Two-stage video-based convolutional neural networks for adult spinal deformity classification.

Introduction: Assessment of human gait posture can be clinically effective in diagnosing human gait deformities early in life. Currently, two methods-static and dynamic-are used to diagnose adult spinal deformity (ASD) and other spinal disorders. Full-spine lateral standing radiographs are used in the standard static method. However, this is a static assessment of joints in the standing position and does not include information on joint changes when the patient walks. Careful observation of long-distance walking can provide a dynamic assessment that reveals an uncompensated posture; however, this increases the workload of medical practitioners. A three-dimensional (3D) motion system is proposed for the dynamic method. Although the motion system successfully detected dynamic posture changes, access to the facilities was limited. Therefore, a diagnostic approach that is facility-independent, has low practice flow, and does not involve patient contact is required. Methods: We focused on a video-based method to classify patients with spinal disorders either as ASD, or other forms of ASD. To achieve this goal, we present a video-based two-stage machine-learning method. In the first stage, deep learning methods are used to locate the patient and extract the area where the patient is located. In the second stage, a 3D CNN (convolutional neural network) device is used to capture spatial and temporal information (dynamic motion) from the extracted frames. Disease classification is performed by discerning posture and gait from the extracted frames. Model performance was assessed using the mean accuracy, F1 score, and area under the receiver operating characteristic curve (AUROC), with five-fold cross-validation. We also compared the final results with professional observations. Results: Our experiments were conducted using a gait video dataset comprising 81 patients. The experimental results indicated that our method is effective for classifying ASD and other spinal disorders. The proposed method achieved a mean accuracy of 0.7553, an F1 score of 0.7063, and an AUROC score of 0.7864. Additionally, ablation experiments indicated the importance of the first stage (detection stage) and transfer learning of our proposed method. Discussion: The observations from the two doctors were compared using the proposed method. The mean accuracies observed by the two doctors were 0.4815 and 0.5247, with AUROC scores of 0.5185 and 0.5463, respectively. We proved that the proposed method can achieve accurate and reliable medical testing results compared with doctors' observations using videos of 1 s duration. All our code, models, and results are available at https://github.com/ChenKaiXuSan/Walk_Video_PyTorch. The proposed framework provides a potential video-based method for improving the clinical diagnosis for ASD and non-ASD. This framework might, in turn, benefit both patients and clinicians to treat the disease quickly and directly and further reduce facility dependency and data-driven systems.

Integration of Independent Heat Transfer Mechanisms for Non-Contact Cold Sensation Presentation With Low Residual Heat.

Thermal sensation is crucial to enhancing our comprehension of the world and enhancing our ability to interact with it. Therefore, the development of thermal sensation presentation technologies holds significant potential, providing a novel method of interaction. Traditional technologies often leave residual heat in the system or the skin, affecting subsequent presentations. Our study focuses on presenting thermal sensations with low residual heat, especially cold sensations. To mitigate the impact of residual heat in the presentation system, we opted for a non-contact method, and to address the influence of residual heat on the skin, we present thermal sensations without significantly altering skin temperature. Specifically, we integrated two highly responsive and independent heat transfer mechanisms: convection via cold air and radiation via visible light, providing non-contact thermal stimuli. By rapidly alternating between perceptible decreases and imperceptible increases in temperature on the same skin area, we maintained near-constant skin temperature while presenting continuous cold sensations. In our experiments involving 15 participants, we observed that when the cooling rate was -0.2 to -0.24 °C/s and the cooling time ratio was 30 to 50%, more than 86.67% of the participants perceived only persistent cold without any warmth.

Intensity-Adjustable Non-Contact Cold Sensation Presentation Based on the Vortex Effect.

Cold sensations of varying intensities are perceived when human skin is subject to diverse environments. The accurate presentation of temperature changes is important to elicit immersive sensations in applications such as virtual reality. In this article, we developed a method to elicit intensity-adjustable non-contact cold sensations based on the vortex effect. We applied this effect to generate cold air at approximately 0 °C and varied the skin temperature over a wide range. The perception of different temperatures can be elicited by adjusting the volume flow rate of the cold air. Additionally, we introduced a cooling model to relate the changes in skin temperature to various parameters such as the cold air volume flow rate and distance from the cold air outlet to the skin. For validation, we conducted measurement experiments and found that our model can estimate the change in skin temperature with a root mean-square error of 0.16 °C. Furthermore, we evaluated the performance of a prototype in psychophysical cold discrimination experiments based on the discrimination threshold. Thus, cold sensations of varying intensities can be generated by varying the parameters. These cold sensations can be combined with images, sounds, and other stimuli to create an immersive and realistic artificial environment.

Radiography education with VR using head mounted display: proficiency evaluation by rubric method.

BACKGROUND: The use of head mounted display (HMD)-based immersive virtual reality (VR) coaching systems (HMD-VRC) is expected to be effective for skill acquisition in radiography. The usefulness of HMD-VRC has been reported in many previous studies. However, previous studies have evaluated the effectiveness of HMD-VRC only through questionnaires. HMD-VRC has difficulties in palpation and patient interaction compared to real-world training. It is expected that these issues will have an impact on proficiency. The purpose of this study is to determine the impact of VR constraints in HMD-VRC, especially palpation and patient interaction, on radiographic skills proficiency in a real-world setting. METHODS: First-year students (n = 30) at a training school for radiology technologists in Japan were randomly divided into two groups, one using HMD-VRC (HMD-VRC group) and the other practicing with conventional physical equipment (RP group) and trained for approximately one hour. The teachers then evaluated the students for proficiency using a rubric method. RESULTS: In this study, it was found that some skills in the HMD-VRC group were equivalent to those of the RP group and some were significantly lower than those of the RP group. There was a significant decrease in proficiency in skills related to palpation and patient interaction. CONCLUSIONS: This study suggests that HMD-VRC can be less effective than real-world training in radiographic techniques, which require palpation and patient interaction. For effective training, it is important to objectively evaluate proficiency in the real world, even for HMD-VRC with new technologies, such as haptic presentation and VR patient interaction. TRIAL REGISTRATION: The study was conducted with the approval of the Ethics Committee of International University of Health and Welfare (Approval No.21-Im-035, Registration date: September 28, 2021).

ConvLSTM based Estimation Method of Incision Trajectory with Electric Knife by Connecting Restored Thermal Sources.

Surgical navigation for understanding the internal structure of an organ is being actively studied, and it is necessary to estimate the incision trajectory to update the structure information dynamically. In this study, we focused on the fact that the region incised by the electric knife becomes high in temperature. Thus, we propose an estimation method of incision trajectory by restoring thermal source from diffused thermal images using a ConvLSTM and connecting the restored thermal sources. We first verified the possibility of thermal source restoration, and confirmed that the method enabled to restore the thermal source with high PSNR equivalent to 42.61. Next, we verified the accuracy of the incision trajectory from proposed method by comparing with the traditional method. The results suggested a better performance compared with the traditional method.

Hemodynamic Sensing of 3-D Fingertip Force by Using Nonpulsatile and Pulsatile Signals in the Proximal Part.

This study proposed a novel sensing method of 3-D contact force at a fingertip by using a photoplethysmogram (PPG) device on the proximal part of a finger. The proposed system detects nonpulsatile and pulsatile components of PPG signals from both sides of the proximal part, extracts 16 feature values related to the contact force, and estimates the 3-D force by using a multiple linear regression model. In the validation experiments, the participants wore a PPG device at the proximal parts of their index fingers and applied a contact force at the fingertips for the 11 types of touch actions. The results indicated that satisfactory agreements are observed between the system outputs and the reference forces by the calibrated force sensor. Moreover, the results revealed that the most effective number of feature values corresponded to six for the higher reproducible sensing. Although the development of the effective calibration method is expected to increase robustness, we realized that the proposed method can potentially be used for a 3-D input user interface.

Estimation of Object Elasticity by Capturing Fingernail Images During Haptic Palpation.

In this study, we present a system that performs natural-touch-based elasticity estimation for an object by using a depth camera. To estimate elasticity, which is defined as an object's Young's modulus, a strain-stress curve is obtained from fingernail images during haptic palpation. From a color image, the proposed system detects a fingernail and extracts 10 feature values related to the contact force; then, it estimates the force using a multiple regression model. Deformation of the object was estimated from the finger's three-dimensional position obtained from both color and depth images. Then, a strain-stress curve was determined using the force and deformation data. Evaluation experiments were designed to obtain the strain-stress curves of five objects from 10 participants; then, the estimation performance was investigated. The results show that the reliable range of sensing was within Young's modulus values of 0.12-5.6 MPa and the precision of the measurement was 55 percent of the estimated elasticity.

Force Rendering and its Evaluation of a Friction-Based Walking Sensation Display for a Seated User.

Most existing locomotion devices that represent the sensation of walking target a user who is actually performing a walking motion. Here, we attempted to represent the walking sensation, especially a kinesthetic sensation and advancing feeling (the sense of moving forward) while the user remains seated. To represent the walking sensation using a relatively simple device, we focused on the force rendering and its evaluation of the longitudinal friction force applied on the sole during walking. Based on the measurement of the friction force applied on the sole during actual walking, we developed a novel friction force display that can present the friction force without the influence of body weight. Using performance evaluation testing, we found that the proposed method can stably and rapidly display friction force. Also, we developed a virtual reality (VR) walk-through system that is able to present the friction force through the proposed device according to the avatar's walking motion in a virtual world. By evaluating the realism, we found that the proposed device can represent a more realistic advancing feeling than vibration feedback.

Hemodynamic sensing of 3D fingertip force using PPG device on proximal part.

This research proposed a novel method to estimate the 3D contact force of a fingertip using a photoplethysmogram (PPG) device on the proximal part of a finger. The proposed method detects non-pulsatile and pulsatile components of PPG signals, extracts eight feature values related to the contact force, and estimates the force by multiple linear regression. In the validation experiments, the participants wore a PPG device at the proximal part of the index finger and applied contact force. Then, the relationship between the contact force and PPG feature values was investigated. As a result, the system was found to be capable of estimating the 3D contact force with a root-mean-square error of 3.5 % of the maximum contact force.

Evaluation of rectum and bladder dose accumulation from external beam radiotherapy and brachytherapy for cervical cancer using two different deformable image registration techniques.

We evaluated dose-volume histogram (DVH) parameters based on deformable image registration (DIR) between brachytherapy (BT) and external beam radiotherapy (EBRT) that included a center-shielded (CS) plan. Eleven cervical cancer patients were treated with BT, and their pelvic and CS EBRT were studied. Planning CT images for EBRT and BT (except for the first BT, used as the reference image) were deformed with DIR to reference image. We used two DIR parameter settings: intensity-based and hybrid. Mean Dice similarity coefficients (DSCs) comparing EBRT with the reference for the uterus, rectum and bladder were 0.81, 0.77 and 0.83, respectively, for hybrid DIR and 0.47, 0.37 and 0.42, respectively, for intensity-based DIR (P < 0.05). D1 cm3 for hybrid DIR, intensity-based DIR and DVH addition were 75.1, 81.2 and 78.2 Gy, respectively, for the rectum, whereas they were 93.5, 92.3 and 94.3 Gy, respectively, for the bladder. D2 cm3 for hybrid DIR, intensity-based DIR and DVH addition were 70.1, 74.0 and 71.4 Gy, respectively, for the rectum, whereas they were 85.4, 82.8 and 85.4 Gy, respectively, for the bladder. Overall, hybrid DIR obtained higher DSCs than intensity-based DIR, and there were moderate differences in DVH parameters between the two DIR methods, although the results varied among patients. DIR is only experimental, and extra care should be taken when comparing DIR-based dose values with dose-effect curves established using DVH addition. Also, a true evaluation of DIR-based dose accumulation would require ground truth data (e.g. measurement with physical phantom).

Evaluation of deformable image registration between external beam radiotherapy and HDR brachytherapy for cervical cancer with a 3D-printed deformable pelvis phantom.

PURPOSE: In this study, we developed a 3D-printed deformable pelvis phantom for evaluating spatial DIR accuracy. We then evaluated the spatial DIR accuracies of various DIR settings for cervical cancer. METHODS: A deformable female pelvis phantom was created based on patient CT data using 3D printing. To create the deformable uterus phantom, we first 3D printed both a model of uterus and a model of the internal cavities of the vagina and uterus. We then made a mold using the 3D printed uterus phantom. Finally, urethane was poured into the mold with the model of the internal cavities in place, creating the deformable uterus phantom with a cavity into which an applicator could be inserted. To create the deformable bladder phantom, we first 3D printed models of the bladder and of the same bladder scaled down by 2 mm. We then made a mold using the larger bladder model. Finally, silicone was poured into the mold with the smaller bladder model in place to create the deformable bladder phantom with a wall thickness of 2 mm. To emulate the anatomical bladder, water was poured into the created bladder. We acquired phantom image without applicator for EBRT. Then, we inserted the applicator into the phantom to simulate BT. In this situation, we scanned the phantom again to obtain the phantom image for BT. We performed DIR using the two phantom images in two cases: Case A, with full bladder (170 ml) in both EBRT and BT images; and Case B with full bladder in the BT image and half-full bladder (100 ml) in the EBRT image. DIR was evaluated using Dice similarity coefficients (DSCs) and 31 landmarks for the uterus and 25 landmarks for the bladder. A hybrid intensity and structure DIR algorithm implemented in RayStation with four DIR settings was evaluated. RESULTS: On visual inspection, reasonable agreement in shape of the uterus between the phantom and patient CT images was observed for both EBRT and BT, although some regional disagreements in shape of the bladder and rectum were apparent. The created phantom could reproduce the actual patient's uterus deformation by the applicator. For both Case A and B, large variation was seen in landmark error among the four DIR parameters. In addition, although DSCs were comparable, moderate differences in landmark error existed between the two different DIR parameters selected from the four DIR parameters (i.e., DSC = 0.96, landmark error = 13.2 ± 5.7 mm vs. DSC = 0.97, landmark error = 9.7 ± 4.0 mm). This result suggests that landmark error evaluation might thus be more effective than DSC for evaluating DIR accuracy. CONCLUSIONS: Our developed phantom enabled the evaluation of spatial DIR accuracy for the female pelvic region for the first time. Although the DSCs are high, the spatial errors can still be significant and our developed phantom facilitates their quantification. Our results showed that optimization is needed to identify suitable DIR settings. For determining suitable DIR settings, our method of evaluating spatial DIR accuracy using the 3D-printed phantom may prove helpful.

Design and Psychophysical Evaluation of the HapSticks: A Novel Non-Grounded Mechanism for Presenting Tool-Mediated Vertical Forces.

Force feedback in tool-mediated interactions with the environment is important for the successful performance of complex tasks in daily life as well as in specialized fields such as medicine. Most stylus-based haptic devices require either grounding or attachment to the user's body. Recently, non-grounded haptic devices have attracted a growing interest. In this study, we propose a non-grounded rotation mechanism to represent the vertical forces applied on the tip of a tool by mimicking the cutaneous sensations that are caused by such forces. As an example of an application of our method, we developed a non-grounded haptic device called HapSticks, which mimicked the sensation of manipulating objects using chopsticks. First, using an adjustment paradigm, we directly compared a virtual weight rendered by our device and a real weight to investigate the relation of real weight and virtual weight. Next, we used a forced choice constant stimuli paradigm in a virtual and a real weight discrimination task. We conclude that our novel device renders a reliable illusion of sensed weight that leads to a discrimination ability that is typical of virtual-reality applications but worse than the discrimination between real weights.

Estimation of Finger Joint Angles Based on Electromechanical Sensing of Wrist Shape.

An approach to finger motion capture that places fewer restrictions on the usage environment and actions of the user is an important research topic in biomechanics and human-computer interaction. We proposed a system that electrically detects finger motion from the associated deformation of the wrist and estimates the finger joint angles using multiple regression models. A wrist-mounted sensing device with 16 electrodes detects deformation of the wrist from changes in electrical contact resistance at the skin. In this study, we experimentally investigated the accuracy of finger joint angle estimation, the adequacy of two multiple regression models, and the resolution of the estimation of total finger joint angles. In experiments, both the finger joint angles and the system output voltage were recorded as subjects performed flexion/extension of the fingers. These data were used for calibration using the least-squares method. The system was found to be capable of estimating the total finger joint angle with a root-mean-square error of 29-34 degrees. A multiple regression model with a second-order polynomial basis function was shown to be suitable for the estimation of all total finger joint angles, but not those of the thumb.

Feature detection in biological tissues using multi-band and narrow-band imaging.

PURPOSE: In the past decade, augmented reality systems have been expected to support surgical operations by making it possible to view invisible objects that are inside or occluded by the skull, hands, or organs. However, the properties of biological tissues that are non-rigid and featureless require a large number of distributed features to track the movement of tissues in detail. METHODS: With the goal of increasing the number of feature points in organ tracking, we propose a feature detection using multi-band and narrow-band imaging and a new band selection method. The depth of light penetration into an object depends on the wavelength of light based on optical characteristics. We applied typical feature detectors to detect feature points using three selected bands in a human hand. To consider surgical situations, we applied our method to a chicken liver with a variety of light conditions. RESULTS: Our experimental results revealed that the image of each band exhibited a different distribution of feature points. In addition, the total number of feature points determined by the proposed method exceeded that of the R, G, and B images obtained using a normal camera. The results using a chicken liver with various light sources and intensities also show different distributions with each selected band. CONCLUSIONS: We have proposed a feature detection method using multi-band and narrow-band imaging and a band selection method. The results of our experiments confirmed that the proposed method increased the number of distributed feature points. The proposed method was also effective for different light conditions.

Postoperative gluteal skin damage associated with latent development of gluteal muscle damage.

Preceding this study, we observed two cases of concurrent postoperative gluteal skin and muscle damage with extremely high serum creatine kinase (CK) levels, both of which were unrelated to pressure-induced tissue injury. However, postoperative gluteal skin damage accompanied by gluteal muscle damage has not been previously reported and the association between gluteal skin damage, gluteal muscle damage and pressure-induced tissue injury has not previously been investigated. Therefore, we conducted this study to determine the postoperative incidence of gluteal skin damage associated with gluteal muscle damage and assess associations with postoperative serum CK levels and pressure-induced tissue injury. We prospectively evaluated postoperative incidence of gluteal skin damage and measured serum CK levels in 929 consecutive patients who underwent abdominal, urological or gynecological surgery at our hospital. Magnetic resonance imaging (MRI) of the pelvis was performed in 67 patients who consented. As a result, two of 929 patients developed postoperative gluteal skin damage accompanied by gluteal muscle damage. Gluteal muscle damage without gluteal skin damage was observed in 23 of the 67 patients who underwent MRI, and volumes of damaged gluteal muscle and postoperative serum CK levels were positively correlated. Both gluteal skin and muscle damage were distinguishable from pressure-induced tissue injury. Based on the results of this study, we could confirm the occurrence of postoperative gluteal skin damage, distinct from pressure sores, accompanied by gluteal muscle damage. We also revealed latent development of postoperative gluteal muscle damage, distinguishable from compression-induced tissue injury, without accompanying gluteal skin damage.

NGF/TrkA Signaling as a Therapeutic Target for Pain.

Nerve growth factor (NGF) was first discovered approximately 60 years ago by Rita Levi-Montalcini as a protein that induces the growth of nerves. It is now known that NGF is also associated with Alzheimer's disease and intractable pain, and hence, it, along with its high-affinity receptor, tropomyosin receptor kinase (Trk) A, is considered to be 1 of the new targets for therapies being developed to treat these diseases. Anti-NGF antibody and TrkA inhibitors are known drugs that suppress NGF/TrkA signaling, and many drugs of these classes have been developed thus far. Interestingly, local anesthetics also possess TrkA inhibitory effects. This manuscript describes the development of an analgesic that suppresses NGF/TrkA signaling, which is anticipated to be 1 of the new methods to treat intractable pain.

Electrotactile Augmentation for Carving Guidance.

In this study, we presented an efficient and unobtrusive tactile feedback system, which is used to train dental technicians in carving tasks using a wax stick and knife. First, we developed a method for generating performance metrics using a model-based estimation of clearance angles between an object's surface and the carving blade. The calculated clearance angles are compared with desired angles obtained from expert operators. Then, angular errors are presented as tactile cues to the user's finger pads through electrical stimuli at the middle phalanx of the index finger and the thumb. Subsequently, we conducted a feasibility test with novice dental technicians, who showed improvement in initial clearance angles of carving strokes. Moreover, the results showed significant reduction in the occurrence rate of poor-carving when using the proposed system. From these results, we concluded that electrotactile augmentation can provide effective guidance for carving tasks.

Noninvasive simultaneous measurement of blood pressure and blood flow velocity for hemodynamic analysis.

We describe a noninvasive and simultaneous measurement method of beat-by-beat blood pressure and blood flow velocity waveforms in the radial artery using tonometry and Doppler flowmetry. We conducted a subjective experiment in which hold-down pressure of tonometry was controlled for determining optimal hold-down pressure and the measurement accuracy under the optimal hold-down pressure was evaluated. As a result, blood pressure and blood flow velocity could be measured simultaneously without the influence of the hold-down pressure on the blood flow velocity. It was possible to analyze hemodynamic indicators, such as wave intensity and vascular impedance, with blood pressure and blood flow using the system. The proposed system for detecting unexpected fluctuations in blood pressure and the involved mechanisms may contribute to the treatment of cardiovascular diseases.

Suppressive effect of membrane-permeable peptides derived from autophosphorylation sites of the IGF-1 receptor on breast cancer cells.

Insulin-like growth factor-1 (IGF-1) receptors play a crucial role in the biology of human cancer, making them an attractive target for anti-cancer agents. We previously designed oligopeptides containing the amino-acid sequences surrounding the autophosphorylation sites of the insulin receptor and found that two of them, namely, Ac-DIYET-NH2 and Ac-DYYRK-NH2, suppressed phosphorylation of purified insulin receptors in a non-ATP-competitive manner, whereas Ac-NIYQT-NH2 and Ac-NYYRK-NH2 suppressed in an ATP-competitive manner. Because the IGF-1 receptor is closely related to the insulin receptor, the aim of this study was to observe the effects of these peptides, which correspond to the amino-acid sequences of the autophosphorylation sites of the IGF-1 receptor, on the activity of the human breast cancer cell lines MCF-7, T47D, MDA-MB-231, and MDA-MB-453. To facilitate peptide delivery into breast cancer cells, the cell-penetrating peptide, human immunodeficiency virus type 1-transactivator of transcription (Tat), was linked to these peptides. When breast cancer cells were treated with each of these synthetic Tat-conjugated peptides, the conjugated peptides penetrated into the cells and suppressed cell proliferation. An inhibitory effect of Tat-conjugated peptides against IGF-1-stimulated phosphorylation of IGF-1 receptors was observed. In addition, we found that combinations of these peptides suppressed phosphorylation of IGF-1 receptors to a greater extent than the peptides did individually. In conclusion, IGF-1 receptor autophosphorylation site-derived membrane-permeable peptides have the potential to suppress IGF-1 receptor function in breast cancer cells and to be developed into novel and useful agents for cancer therapy.

Material Roughness Modulation via Electrotactile Augmentation.

Tactile exploration of a material's texture using a bare finger pad is a daily human activity. However, modern tactile displays do not allow users to experience the natural sensations of a material when artificial sensations are presented. We propose an electrotactile augmentation technique capable of superimposing vibrotactile sensations in a finger pad, thereby allowing the texture modulation of real materials. Users attach two stimulus electrodes to the middle phalanx of a finger and a grounded electrode at the base of the finger in order to evoke nerve activity. This paper evaluates the proposed electrotactile augmentation for roughness modulation of real materials. First, we introduce the principle of the electrotactile display, which presents artificial sensations at the finger pad. We then confirm that the perceived frequency of mechanical vibration at the finger pad can be shifted using electrotactile augmentation. Finally, we discuss a user study, wherein participants rated the roughness of real materials explored using the proposed system. Experimental results indicate that fine- and macro-roughness perceptions of real materials can be altered using electrotactile augmentation.