Investigation of Appropriate Scaling of Networks and Images for Convolutional Neural Network-Based Nerve Detection in Ultrasound-Guided Nerve Blocks.

Ultrasound imaging is an essential tool in anesthesiology, particularly for ultrasound-guided peripheral nerve blocks (US-PNBs). However, challenges such as speckle noise, acoustic shadows, and variability in nerve appearance complicate the accurate localization of nerve tissues. To address this issue, this study introduces a deep convolutional neural network (DCNN), specifically Scaled-YOLOv4, and investigates an appropriate network model and input image scaling for nerve detection on ultrasound images. Utilizing two datasets, a public dataset and an original dataset, we evaluated the effects of model scale and input image size on detection performance. Our findings reveal that smaller input images and larger model scales significantly improve detection accuracy. The optimal configuration of model size and input image size not only achieved high detection accuracy but also demonstrated real-time processing capabilities.

A Small and High-Speed Driving Mechanism for 3D Shape Measurement in Monocular Endoscopy.

Three-dimensional (3D) shape acquisition has been widely introduced to enrich quantitative analysis with the combination of object shape and texture, for example, surface roughness evaluation in industry and gastrointestinal endoscopy in medicine. Shape from focus is a promising technique to measure substance surfaces in 3D space because no occlusion problem appears in principle, as does with stereo shape measurement, which is another commonly used option. We have been developing endoscopic shape measurement devices and shape reconstruction algorithms. In this paper, we propose a mechanism for driving an image sensor reciprocated for the shape from focus of 3D shape measurement in monocular endoscopy. It uses a stepping motor and a planar-end cam, which transforms the motor rotation to imaging sensor reciprocation, to implement the shape from focus of 3D shape measurement in endoscopy. We test and discuss the device in terms of its driving accuracy and application feasibility for endoscopic 3D shape measurement.

A new robotic-assisted flexible endoscope with single-hand control: endoscopic submucosal dissection in the ex vivo porcine stomach.

BACKGROUND: Difficulties in endoscopic operations and therapeutic procedures seem to occur due to the complexity of operating the endoscope dial as well as difficulty in performing synchronized movements with both hands. We developed a prototype robotic-assisted flexible endoscope that can be controlled with a single hand in order to simplify the operation of the endoscope. The aim of this study was to confirm the operability of the robotic-assisted flexible endoscope (RAFE) by performing endoscopic submucosal dissection (ESD). METHODS: Study 1: ESD was performed manually or with RAFE by an expert endoscopist in ex vivo porcine stomachs; six operations manually and six were performed with RAFE. The procedure time per unit circumferential length/area was calculated, and the results were statistically analyzed. Study 2: We evaluated how smoothly a non-endoscopist can move a RAFE compared to a manual endoscope by assessing the designated movement of the endoscope. RESULTS: Study 1: En bloc resection was achieved by ESD using the RAFE. The procedure time was gradually shortened with increasing experience, and the procedure time of ESD performed with the RAFE was not significantly different from that of ESD performed with a manual endoscope. Study 2: The time for the designated movement of the endoscope was significantly shorter with a RAFE than that with a manual endoscope as for a non-endoscopist. CONCLUSIONS: The RAFE that we developed enabled an expert endoscopist to perform the ESD procedure without any problems and allowed a non-endoscopist to control the endoscope more easily and quickly than a manual endoscope. The RAFE is expected to undergo further development.

Improving the strength of sutureless laser-assisted vessel repair using preloaded longitudinal compression on tissue edge.

BACKGROUND AND OBJECTIVE: Little is known about the approximation of coapted edges in sutureless laser-assisted vessel welding. Tissue shrinkage by laser irradiation may cause coapted edges to separate, reducing strength of welding. This may be avoided by preloaded longitudinal compression on the tissue edges to be welded. This study compared welding strength with and without preloaded compression in ex vivo animal experiments. MATERIALS AND METHODS: This study evaluated 24 samples of harvested porcine carotid arteries, each having a length of 3 cm and an inner diameter of 1.0-2.0 mm. A half circumferential incision was made at the center of each sample. A steel shaft 2.0 mm in diameter was inserted into each sample to approximate the incised edges. The samples were longitudinally compressed to 6 mm. Incision sites were repaired by irradiation with a 970-nm diode laser. No glue or die was used. The repair strength was evaluated by measuring the bursting point (BP) of all samples. In a pilot study, the welding conditions, including power, duration, and interval of the laser spots, were tested by trial and error in 18 samples, including six treated under optimum conditions. As a control group, six samples were welded under optimum conditions, but without compression. RESULTS: Optimum conditions, consisting of 2.4 W power, 30-second duration, and 1-mm intervals of laser spots, yielded the highest BP (623 ± 236 mmHg), which was significantly higher than in the control group without compression (204 ± 208 mmHg, P = 0.009). Defining BP > 400 mmHg as successful repair, the success rates in the compression and control groups were 83% and 17%, respectively. CONCLUSION: Preloaded longitudinal compression on the coapted edges may be important for sutureless laser-assisted vessel repair and anastomosis and may affect the strength of welding. Lasers Surg. Med. 49:533-538, 2017. © 2017 Wiley Periodicals, Inc.

Loss Weightings for Improving Imbalanced Brain Structure Segmentation Using Fully Convolutional Networks.

Brain structure segmentation on magnetic resonance (MR) images is important for various clinical applications. It has been automatically performed by using fully convolutional networks. However, it suffers from the class imbalance problem. To address this problem, we investigated how loss weighting strategies work for brain structure segmentation tasks with different class imbalance situations on MR images. In this study, we adopted segmentation tasks of the cerebrum, cerebellum, brainstem, and blood vessels from MR cisternography and angiography images as the target segmentation tasks. We used a U-net architecture with cross-entropy and Dice loss functions as a baseline and evaluated the effect of the following loss weighting strategies: inverse frequency weighting, median inverse frequency weighting, focal weighting, distance map-based weighting, and distance penalty term-based weighting. In the experiments, the Dice loss function with focal weighting showed the best performance and had a high average Dice score of 92.8% in the binary-class segmentation tasks, while the cross-entropy loss functions with distance map-based weighting achieved the Dice score of up to 93.1% in the multi-class segmentation tasks. The results suggested that the distance map-based and the focal weightings could boost the performance of cross-entropy and Dice loss functions in class imbalanced segmentation tasks, respectively.

Complex-Valued Wavelet Spectrum Analysis of Respiratory Conditions and Its Feasibility in the Detection of Low-Functional Respiration.

Respiratory monitoring is a significant issue to reduce patient risks and medical staff labor in postoperative care and epidemic infection, particularly after the COVID-19 pandemic. Oximetry is widely used for respiration monitoring in the clinic, but it sometimes fails to capture a low-functional respiratory condition even though a patient has breathing difficulty. Another approach is breathing-sound monitoring, but this is unstable due to the indirect measurement of lung volume. Kobayashi in our team is developing a sensor measuring temporal changes in lung volume with a displacement sensor attached across the sixth and eighth ribs. For processing these respiratory signals, we propose the combination of complex-valued wavelet transform and the correlation among spectrum sequences. We present the processing results and discuss its feasibility to detect a low-functional condition in respiration. The result for detecting low-functional respiration showed good performance with a sensitivity of 0.88 and specificity of 0.88 to 1 in its receiver operating characteristic (ROC) curve.

Label cleaning and propagation for improved segmentation performance using fully convolutional networks.

PURPOSE: In recent years, fully convolutional networks (FCNs) have been applied to various medical image segmentation tasks. However, it is difficult to generate a large amount of high-quality annotation data to train FCNs for medical image segmentation. Thus, it is desired to achieve high segmentation performances even from incomplete training data. We aim to evaluate performance of FCNs to clean noises and interpolate labels from noisy and sparsely given label images. METHODS: To evaluate the label cleaning and propagation performance of FCNs, we used 2D and 3D FCNs to perform volumetric brain segmentation from magnetic resonance image volumes, based on network training on incomplete training datasets from noisy and sparse annotation. RESULTS: The experimental results using pseudo-incomplete training data showed that both 2D and 3D FCNs could provide improved segmentation results from the incomplete training data, especially by using three orthogonal annotation images for network training. CONCLUSION: This paper presented a validation for label cleaning and propagation based on FCNs. FCNs might have the potential to achieve improved segmentation performances even from sparse annotation data including possible noises by manual annotation, which can be an important clue to more efficient annotation.

Suction-fixing surgical device for assisting liver manipulation with laparoscopic forceps.

PURPOSE: Noninvasiveness and stability are significant issues in laparoscopic liver resection. Inappropriate grasping force can cause damage or serious bleeding to the liver. In addition, instability of grasping can result unsafe operations or wavered cutting. We propose a surgical device to improve stability of liver manipulation. METHODS: A proposed device adheres to the liver surface with suction fixation, then tunes its stiffness to being hard and shapes like as a bulge on the liver surface to be grasped with laparoscopic forceps. It consists of two soft beams, a chamber sponge, membrane covering the device upper, suburb extrusion wing membrane, a vacuuming tube and to-be-grasped bars. The beams are designed as being non-stretchable and easy to bend. The device is connected to a medical vacuuming pump to vacuum air in the device and then gets hard to transfer forceps operation well. This stiffness tuning mechanism by pneumatic control features the device for achieving good liver shape followability and forceps operation propagation less invasively. The proposed device was tested with rubber phantoms and porcine livers on shape followability, stiffness transition, liver invasiveness and operational usability in the experiments. RESULTS: Performance of the proposed device was assessed in experiments. The device showed good object-shape followability. It held the liver with 2.43-N force for vertical lifting and 4.90-N shear force with - 80 kPa vacuuming pressure. Invasiveness was reduced to acceptable level of liver damage. In usability test, the device grasped the liver stably and transferred surgical forceps operations to the liver surface well. CONCLUSION: The proposed device showed effective performance to improve laparoscopic liver manipulation. It held the liver stably and less invasively and transferred forceps operation force to the liver surface well.

Surgical Robot for Intraluminal Access: An Ex Vivo Feasibility Study.

Early-stage gastrointestinal cancer is often treated by endoscopic submucosal dissection (ESD) using a flexible endoscope. Compared with conventional percutaneous surgery, ESD is much less invasive and provides a high quality of life for the patient because it does not require a skin incision, and the organ is preserved. However, the operator must be highly skilled because ESD requires using a flexible endoscope with energy devices, which have limited degrees of freedom. To facilitate easier manipulation of these flexible devices, we developed a surgical robot comprising a flexible endoscope and two articulating instruments. The robotic system is based on a conventional flexible endoscope, and an extrapolated motor unit moves the endoscope in all its degrees of freedom. The instruments are thin enough to allow insertion of two instruments into the endoscope channel, and each instrument has a bending section that allows for up-down, right-left, and forward-backward motion. In this study, we performed an ex vivo feasibility evaluation using the proposed robotic system for ESD in a porcine stomach. The procedure was successfully performed by five novice operators without complications. Our findings demonstrated the feasibility of the proposed robotic system and, furthermore, suggest that even operators with limited experience can use this system to perform ESD.

Laparoscopic ultrasound manipulator with a spring-based elastic mechanism.

PURPOSE: Image guidance is a key technology that can improve the outcome of laparoscopic surgery. However, due to the large deformation caused by digestive organs, a computer-aided navigation system based on preoperative imaging data cannot indicate the correct target position of the lesion (e.g., liver tumors and vessels invisible from the organ surface). To overcome this issue, we developed a laparoscopic ultrasound manipulator with two motorized degrees of freedom at the tip, allowing for the performance of a dexterous ultrasound scan in a confined laparoscopic surgical area. METHOD: The developed manipulator consists of a compact and elastic structure using springs, enabling a safe ultrasound scan and avoiding excess force on the inspected organs. The manipulator is a handheld device equipped with four buttons at the handle, which the surgeon directly grasps to send a motion command to the tip structure. The developed prototype realizes two motorized degree-of-freedom motion at the tip. The size of prototype is 15.0 mm in diameter that is usable in conventional laparoscopy. The tip of the manipulator was carefully designed by considering the kinematic model and the results of the finite element analysis. RESULTS: To assess the prototype, accuracy and rigidity were measured by using a motion processing microscope. The accuracy test showed that the proposed device has a fairly accurate characteristic as a handheld device. This was supposedly caused by the nature of compliant mechanism, which does not have mechanical play in motion. In addition, the intrinsic elastic structure (approximately 2.0 N/mm in most of the range of motion) allowed the ultrasound probe to adequately fit on the curved organ surface without extra effort of manipulation during the inspection. In the in vivo experiment, the yaw motion was found to be effective for investigating the vascular network because the manipulator allows the probe to be rotated while maintaining the same position. CONCLUSION: The mechanical evaluation and in vivo test results showed high feasibility of the prototype. We are currently working on further mechanical improvement for commercialization and development of a real-time navigation system that can perform three-dimensional reconstruction of ultrasonographic images by implementing a magnetic position sensor at the tip of the manipulator.

Registration using 3D-printed rigid templates outperforms manually scanned surface matching in image-guided temporal bone surgery.

PURPOSE: Image-guided surgery (IGS) for otological procedures requires minimal invasiveness and a high degree of accuracy. We have recently developed a noninvasive registration method, the Surface Template-Assisted Marker Positioning (STAMP) method, which uses a rigid template of the surface of the temporal bone. However, the STAMP method is not applicable when the bony surface is not exposed, such as in endoscopic surgery. Thus, we extended our research to apply the STAMP method onto the skin and tested its feasibility in this study. METHODS: We designed a phantom made of a rigid box and soft material for the study. The target registration error (TRE) was measured at preset measuring points in the phantom. We modified the STAMP method to be applicable for use on the skin around the ears (S-STAMP). The same phantom was also registered using the conventional, manually scanned surface matching method. We compared the TRE after the different registration methods. RESULTS: The TRE after the S-STAMP registration method was significantly smaller than that of the conventional surface matching method at all error measurement points in the phantom. However, the TRE after the S-STAMP registration method was significantly larger than that of paired point registration using invasive fiducial markers. CONCLUSIONS: The S-STAMP method using a rigid template on the soft surface yields a significantly smaller TRE than that of conventional, manually scanned surface matching registration. This strategy provides an alternative option to improve the accuracy of IGS without loading patients with additional invasive procedures.

Fluoroscopy-based laser guidance system for linear surgical tool insertion depth control.

PURPOSE: In most orthopedic surgeries, knowing how far to insert surgical tools is crucial. The objective of this study was to provide guidance information on depth without tracking surgical tools. A previously developed laser guidance system for linear surgical tool insertion uses two laser beams that display the insertion point and orientation on the skin surface. However, the system only provides 4 degrees of freedom guidance (an entry point on the planned pathway line and the orientation) but do not inform surgeons on the ideal insertion depth. METHOD: A 5-DOF guidance method was developed to provide guidance information by direct projection onto the surgical area using laser beams without tracking markers. A position and orientation guidance performed by two laser beams and depth guidance performed by a single laser beam are appeared on the surgical area in turn. However, depth point appears on the surgical tool side face with some error because of tool radius. Using the actual depth position, insertion path vector and location of the laser sources, the correct depth point on the tool's surface is calculated by the proposed method. So, this system can indicate and navigate the 5-DOF which is planning path and the correct depth point. RESULTS: An evaluation of the accuracy of depth guidance revealed a depth guidance error of 0.55±0.29 mm and results from phantom target insertions revealed overall system accuracies of 1.44 ± 1.09 mm, 0.91° ± 0.82°. In addition, overall system accuracies of application feasibility experiment under the X-ray condition were 1.94 ± 0.98mm, 1.39° ± 1.30°. CONCLUSION: A new surgical tool depth insertion method was developed using a fluorolaser guidance system. This tool informs surgeons of the surgical tool tip depth assuming that the insertion point and orientation are correct. The new method was tested successfully in vitro.

Production of acoustic field with multiple focal points to control high amount of microbubbles in flow using a 2D array transducer.

We have newly developed a 2D array transducer to control the behavior of microbubbles, which is different from that for HIFU therapy, to emit continuous wave by designing acoustic field including multiple focal points. In the experiment using a straight path model, we have confirmed that higher concentration of acoustic energy does not result more aggregation. We also have confirmed the trapped areas of microbubbles are located not in the peak of the distribution of sound pressure, but in the middle range. The dispersion of acoustic energy is important because there was a relation in the trapping performance of microbubbles and the shape of acoustic field.

Image plane positioning by pneumatic actuators for ultrasound guidance.

Image guided procedures such as percutaneous needle insertion or high intensity focused ultrasound, have become quite widespread. In images acquisition, ultrasound (US) is convenient to use in a conventional operating room, and inexpensive compared to CT and MRI. However, US requires to handle an US probe and do not have the base coordinate system. Therefore, intraoperative image position is unclear and cannot position to interested area. To address the issues, we have developed a robotic system based on US calibration and a probe scanning robot. In this study, to validate the implement system, positioning accuracy of an image plane was evaluated. Moreover, we developed an automated US guidance system with a conventional US probe. The system enables image plane positioning to visualize a therapeutic tool automatically. From the results, positioning accuracy of the image plane was 1.6 mm and 1.5 deg, maximally. In the phantom test, the error between the positions of the image plane and the mock needle was 2.5 mm and 0.9 deg. We have confirmed that the proposed system is greatly applicable for an intraoperative US guidance.

Active control of microbubbles stream in multi-bifurcated flow by using 2D phased array ultrasound transducer.

We have previously reported our attempt to propel microbbles in flow by a primary Bjerknes force, which is a physical phenomenon where an acoustic wave pushes an obstacle along its direction of propagation. However, when ultrasound was emitted from surface of the body, controlling bubbles in against flow was needed. It is unpractical to use multiple transducers to produce the same number of focal points because single element transducer cannot produce more than two focal points. In this study, we introduced a complex artificial blood vessel according to a capillary model and a 2D array transducer to produce multiple focal points for active control of microbubbles in against flow. Furthermore, we investigated bubble control in viscous fluid. As the results, we confirmed clearly path selection of MBs in viscous fluid as well as in water.

An accurate calibration method of ultrasound images by center positions of a metal ball.

This paper provides a novel method for three-dimensional tracking of ultrasound images. One of the issues to determine the position of a ultrasound image plane is the thickness of the image plane. The proposed methodology address the issue by the calibration phantom using a fiducial sphere with the diameter of 5.5 mm because comet-trail artifact can be observed in the image plane through the center of the sphere. Meanwhile, to measure the sphere center accurately by a tracking device, a pointer tool with the same sphere at the tip is also proposed. To validate the feasibility of the method, simulation and phantom tests were conducted. From the results of the phantom test, the accuracy of the calibration was 0.65, 0.40, and 0.42 mm in 10, 50, 100 points calibration. The results demonstrate that the proposed method has a great potential for accurate US probe calibration.

A fluorolaser navigation system to guide linear surgical tool insertion.

PURPOSE: Conventional navigation systems for minimally invasive orthopedic surgery require a secondary monitor to display guidance information generated with CT or MRI images. Newer systems use augmented reality to project surgical plans into binocular glasses. These surgical procedures are often mentally challenging and cumbersome to perform. METHOD: A comprehensive surgical navigation system for direct guidance while minimizing radiation exposure was designed and built. System accuracy was evaluated using in vitro needle insertion experiments. The fluoroscopic-based navigation technique is combined with an existing laser guidance technique. As a result, the combined system is capable of surgical planning using two or more X-ray images rather than CT or MRI scans. Guidance information is directly projected onto the patient using two laser beams and not via a secondary monitor. RESULTS: We performed 15 in vitro needle insertion experiments as well as 6 phantom pedicle screw insertion experiments to validate navigation system accuracy. The planning accuracy of the system was found to be 2.32 mm and 2.28°, while its overall guidance accuracy was found to be 2.40 mm and 2.39°. System feasibility was demonstrated by successfully performing percutaneous pin insertion on phantoms. CONCLUSION: Quantitative and qualitative evaluations of the fluorolaser navigation system show that it can support accurate guidance and intuitive surgical tool insertion procedures without preoperative 3D image volumes and registration processes.