Angle Variance Research Articles

Multiscale Kernel Entropy Component Analysis With Application to Complex Industrial Process Monitoring

Modern industrial processes are characterized by numerous measurement points and wide operating ranges, resulting in extremely complex correlations among variables. Therefore, an effective monitoring system should balance diverse process characteristics such as nonlinearity, non-Gaussianity, and multiscale simultaneously. Moreover, it should have the ability to detect and diagnose faults in the incipient stage, thus avoiding accident escalation. With these goals in mind, this paper proposes an integrated monitoring solution based on multiscale kernel entropy component analysis (MSKECA). Specifically, process variables are first decomposed into approximations and details at different scales in real-time using the moving window-based wavelet, and contributions from each scale are collected in separate matrices. Then, KECA-based local-scale models are built to sift out important detail scales for reconstruction along with the approximate scale. Lastly, a KECA-based global-scale model is developed to monitor the reconstructed data. To improve the fault detection performance, a novel monitoring index based on the angle metric called angle variance index (AVI) is designed. In addition, to achieve effective diagnosis, MSKECA-based contribution plots are constructed, which depict the contributions of variables to faults at each scale, thus comprehensively revealing the root causes. Finally, the effectiveness and superiority of the proposed solution are validated by comparisons with other advanced counterparts in two industrial scenarios. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —This paper proposes an integrated monitoring solution for complex industrial processes, i.e., MSKECA-based fault detection and diagnosis. The solution takes into account the diversity of process characteristics, the effectiveness of incipient fault detection, and the richness of diagnostic information. Specifically, 1) MSKECA can simultaneously handle the nonlinear, non-Gaussian, and multiscale characteristics that are prevalent in real process data. It enables on-line detection of significant events occurring at different scales and extraction of fault-sensitive features for monitoring; 2) based on the angular structure of KECA, the AVI statistic is designed, which exhibits low autocorrelation and is sensitive to faults. Leveraging the statistic, MSKECA allows reliable and prompt responses to faults; 3) MSKECA-based contribution plots not only convey diverse diagnostic information including fault variable, type, and grade but also are not susceptible to the smearing effect, which is helpful for practitioners to achieve fault repair. The solution has proven useful for a real hot rolling process. It can also be extended to other industrial processes.

Chewing efficiency and contact area discrepancies in Angle's class I and II malocclusion: A comparative study

ObjectiveThe study aimed to assess and compare both the chewing efficiency and the contact area between class I and class II of Angle’s malocclusions. Material and methodsA total of 120 individuals aged between 19 and 30 years were examined and were divided into two groups according to Angle’s class (Class I and class II). The chewing efficiency was quantified using the ViewGum software and two-colored chewing gum. The contact area was quantified using modeling wax and MATLAB software. All data were collected, then, analyzed using SPSS software 21. Data normality was checked through kurtosis test. Descriptive results were calculated. Matched sample t-tests were used to compare chewing efficiency measurements between right and left sides. Independent t-tests were used to compare chewing efficiency and the contact area between class I and class II of Angle’s malocclusions. Linear regression and Pearson correlation were used to assess the correlation between chewing efficiency and the contact area. The significance level was fixed at p = 0.05. ResultsFor group 1, the mean hue value on both sides was 0.086 ± 0.058. For group 2, it was 0.095 ± 0.055. The difference between both groups was statistically significant (p = 0.03). For group 1, the mean contact area was 49.91 ± 21.47 mm2. For group 2, it was 51.42 ± 19.76 mm2. The difference was statistically not significant (p = 0.4). The correlation between the contact area and the chewing efficiency in both groups was statistically significant and it was negative (in class I p = 0 and R = -0.616; in class II p = 0.01 and R = −0.408). ConclusionThe Chewing is better for patients with Angle’s Class I malocclusion. The contact area is higher. Larger occlusal contact area leads to higher masticatory efficiency. Further studies should be conducted.

Power loss measurement system for soft magnetic materials in megahertz frequencies

Soft magnetic materials play a key role in developing integrated and miniaturized power electronics. However, it is still challenging to accurately evaluate the high-frequency (MHz) power loss of soft magnetic materials, which determines the energy efficiency of power electronics. In this paper, we firstly trace the high-frequency error origination in traditional electrometric method and propose a two-step calibration method for the system to compensate the phase angle discrepancy between exciting current and induced voltage. The first-step calibration eliminates the phase angle discrepancies caused by propagation delay of current and voltage probes, while the second-step calibration compensates residual phase angle discrepancies. A calibration factor k is defined to determine whether the second-step calibration is needed. The experimental results demonstrate that the system’s phase discrepancy is compensated within 0.13°, enabling power loss measurement frequency up to 20 MHz that much higher than present apparatus. Our results will facilitate the development of high-frequency soft magnetic materials and promote the miniaturization of power electronics.

Accuracy analysis of artificial intelligence-assisted three-dimensional preoperative planning in total hip replacement.

This study aims to assess the outcome of total hip arthroplasty (THA) using artificial intelligence (AI)-assisted three-dimensional (3D) preoperative planning technology in terms of predicting prosthesis size, acetabular cup positioning, and lowerlimb length restoration. Between January 2020 and July 2022, a total of 161 patients (101 males, 60 females; mean age: 57.6±10.5 years; range, 31 to 80 years) who underwent primary unilateral THA were retrospectively analyzed. The patients were divided into two groups as those who were treated with AI-assisted 3D preoperative planning technology (the observation group, n=95) and patients who were treated with traditional two-dimensional (2D) X-ray template planning technology (the control group, n=66). The accuracy of the planning was based on the consistency of the preoperative planning and intraoperative models. The difference between the observation group and the control group was statistically significant in terms of the accuracy of the preoperative planning of acetabular prostheses (54% vs. 38%, p=0.048) and femoral prostheses (64% vs. 44%, p=0.011), with both values significantly higher in the observation group. The mean inclination angle, anteversion angle, and limb length discrepancy (LLD) in the observation group were 36.85°±4.82°, 12.10°±5.33°, and 2.18±2.70 mm, respectively, while those in the control group were 35.06°±6.07°, 10.95°±5.09°, and 4.42±3.85 mm, respectively. There was a statistically significant difference between the two groups in terms of inclination angle and LLD (p<0.05 for both), but there was no significant difference in terms of anteversion angle (p>0.05). In the observation group, 86.3% (82/95) of acetabular cups were implanted within the Lewinnek safe zone (72.7% [48/66] in the control group), while 83.2% (79/95) were within the Callanan safe zone (69.7% [46/66] in the control group), with both values higher in the observation group (p<0.05). Overall, AI-assisted 3D preoperative planning is superior to traditional 2D X-ray template planning for predicting prosthesis size, and it also has the advantage in terms of acetabular cup positioning and lower-limb length restoration.

PolSAR Image Classification by Introducing POA and HA Variances

A polarimetric synthetic aperture radar (PolSAR) has great potential in ground target classification. However, current methods experience difficulties in separating forests and buildings, especially oriented buildings. To address this issue, inspired by the three-component decomposition method, multiple new scattering models were proposed to describe the difference between forest scattering and building scattering. However, this problem cannot effectively be solved with scattering power alone since HV polarization records significant scattering powers from building areas that are similar to vegetation. Therefore, in this study, two new parameters, the polarization orientation angle (POA) variance and helix angle (HA) variance, were defined to describe the distributions of buildings and forests. By combining scattering power with POA and HA variances, the random forest algorithm was used to conduct the land cover classification, focusing on distinguishing between forests and oriented buildings. Finally, the C- and L-band polarimetric SAR data acquired by the GF-3, ALOS1 PALSAR, and SAOCOM systems were selected to test the proposed method. The results indicate that it is feasible to improve PolSAR classification accuracy by introducing polarimetric parameters. Quantitatively, the classification accuracies increased by 23.78%, 10.80%, and 12.97% for the ALOS1 PALSAR, GF-3, and SAOCOM data, respectively.

Fused-Deposition Modeling 3D Printing of Short-Cut Carbon-Fiber-Reinforced PA6 Composites for Strengthening, Toughening, and Light Weighting.

Additive manufacturing of carbon-fiber-reinforced polymer (CFRP) has been widely used in many fields. However, issues such as inconsistent fiber orientation distribution and void formation during the layer stacking process have hindered the further optimization of the composite material's performance. This study aimed to address these challenges by conducting a comprehensive investigation into the influence of carbon fiber content and printing parameters on the micro-morphology, thermal properties, and mechanical properties of PA6-CF composites. Additionally, a heat treatment process was proposed to enhance the interlayer bonding and tensile properties of the printed composites in the printing direction. The experimental results demonstrate that the PA6-CF25 composite achieved the highest tensile strength of 163 MPa under optimal heat treatment conditions: 120 °C for 7.5 h. This corresponds to a significant tensile strength enhancement of 406% compared to the unreinforced composites, which represents the highest reported improvement in the current field of CFRP-fused deposition 3D printing. Additionally, we have innovatively developed a single-layer monofilament CF-OD model to quantitatively analyze the influence of fiber orientation distribution on the properties of the composite material. Under specific heat treatment conditions, the sample exhibits an average orientation angle μ of 0.43 and an orientation angle variance of 8.02. The peak frequency of fiber orientation closely aligns with 0°, which corresponds to the printing direction. Finally, the study explored the lightweight applications of the composite material, showcasing the impressive specific energy absorption (SEA) value of 17,800 J/kg when implementing 3D-printed PA6-CF composites as fillers in automobile crash boxes.

A feasibility study of applying two-dimensional photogrammetry for screening and monitoring of patients with adolescent idiopathic scoliosis in clinical practice

Standing posteroanterior radiographs have been the golden standard to quantify the severity of scoliosis deformity. However, it exposes ionizing radiation to scoliosis patients, and cannot be used for routine screening and monitoring. This study aimed to develop a protocol of measuring postural indexes by using the noninvasive and radiation-free two-dimensional (2D) photogrammetry method and identify its clinical value in scoliosis screening and monitoring. The five postural indexes were measured from the posterior view of 110 participants. One-way ANOVA with post hoc Tukey HSD/Games–Howell analysis was used to compare the differences between the participants in the scoliosis group and the non-scoliosis group. Pearson coefficients of correlation were analyzed to identify the relationships between Cobb angles and each of the five quantitative postural indexes. Based on 2D photogrammetry, the postural indexes of C7 deviation (p = 0.02), shoulder alignment (p < 0.001), scapula alignment (p < 0.001), waist angle discrepancy (p < 0.001), and PSIS alignment (p < 0.001) could significantly differentiate scoliosis and non-scoliosis patients during screening. The waist angle discrepancy (r = 0.4, p = 0.01; r = 0.8, p = 0.03; r = 0.7, p = 0.01) and shoulder alignment (r = 0.6, p = 0.03) had moderate to strong positive correlations with the Cobb angles, which supported their clinical values in monitoring scoliotic curvature changes of adolescent idiopathic scoliosis (AIS) patients.

Abnormal coordination of upper extremity during target reaching in persons post stroke

Understanding abnormal synergy of the upper extremity (UE) in stroke survivors is critical for better identification of motor impairment. Here, we investigated to what extent stroke survivors retain the ability to coordinate multiple joints of the arm during a reaching task. Using an exoskeleton robot, 37 stroke survivors’ arm joint angles (θ) and torques (τ) during hand reaching in the horizontal plane was compared to that of 13 healthy controls. Kinematic and kinetic coordination patterns were quantified as variances of the multiple-joint angles and multiple-joint torques across trials, respectively, that were partitioned into task-irrelevant variance (TIVθ and TIVτ) and task-relevant variance (TRVθ and TRVτ). TIVθ and TRVθ (or TIVτ and TRVτ) led to consistent and inconsistent hand position (or force), respectively. The index of synergy (ISθ and ISτ) was determined as {IS}_{theta }=({TIV}_{theta }-{TRV}_{theta })/({TIV}_{theta }+{TRV}_{theta }) and {IS}_{tau }=({TIV}_{tau }-{TRV}_{tau })/({TIV}_{tau }+{TRV}_{tau }) for kinematic and kinetic coordination patterns, respectively. Both kinematic ISθ and kinetic ISτ in the stroke group were significantly lower than that of the control group, indicating stroke survivors had impaired reaching abilities in utilizing the multiple joints of the UE for successful completion of a reaching task. The reduction of kinematic ISθ in the stroke group was mainly attributed to the lower TIVθ as compared to the control group, while the reduction of kinetic ISτ was mainly due to the higher {TRV}_{tau } as well as lower TIVτ. Our results also indicated that stroke may lead to motor deficits in formation of abnormal kinetic synergistic movement of UE, especially during outward movement. The findings in abnormal synergy patterns provides a better understanding of motor impairment, suggesting that impairment-specific treatment could be identified to help improve UE synergies, focusing on outward movements.

Characterizing the Sound-Scattering Layer and Its Environmental Drivers in the North Equatorial Current of the Central and Western Pacific Ocean

Acoustic technology is an essential tool for detecting marine biological resources and has been widely used in sound-scattering layer (SSL) research. The North Equatorial Current (NEC) warm pool region of the Central and Western Pacific Ocean has a vast distribution of micronekton and zooplankton; analyzing the SSL characteristics in this region is vital for monitoring the marine environment and studying the marine ecosystem. In this study, we statistically analyzed the spatiotemporal factors of 10–200 m SSL in the NEC of the Central and Western Pacific Ocean using acoustic survey data collected by the “Songhang” research vessel (RV) in 2022, and the influence of environmental factors on the scattering layer distribution was analyzed using the Generalized Additive Model (GAM). The results showed that the SSL in the warm pool area of the NEC is distributed in shallow waters above 100 m. The primary scatterers are micronekton and zooplankton, and this SSL had diel vertical migration behavior. By comparing Akaike’s Information Criterion of different GAMs, the model consisting of six factors, namely, temperature, current velocity, turbidity, solar altitude angle, longitude, and latitude, was remarkable. Each model’s factor effects primarily influence the contribution of the volume-backscatter strength (Sv). The cumulative deviation explanation rate of the Sv was 67.2%, among which the highest explanation rate of solar altitude angle variance was 35.4%, the most critical environmental factor. The results of this study can provide a reference for long-term studies on ecological changes and their effects on micronekton and zooplankton distribution.

Interaction of Relativistic Electrons with Packets of the Electromagnetic Ion Cyclotron Waves of Finite Length and Low Amplitude

Interaction of relativistic electrons with packets of the electromagnetic ion cyclotron waves of finite length and low amplitude in the Earth’s radiation belts is analyzed. The variance of the equatorial pitch angle of electrons for wave packets located near the Earth’s geomagnetic equator is estimated analytically within the linear approximation. The analytical estimates agree with the results of numerical test-particle simulation. It is demonstrated that reduction in the packet length extends the interaction range to lower energies beyond the range of resonant energies. Such an interaction can result in precipitation of electrons with energies on the order of several hundred kiloelectronvolts into the ionosphere.

Measuring Arm and Hand Joint Kinematics to Estimate Impairment During a Functional Reach and Grasp Task after Stroke

Background Current approaches to characterizing deficits in upper limb movements after stroke typically focus either on changes in a functional measure, for example, how well a patient can complete a task, or changes in impairment, for example, isolated measurements of joint range of motion. However, there can be notable dissociations between static measures of impairment versus those of function. Objective We develop a method to measure upper limb joint angles during performance of a functional task and use measurements to characterize joint impairment in the context of a functional task. Methods We developed a sensorized glove that can precisely measure select finger, hand, and arm joints while participants complete a functional reach-to-grasp task involving manipulation of a sensorized object. Results We first characterized the accuracy and precision of the glove’s joint angle measurements. We then measured joint angles in neurologically intact participants (n = 4 participants, 8 limbs) to define the expected distribution of joint angle variation during task execution. These distributions were used to normalize finger, hand, and arm joint angles in stroke participants (n = 6) as they performed the task. We present a participant-specific visualization of functional joint angle variance which illustrated that stroke participants with nearly identical clinical scores exhibited unique patterns of joint angle variation. Conclusions Overall, measuring individual joint angles in the context of a functional task may inform whether changes in functional scores over recovery or rehabilitation are driven by changes in impairment or the development of compensatory strategies, and provide a quantified path toward personalized rehabilitative therapy.

Effects of Toolpath Parameters on Engagement Angle and Cutting Force in Ellipse-Based Trochoidal Milling of Titanium Alloy Ti-6Al-4V

Trochoidal milling is an efficient strategy for the rough machining of difficult-to-cut materials. The true trochoidal toolpath has C2 continuity and avoids sharp changes in engagement angle and cutting load, resulting in smooth machine tool movement. However, its total length is too long, and its engagement angle is uneven. These factors limit further improvements in the material removal rate. Based on the true trochoidal toolpath model, this paper develops an ellipse-based trochoidal toolpath generation method by introducing a compression ratio in the trochoidal step direction. The analytical model of engagement angle and the mechanistic model of the cutting force are proposed. A series of simulations and milling experiments were conducted to analyze the effects of toolpath parameters on the engagement angle and the cutting force. The results show that the compression ratio has the most significant effects. A compression ratio of 50% is optimal, using which the total toolpath length is reduced by 34.0%, and the variance of the engagement angle is reduced by 31.2% compared with that of the true trochoidal toolpath. The profile of the total cutting force corresponds to that of the engagement angle.

Lack of visual information alters lower limb motor coordination to control center of mass trajectory during walking

Vision, as queen of the senses, plays a critical role in guiding locomotion. Little is known about the effects of vision on gait coordination in terms of variability. The uncontrolled manifold (UCM) approach offers a window to the structure of motor variability that has been difficult to obtain from the traditional correlation analysis. In this study, we used the UCM analysis to quantify how the lower limb motion is coordinated to control the center of mass (COM) while walking under different visual conditions. We also probed how synergy strength evolved along the stance phase. Ten healthy participants walked on the treadmill with and without visual information. Leg joint angle variance with respect to the whole-body COM was partitioned into good (i.e., the one that kept the COM) and bad (i.e., the one that changed the COM) variances. We observed that after vision was eliminated, both variances increased throughout the stance phase while the strength of the synergy (the normalized difference between the two variances) decreased significantly and even reduced to zero at heel contact. Thus, walking with restricted vision alters the strength of the kinematic synergy to control COM in the plane of progression. We also found that the strength of this synergy varied across different walking phases and gait events in both visual conditions. We concluded that the UCM analysis can quantify altered coordination of COM when vision is blocked and sheds insights on the role of vision in the synergistic control of locomotion.

Internal Acoustic Meatus Localization During Anterior Petrosectomy: A Novel Technique

Anterior petrosectomy demands localization of the internal auditory canal (IAC) for safe bone drilling and maximum exposure. Various techniques have been described in the literature, each with shortcomings. We propose a new technique to localize the internal acoustic meatus (IAM) using more consistent anatomical landmarks. The study was done in three phases. In phase-I (radiological), computed tomography scan heads of fifty patients (100 sides) were analyzed. Arcuate eminence-Greater Superficial Petrosal Nerve bifurcation angle(Garcia-Ibanez technique), Arcuate eminence-IAC angle(Fisch technique) and a new angle formed between foramen ovale (FO) and foramen spinosum (FS) line, and FS and IAM line (FO-FS-IAM angle) was measured. The mean, standard deviation, and variance were calculated. In phase-II (cadaveric), the FO-FS-IAM angle was measured on five (10 sides) dry skulls. In phase-III (clinical), the IAM was localized using the FO-FS-IAM angle in 13 patients. The mean angle between arcuate eminence and Greater Superficial Petrosal Nerve (Garcia-Ibanez technique) was 126.20±11.63°(range 106-156) with a variance of 135.20. The mean bifurcation angle was 63±5.81°(range 53-78). By the Fisch technique, the mean arcuate-IAM angle was 73.5±11.70°(range 51-105) with a variance of 137.18. By our technique, the mean FO-FS-IAM angle was 94.72±5.89°(range 84-108). The variance was 34.73. The mean FO-FS-IAM angle on dry skulls was identical (95±1.97°) to our radiological measurements. This angle was reproduced reliably in clinical cases for localizing the IAM during anterior petrosectomy. The FO-FS-IAM angle variance was much lower than the analogous angles measured by Garcia-Ibanez and Fisch techniques, making it a more reliable and effective tool for localizing the IAM.

Divergence of the Variance of the Optical Phase in Gain-Switched Semiconductor Lasers Described by Stochastic Rate Equations

We report a theoretical study of the phase diffusion in a gain-switched single-mode semiconductor laser. We use stochastic rate equations for the electric field to analyze the phase statistics of the gain-switched laser. Their use avoids the instabilities obtained with rate equations for the photon number and optical phase when the photon number is small. However, we show that a new problem appears when the field equations are integrated: the variance of the optical phase becomes divergent. This divergence cannot be observed with the numerical integration of the commonly used equations for the photon number and optical phase because of the previous instabilities. The divergence of the phase variance means that this quantity does not reach a fixed value as the integration time step is decreased. We find that the phase variance increases as the integration time step decreases, with no sign of saturation behavior even for tiny steps. We explain the divergence by making the analogy of our problem with two-dimensional Brownian motion. The fact that the divergence appears is not surprising because in 1940 Paul L\`evy demonstrated that the variance of the polar angle in two-dimensional Brownian motion is a divergent quantity. Our results show that stochastic rate equations for the photon number and phase are not appropriate for describing the phase statistics when the photon number is small. Simulation of the stochastic rate equations for the electric field are consistent with L\`evy's results but gives unphysical results since an infinite value is obtained for a quantity that can be measured.

Multitask Sparse Representation of Two-Dimensional Variational Mode Decomposition Components for SAR Target Recognition

A synthetic aperture radar (SAR) automatic target recognition (ATR) method is developed based on the two-dimensional variational mode decomposition (2D-VMD). 2D-VMD decomposes original SAR images into multiscale components, which depict the time-frequency properties of the targets. The original image and its 2D-VMD components are highly correlated, so the multitask sparse representation is chosen to jointly represent them. According to the resulted reconstruction errors of different classes, the target label of test sample can be classified. The moving and stationary target acquisition and recognition (MSTAR) dataset is used to set up the standard operating condition (SOC) and several extended operating conditions (EOCs) including configuration variants, depression angle variances, noise corruption, and partial occlusion to test and validate the proposed method. The results confirm the effectiveness and robustness of the proposed method compared with several state-of-the-art SAR ATR references.

Optimizing the results of facial animation surgery: Botulinum toxin injection into free functional gracilis flap transfer

Although neuromuscular gracilis transplantation is the best choice for facial reanimation in patients with congenital or inveterate palsy, the results are not completely satisfactory. Ancillary procedures developed to achieve better symmetry of the smile and reduce the hypercontractility of the transplanted muscle have been reported. However, the intramuscular injection of botulinum toxin has not been described for this purpose. Patients undergoing gracilis injections of botulinum toxin after facial reanimation surgery between September 1, 2020, and June 1, 2022, were retrospectively enrolled in this study. We collected photographs taken before and 20-30 days after injection and compared the symmetry of the face using software. Nine patients with a mean age of 23.56 years (range, 7-56 years) were enrolled. Reinnervation of the muscle was provided by the contralateral healthy facial nerve via a sural cross-graft (four patients), by the ipsilateral masseteric nerve (three cases), and by the contralateral masseteric and facial nerve (two). Using Emotrics software, we identified differences in the commissure excursion discrepancy of 3.82mm, the smile angle discrepancy of 0.084°, and the dental show discrepancy of 1.49mm; the average difference in the commissure height deviation was 2.26mm (P=0.02), and those in the upper- and lower-lip height deviation were 1.05mm and 1.49mm, respectively. Gracilis injection of botulinum toxin after gracilis transplantation is a safe and feasible procedure that could be applicable to all patients with asymmetric smiles related to excessive transplant contraction. It yields good esthetic results with little to no related morbidity.

How useful is pre-referral pediatric spine imaging?

PurposePrimary care physicians rely on radiology reports to confirm a scoliosis diagnosis and inform the need for spine specialist referral. In turn, spine specialists use these reports for triage decisions and planning of care. The objectives were to determine (1) the adequacy of index images to inform treatment decisions at the initial consultation and (2) the utility of index radiology reports for appropriate triage decisions.MethodsA retrospective chart review was conducted including all idiopathic scoliosis patients seen for initial consultation, aged three to 18 years, between January 1 and April 30, 2021. A score out of five was generated based on the adequacy of index images to provide accurate Cobb angle measurements and determine skeletal maturity. Index images were considered inadequate if repeat imaging was necessary. Index radiology reports, associated imaging, and new imaging, if obtained at the initial consultation, were compared.ResultsOf the 94 patients reviewed, 79% (n = 74) required repeat imaging at the initial consultation, of which 74% (n = 55) were due to insufficient quality and/or limited field of view. Of index images available for review at the initial consult (n = 80), 41.2% scored five out of five, and 32.5% scored two or below. Comparing index radiology reports to initial visit evaluation with ≤ 60 days between imaging (n = 49), discrepancies in Cobb angle were found in 24.5% (95% CI 14.6, 38.1) of patients. The Risser stage was reported in 14% of index radiology reports.ConclusionAlthough pre-referral pediatric spine radiographs serve a diagnostic purpose, most are inadequate for comprehensive idiopathic scoliosis evaluation.Level of evidenceIII.

Effects of postural instability on the coordination between posture and arm reaching

Reaching from standing requires adjustments of hand movement and posture, which are assured by redundant kinematic degrees of freedom. However, the increased demand for postural adjustments may interfere with the stability of reaching. The objective of this study was to investigate the effect of postural instability on the use of kinematic redundancy to stabilize the finger and center-of-mass trajectories during reaching from standing in healthy adults. Sixteen healthy young adults performed reaching movements from standing with and without postural instability induced by small base-of-support. The three-dimensional positions of 48 markers were recorded at 100 Hz. The uncontrolled manifold (UCM) analysis was performed separately with the finger and center-of-mass positions being the performance variables, and joint angles being the elemental variables. ΔV, the normalized difference between the variance in joint angle that does not affect task performance (VUCM) and the variance that does affect task performance (VORT), was calculated separately for finger (ΔVEP) and center-of-mass (ΔVCOM) positions, and was compared between stable and unstable base-of-support conditions. ΔVEP decreased after movement onset and reached its minimum value at around 30–50% of the normalized movement time, and increased until movement offset, while ΔVCOM remained stable. At 60%–100% normalized movement time, ΔVEP was significantly reduced in the unstable base-of-support, compared to the stable base-of-support condition. ΔVCOM remained similar between the two conditions. At movement offset, ΔVEP was significantly reduced in the unstable base-of-support, compared to the stable base-of-support condition, and was associated with a substantial increase in VORT. Postural instability might reduce the ability to use kinematic redundancy to stabilize the reaching movement. The central nervous system may prioritize the maintenance of postural stability over focal movement when postural stability is challenged.

CAMRI Loss: Improving the Recall of a Specific Class without Sacrificing Accuracy

In real-world applications of multi-class classification models, misclassification in an important class (e.g., stop sign) can be significantly more harmful than in other classes (e.g., speed limit). In this paper, we propose a loss function that can improve the recall of an important class while maintaining the same level of accuracy as the case using cross-entropy loss. For our purpose, we need to make the separation of the important class better than the other classes. However, existing methods that give a class-sensitive penalty for cross-entropy loss do not improve the separation. On the other hand, the method that gives a margin to the angle between the feature vectors and the weight vectors of the last fully connected layer corresponding to each feature can improve the separation. Therefore, we propose a loss function that can improve the separation of the important class by setting the margin only for the important class, called Class-sensitive Additive Angular Margin Loss (CAMRI Loss). CAMRI loss is expected to reduce the variance of angles between features and weights of the important class relative to other classes due to the margin around the important class in the feature space by adding a penalty to the angle. In addition, concentrating the penalty only on the important classes hardly sacrifices the separation of the other classes. Experiments on CIFAR-10, GTSRB, and AwA2 showed that the proposed method could improve up to 9% recall improvement on cross-entropy loss without sacrificing accuracy.