With the continuous development of science, technology and improvement of people's living standards, the requirements for differentiated medical care, treatment quality and quality of life are gradually improved. People not only care about the treatment results, but also the comfort, privacy protection and minimally invasive during the treatment. Thus, minimally invasive surgery (MIS) techniques have achieved significant improvement in recent years. In MIS, doctors use a variety of techniques to operate with less damage to the body than with open surgery. In general, MIS is associated with less pain, a shorter time to be hospitalized and fewer complications than conventional techniques. Robotic surgery is one of the most typical MIS types, which provides a magnified, 3 D view of the surgical site and helps the surgeon operate with precision, flexibility and control. Another typical MIS type is laparoscopy – surgery is done through one or more small incisions, using small tubes and tiny cameras and surgical instruments. Other typical MIS types include endoscopic surgery, interventional surgery, etc. Nevertheless, image processing and signal processing are important for the surgery-related and clinical application. Novel clinical measurement and advanced signal processing methods are important or even essential in the clinic. In this supplement issue, researchers have addressed the recent technology advances in MIS, surgery technology, clinical measurement, device development, modeling and signal processing.
Jinao Zhang et al. proposed a temperature distribution and thermal damage prediction model in soft tissues according to a combination of Galerkin method and explicit forward finite difference method. In order to address the expressive computation load involved in finite element analysis, graphics processing unit (GPU) acceleration was implemented using the high-level shader language, leading to a maximum reduction of 55.3 times in computation time comparing to standalone central processing unit (CPU) execution. Tissue at the heat source point can be quickly necrotized in a matter of seconds, while it required several minutes to fully necrotize the entire neighboring tissues.
Weijian Ren et al. compared three different minimally invasive techniques, percutaneous endoscopic lumbar discectomy (PELD), mis-tlif combined with a contralateral translaminar screw (MIS-TLIF CTS) and mis-tlif combined with bilateral pedicle screws (MIS-TLIF BPS) for the therapy of far lateral disc herniation in middle-aged and elderly patients via a retrospective chart view. The test contained various features from 74 patients, such as the length of the incision, duration of the operation, estimated blood loss, hospitalization time, operation cost, recurrence rate,. Results showed that the PELD method was better than the other two comparable ones.
Zijian Zhao et al. proposed an automatic real-time algorithm for 2 D tool detection and tracking on the basis of a spatial transformer network and spatiotemporal context in computer-assisted MIS. This study achieved great performance with respect to the accuracy and the speed after compared with other four general visual tracking methods in eight existing datasets.
Laparoscopy surgery is an important trend in the future development of surgical methods since it has advantages of small trauma and rapid recovery Its’ role in the surgical treatment of diseases has attracted more and more attention. Ke Xu et al. implemented a study aiming to 3 D reconstruction of the surgical scene based on the disparity map generated by the depth estimation algorithm. They took unsupervised learning autoencoder method to calculate the accurate disparity with a 101-layer residual convolutional network. The method was validated on a Hamlyn center laparoscopic/endoscopic video dataset, which reported the potential to offer dense disparity map and can meet surgical real-time requirement.
The existing surgical robots for laparoscopic surgery offer no or limited force feedback, and there are many problems for the traditional sensor-based solutions. Baoliang Zhao et al. built a teleoperation surgical system and validate the effectiveness of sensorless force feedback. The success rate of stiffness differentiation experiment was 96 and 80% for the tumor detection experiment. The results showed that the new system was available for surgeons regain tactile information and distinguish between the healthy and cancerous tissue.
Zhonghao Han et al. presented a precise and efficient targeting method for robot-assisted percutaneous needle placement under C-arm fluoroscopy. A special end-effector was constructed to perform fluoroscopy calibration and robot to image-space registration simultaneously and automatically. In addition, formulations were given to compute the movement of robot targeting and evaluate targeting accuracy using only one X-ray image. Pre-clinical experiment showed that the maximal angle error was 0.94° and the maximal position error of a target located 80 mm below the end-effector was 1.31 mm.
Medical imaging technology plays a decisive role in the development of modern medicine. A complete inspection requires the collection of hundreds of 2 D data. However, the high dimensionality of data collection has made it difficult to understand such a large amount of information relying on traditional two-dimensional expression, let alone efficient interpretation and analysis. Medical image analysis has become more and more widely studied. It is no longer limited to diseases with obvious diagnostic features in the past. It begins to expand into images of many different organs, anatomical and functional processes and attempts to use computer-aided image analysis with automatic and accurate quantification as well. The main contents of medical image analysis include image registration and information fusion, image visualization, functional analysis of time series images, and content-based image retrieval.
Li Yao et al. developed a novel technique to distinguish between an original image and its’ histogram equalized version according to concepts of histogram equalization and superpixel segmentation. This research can assist in image enhancement process for differentiating between enhanced/histogram equalized and original images.
Ji Zhang et al. proposed a method for characteristics analysis and recognition of digitized tongue pictures and tongue coating texture according to fractal theory in traditional Chinese medicine. They implemented it via a three-layers back propagation (BP) neural network with 587 digitized tongue coating. Results showed the simple fractal dimension is sensitive to the thin/thick and greasy characteristics of digitized tongue pictures and could judge the thickness of tongue coating well. The use of BP neural network could effectively increase accuracy rate of judgment of tongue coating texture.
Liu Shuang et al. came up with a multi-feature fusion method for medical image retrieval using wavelet and bag-of-feature. Wavelet decomposition was adopted to generate different resolution images. Bag-of-feature, texture, and other feature are extracted from three different-level wavelet images. Then generate a similarity measure function with these three features. The proposed multi-feature fusion method can achieve a high retrieval accuracy with an acceptable retrieval time.
To improve the quality of the super-resolution reconstructed medical images, Fang Zhang et al. gave an improved adaptive multi-dictionary learning method, which used the combined information of the medical image itself and the natural images database. A pyramid was generated by the self-similarity of low-resolution images, the upper layer images were used in training dictionary section and the top layer image was taken as the initial reconstruction image. This method can make full use of the same scale and different scale with similar information of medical images. Simulation experiments were validated on both natural images and medical images. The proposed method was effective for improving the effect of medical image super-resolution reconstruction.
Some minimally invasive assist devices were brought. Shidong Zhu et al. developed a novel intra-ventricular assist device whose key parameter, such as, stroke volume was precisely measured under different afterloads (60, 80, 100, and 120 mmHg), drive pressure (from 90 to 300 mmHg at 30 mmHg intervals), and heart rate (45 to 150 beats per minute). Stroke volume increased with increasing drive pressure but decreased with an increasing peripheral resistance and kept consistent with the native heart. When set this device to a heart failure mode in couple with mock system, the device improved left ventricular pressure from 106 to 158 mmHg, and stroke volume from 25.5 to 44 ml at 90 bpm.
Haisheng Zhang et al. developed a non-contact and non-invasive method using a microstrip antenna to construct the detection system for cerebral hemorrhage. They carried out an animal experiment with 13 rabbits in different bleeding states: 1, 2 and 3 ml. All measured data was sent to support vector machine (SVM) to assess the severity of cerebral hemorrhage. The -10 dB workband of the antenna was 1.55–2.05 GHz and the frequency range of the transmission parameters S21 above -30 dB is 1.2–3 GHz according to the test results. In the animal experiment, the phase difference of Transmission coefficient S21 was gradually increased with the increase of bleeding volume.
Xinghui Li et al. used three different types of synchronous control methods based on physiological parameters, electrocardiogram (ECG), fixed systole, ECG, aortic pressure, ECG and phonocardiogram (PCG) to trigger a proposed intra-ventricular assist device (iVAD) in vitro circulation system. The data showed that all three types of synchronous control methods produced trigger pulses synchronously and that the iVAD synchronously beat with the native heart. The combination of three synchronization methods can be applied to the iVAD to improve the reliability of synchronization.
Neural substrates of action to the object or this specific direct route, however, remain unclear, especially for the connection from the visual pathway to the motor cortex. Zijian Wang et al. conducted a study examined this issue by conducting a Functional Magnetic Resonance Imaging (fMRI) experiment, in which two action generation tasks involving pictures of real objects and the object’s nouns were used, with pictures naming and covert noun reading being the control tasks. The result showed that the model predefined for the Posterior Cingulate Cortex (PCC) and precuneus connecting Inferior Parietal Lobule (IPL) to the posterior medial frontal cortex dominated over the others (with 0.45 probability), suggesting that the PCC and the precuneus locate at the neural substrates of action to the object.
In the 3 D reconstruction area, Zhen Shen et al. presented a general and efficient procedure to process the digital skull data to make the printed structures meet the requirements of anatomy education, which combines the use of five 3 D manipulation tools and the procedure can be finished within 6 hours. The model was printed and compared with the cadaveric skull from frontal, left, right and anterior views respectively. The printed model can describe the correct structure and details of the skull clearly, which can be considered as a good alternative to the cadaveric skull.
Ziteng Liu et al. designed a cost-effective Virtual Reality (VR) application system for demonstration of 3 D virtual anatomical models with non-contact interaction and stereo display. The system was integrated with hand gesture interaction and voice interaction to achieve non-contact interaction. A local relationship database was designed to record the anatomical terminologies to speech recognition engine to query these uncommon words. The results showed that their system was more efficient than traditional interactive manner and verified the feasibility and practicability in the sterile operating room.
Yawei Wang et al. discovered a simple geometric method for 3 D morphology reconstruction of a cell according to two orthogonal phase images. The edges of a cell in two orthogonal projection directions were extracted from the corresponding phase images. Then, with the help of geometric rotation method, 3 D surfaces of the cell membrane and its nuclei were reconstructed respectively. Simulation and experimental results demonstrated the validity and accuracy of this method for spheroid cells.
Continuous wave near-infrared spectroscopy (CW-NIRS) can be used to measure cerebral activity because it is noninvasive, simple and portable. However, the performance of the CW-NIRS is distorted by the presence of extracerebral layer. Change of optical parameters in the gray matter layer is inappropriately converted into the brain activity response. In the study of Yan Zhang et al., a five-layer structure model constitute of scalp, skull, cerebrospinal fluid, gray matter and white matter was adopted and the mixture of the Intralipid, India ink and agar were applied to fabricate human brain tissue. The NIRS measurement system was designed to detect the changes in the absorption coefficients of the gray matter and quantitatively analyses the influence of the extracerebral layers. Monte Carlo technique was performed to compensate the partial volume effect introduced by the extracerebral layers. The results of the in-vitro experiments showed that the measured absorption coefficients are about 9% of the standard value and the relative error is about 91% due to the extracerebral layers.
Biomechanics modeling is critical in surgery using surgery robot assistants. However, due to the force of the surgical instruments on the spine, there is a corresponding deformation in the surgical field, which affects the accuracy of the operation. In order to improve the accuracy and safety of the operation. Weixin Guan et al. reconstructed the three-dimensional model of the lumbar spine which included the vertebral body and the intervertebral disc based on the computed tomography (CT) scan data, and then the lumbar spine was analyzed by the finite element method. The mathematical model of the relationship between force and displacement was established by using response surface methodology based on the simulation results. Through the simulation of the control system, the trajectory curve of the end of the manipulator was compared and the validity of the mathematical model was verified.
Physiological signal processing is also involved in this special issue. Electroencephalogram (EEG) signal has been widely used in the research of physiology, psychology, neuroscience, and cognitive science. Surface EEG signal is very weak and extremely susceptible to interference during the collection process, resulting in the unavailability of EEG data usually. Xin Liu et al. developed a method for reducing interference of breathing and cardiac contraction during near infrared spectroscopy brain activity measurement. A novel solution was used by combining the recursive least squares adaptive filtering method with the least squares SVM to suppress physiological interference both in the superficial layers and deeper layers of the head tissue, which obtained a better performance than the traditional recursive least squares method.
Dan Liu et al. focused on the evaluation of EEG signal quality using a developed model, employs amplitude, power frequency ratio, and alpha band power spectral density (PSD) ratio of resting EEG as evaluation factors, and performed a quantitative assessment of the signal quality. It turned out to be effective both before and after short-term resting EEG collection.
Denoising is a most important part in ECG signal processing, Zhaoyang Wang et al. constructed a modified wavelet design with optimized filter coefficients obtained by approximating the amplitude-frequency response of the ideal filter. This work had proven ability to remove the high-frequency component and enhance the dominating complexes in clinical data test.
Overall, MIS and minimally invasive medicine are still in development, there is a lot of room to improve for minimally device and surgery techniques. With the advancement of drugs or the rapid development of micro-robot technology, surgery is becoming more and more invasive, and even some diseases may be cured without surgical intervention. The advancement of surgical treatment is constantly evolving in the interaction between technology and concept progress. With the overall technological leap of human beings, such as computer intelligence technology, cloud computing, virtual reality technology, new materials and life sciences, endoscopic mirror technology, etc., all innovate clinical treatment methods, while at the same time acting on medical concepts. The clinical application of ‘digital human’ anatomy, (CT/MRI), robot and ‘precise medicine’ has made the surgical diagnosis and treatment more individualized and precise in the constant standardization and micro-invasiveness. In short, minimalist surgery, a cross-century new technological revolution in the field of human surgery, is sweeping the globe.
Disclosure statement
No potential conflict of interest was reported by the authors.