity. Robust, wireless gastric optogenetic implants with a tubing-assisted U-shaped tether directly interface with nerve endings in the stomach with chronic stability in operation (> 100 kilocycles) and allows for optogenetic stimulations of vagus nerves in a freely behaving animal. We demonstrated utilities of the proposed wireless device in in vivo experiments. Results suggest the potential for identifying interventions for the treatment of obesity.Clinical Relevance - Identification of the roles of subpopulations in viscerosensory pathways would provide the platform for the development of better therapeutics for the treatment of obesity.

Machine learning in the prognostic appraisal of Class III growth

Machine Learning (ML) is an emerging subfield of artificial intelligence with significant resources being applied to connect computer science, statistics, and medical problems. Currently, and even more so in the future, ML algorithms applied to the orthodontic specialty will offer sophisticated and automatic models able to process and synthesize data in ways orthodontists could never do themselves, and ultimately convert data into intelligent treatment actions. This work focuses on the usefulness of two ML methodologies, LASSO networks (Ln), and Boruta selection (Ba), to simplify information from different types of pathogenic processes leading to the worsening of skeletal Class III malocclusion. Cephalometric analyses of 144 Class III untreated subjects followed longitudinally during the growth process (4–19 years) were performed. After separating subjects into two subgroups of 116 with mild (M) and 28 with very serious (VS) unfavorable growth, cephalometric features were processed using Ba and Ln algorithms for feature selection and regularization. The selection procedure revealed the unexpected predictive importance of the combination of two often overlooked craniofacial variables, SN-PP and L1-MP angles. Ln regularization highlighted additional feature interactions between M and VS growing subjects. Thus, the appropriate removal of redundant cephalometric features from the dataset contributed to the detection of subjects affected by serious unfavorable craniofacial progression and revealed the unexpected prognostic value of some skeletal feature interactions.

Kinect-based integrated physiotherapy mentor application for shoulder damage

With the increasing rate of disabled people in the world, the number of people in need of physiotherapy is increasing. Most of the disabled persons with shoulder damaged should do physiotherapy exercises. The telerehabilitation systems using developed technologies have advantages over conventional physiotherapy methods. Such systems have the potential to improve communication between patients and physiotherapists. They also allow for easier information and medical data sharing. In this study, a telerehabilitation system based on Kinect Sensor (Kinect 2) was developed to observe and evaluate shoulder damaged patient exercises by using the Kinect skeletal monitoring feature. The proposed system consists of two basic components: the web application that enables the communication between patients and physiotherapists, allows the physiotherapist to observe/update patient exercise data, patients information, and the console application which enables the patients to perform their exercises correctly. This console app was developed for patients with shoulder damage. Abduction, flexion, internal rotation, external rotation, extension exercises can be performed correctly thanks to the developed application. The proposed system was evaluated in Bilecik State Hospital for 29 shoulder damaged volunteers. 14 of the volunteers were treated with the conventional methods and 15 of them were treated with the proposed system. The recovery of the volunteers was analyzed for each intragroup and intergroup comparison was made. Statistically, significant decreases were observed in the limitations of the patients treated with the proposed system. For abduction, flexion, internal rotation, and external rotation exercises, the Kinect-based physiotherapy mentor application was more successful than conventional method, while all volunteers treated by the proposed method had no limitations in extension exercise. As a result, the Kinect-based physiotherapy mentor application can be used as an alternative telerehabilitation method for patients with shoulder damaged.

Feet first: Adaptive growth in magellanic penguin chicks.

Growing animals should allocate their limited resources in ways that maximize survival. Seabird chicks must balance the growth of features and fat reserves needed to survive on land with those needed to successfully fledge and survive at sea. We used a large, 34‐year dataset to examine energy allocation in Magellanic penguin chicks. Based on the temporal trends in the selective pressures that chicks faced, we developed predictions relating to the timing of skeletal feature growth (Prediction 1), variation in skeletal feature size and shape (Prediction 2), and responses to periods of high energetic constraint (Prediction 3). We tested our predictions using descriptive statistics, generalized additive models, and principal component analysis. Nearly all of our predictions were supported. Chicks grew their feet first, then their flippers. They continued to grow their bill after fledging (Prediction 1). Variance in feature size increased in young chicks but declined before fledging; this variance was largely driven by overall size rather than by shape (Prediction 2). Chicks that died grew slower and varied more in feature size than those that fledged (Prediction 2). Skeletal features grew rapidly prior to thermoregulation and feet and flippers were 90% grown prior to juvenile feather growth; both thermoregulation and feather growth are energetically expensive (Prediction 3). To avoid starvation, chicks prioritized storing mass during the first 10 days after hatching; then, the body condition of chicks began to decline (Prediction 3). In contrast to our prediction of mass prioritization in young chicks, chicks that were relatively light for their age had high skeletal size to mass ratios. Chicks did not show evidence of reaching physiological growth limits (Prediction 3). By examining energy allocation patterns at fine temporal scales and in the context of detailed natural history data, we provide insight into the trade‐offs faced by growing animals.

Attention-Based Sign Language Recognition Network Utilizing Keyframe Sampling and Skeletal Features

Sign language recognition(SLR) is a multidisciplinary research topic in pattern recognition and computer vision. Due to large amount of data from the continuous frames of sign language videos, selecting representative data to eliminate irrelevant information has always been a challenging problem in data preprocessing of sign language samples. In recent years, skeletal data emerged as a new type of data but received insufficient attention. Meanwhile, due to the increasing diversity of sign language features, making full use of them has also been an important research topic. In this paper, we improve keyframe-centered clips (KCC) sampling to get a new kind of sampling method called optimized keyframe-centered clips (OptimKCC) sampling to select key actions from sign language videos. Besides, we design a new kind of skeletal feature called Multi-Plane Vector Relation (MPVR) to describe the video samples. Finally, combined with the attention mechanism, we also use Attention-Based networks to distribute weights to the temporal features and the spatial features extracted from skeletal data. We implement comparison experiments on our own and the public sign language dataset under the Signer-Independent and the Signer-Dependent circumstances to show the advantages of our methods.

The study of facial morphology in patients with vertical growth pattern (hyperdivergent) lacking or showing temporomandibular disorders symptoms

AimAccording to the large number of studies, there seems to be a significant relation between hyperdivergence growth pattern and temporomandibular disorders. However, it is not clearly scrutinized which morphological factors can contribute to the development of temporomandibular disorders. The aim of this study was to investigate the relation between some skeletal and facial morphologic features of hyperdivergent facial growth pattern patients and temporomandibular disorders. Materials and MethodsThis case-control study was performed on a population of 50 patients including 25 patients with temporomandibular disorder and 25 non- temporomandibular disorder individuals aged between 15–30 years with hyperdivergent growth pattern (Frankfurt Mandibualr plane angle between 28 and 35 degree). Two groups of symptomatic and asymptomatic patients (case and control) were matched for age and gender. In order to evaluate the skeletal and dental variables, lateral cephalometries was taken and traced after obtaining written consent from the patients. The measured variables were A point to Nasion to B point, Frankfurt plane to Nasion to Pogonion, Nasion to A point to Pogonion, Sella to Gnathion to Frankfurt plan, Articular point to Gonion-Menton, Sella to Articular point to Gonion, Palatal plane to Mandibular plane, and Articular point to Gonion angles. The intervals of Basion to Nasion, Sella to Basion, Gonion to Menton, and amounts of Anterior facial height meas, Upper Anterior facial height ratio to Lower Anterior facial height, posterior facial height, overjet, and overbite Data were analyzed by Chi2 test, t-test, and multivariate test. ResultsAccording to the descriptive statistics, the age of the control and case groups averaged 21.12±1.99 and 21.63±1.58 years respectively. Among the people referred to the dental school, the frequency of males and females in the control and case groups were 6, 7 and 19, 18 respectively. The results of t-test and multivariate tests indicated significant differences between the two variables of overbite and mandibular length. ConclusionThe present study revealed increased overbite (dental feature) and mandibular length (skeletal feature) is more likely to be associated with a higher risk of temporomandibular disorders joint disease in patients with hyperdivergent facial growth pattern. Clinical significanceTreatment of the deep bite condition can be helpful in improving temporomandibular disorder.

Organic skeleton correspondence using part arrangements

Shape correspondence between semantically similar organic shapes with large shape variations is a difficult problem in shape analysis. Since part geometries are no longer similar, we claim that the challenge is to extract and compare prominent shape substructures, which are recurring part arrangements among semantically related shapes. Our main premise is that the challenge can be solved more efficiently on curve skeleton graphs of shapes, which provide a concise abstraction of shape geometry and structure. Instead of directly searching exponentially many skeleton subgraphs, our method extracts the intrinsic reflectional symmetry axis of the skeleton to guide the generation of subgraphs as part arrangements. For any two subgraphs from two skeletons, their orientations are aligned and their pose variations are normalized for matching. Finally, the matchings of all subgraph pairs are evaluated and accumulated to the skeletal feature node correspondences. The comparison results with the state-of-the-art work show that our method significantly improves the efficiency and accuracy of the semantic correspondence between a variety of shapes.

The evolution of underwater flight: The redistribution of pectoral fin rays, in manta rays and their relatives (Myliobatidae)

Batoids are a diverse clade of flat cartilaginous fishes that occur primarily in benthic marine habitats. The skates and rays typically use their flexible pectoral fins for feeding and propulsion via undulatory swimming. However, two groups of rays have adopted a pelagic or bentho-pelagic lifestyle and utilize oscillatory swimming-the Myliobatidae and Gymnuridae. The myliobatids have evolved cephalic lobes, anteriorly extended appendages that are optimized for feeding, while their pectoral fins exhibit several modifications that likely arose in association with functional optimization of pelagic cruising via oscillatory flight. Here, we examine variation in fin ray distribution and ontogenetic timing of fin ray development in batoid pectoral fins in an evolutionary context using the following methods: radiography, computed tomography, dissections, and cleared and stained specimens. We propose an index for characterizing variation in the distribution of pectoral fin rays. While undulatory swimmers exhibit symmetry or slight anterior bias, we found a posterior shift in the distribution of fin rays that arose in two distinct lineages in association with oscillatory swimming. Undulatory and oscillatory swimmers occupy nonoverlapping morphospace with respect to fin ray distribution illustrating significant remodeling of pectoral fins in oscillatory swimmers. Further, we describe a derived skeletal feature in anterior pectoral fins of the Myliobatidae that is likely associated with optimization of oscillatory swimming. By examining the distribution of fin rays with clearly defined articulation points, we were able to infer evolutionary trends and body plan remodeling associated with invasion of the pelagic environment. Finally, we found that the number and distribution of fin rays is set early in development in the little skate, round stingray, and cownose ray, suggesting that fin ray counts from specimens after birth or hatching are representative of adults and therefore comparable among species.