InfoColon: A dataset for consecutive informative frames in Colonoscopy.

Background modeling methods in video analysis: A review and comparative evaluation

Video scaling is a crucial video processing component that finds a variety of applications in video display, enhancement and analysis. Thus, video scaling has been an important research topic for several years now. With the emergence of hybrid-cameras that simultaneously capture a low-resolution (LR) video and periodic high-resolution (HR) stills, methods that use information in HR stills for enhancing the LR video are being studied. We present a multi-layer super-resolution (SR) method to rescale the input LR video with the help of a few HR still image guides. Resolution enhancement is achieved by blending the high frequency details into the input video frames that are scaled-up by a conventional, spatial scaling method. The missing high frequency details for each video frame are derived by spatio-temporal warping of the high frequency details extracted from HR still image guides. The Non-Local-Means (NLM) concept is used for computing the warping weights. The proposed SR method handles both global and local motions in a video, without requiring any explicit image registration. Experimental results comparing the proposed method with some of the existing video scaling techniques clearly points its strong potential.

Iris authentication is one of the most successful applications in video analysis and image processing. In this paper, several novel approaches are proposed to improve the overall performance of iris recognition systems. First, this paper proposes a new eyelash detection algorithm based on directional filters, which achieves a low rate of eyelash misclassification. Second, a multiscale and multidirection data fusion method is introduced to reduce the edge effect of wavelet transformation produced by complex segmentation algorithms. Finally, an iris indexing method on the basis of corner detection is presented to accelerate exhausted the 1: N search in a huge iris database. The performance evaluations are carried out on two popular iris databases, and the test results are experimentally more robust and accurate with less elapsed time compared with most existing methods.

Computer vision is nothing but a concept in which a computer or any other machine analyzes a scene like a human eye and performs some tasks accordingly. In the past few years, the research in the area of computer vision has increased on a very large scale due to its numerous applications in video analysis, scene detection, self - driving systems etc. The process of object detection is getting faster and better with development in the fields of neural networks and deep learning domains. In this paper, a comprehensive study is carried out on deep learning and its major component named convolution neural networks (CNN) that are largely used for object detection. At last paper shed lights on the various methods for object detection such as R-CNN, Fast R-CNN and Faster R-CNN and their applications in real world. The object detection applications has leads the world to digital revolution..

This study analyzed 26 Staphylococcus aureus strains, including 16 methicillin-resistant S. aureus (MRSA) and 10 methicillin-susceptible S. aureus (MSSA), collected from eight medical institutions in the Chiba Prefecture that requested a toxin gene analysis between 2015 and 2021. A total of 14 Panton-Valentine leukocidin (PVL) positive strains were identified, including MSSA. PVL-positive strains were classified into seven types according to polymerase chain reaction-based open reading frame typing (POT); of these types, three POT MRSA strains have not been previously reported, and one has been previously reported as PVL-negative. Some strains tested positive for both PVL and toxic shock syndrome toxin 1. One POT type was identified in both PVL-positive and PVL-negative strains. To the best of our knowledge, this is the first report on the regional spread of highly pathogenic S. aureus strains based on the POT method in children from multiple medical institutions. This method is useful for estimating the spread of toxin gene-carrying strains in the community owing to its association with toxin genes. As the number of PVL-positive strains in Japan increases, it is important to analyze the isolates of severe S. aureus infections in children by combining toxin gene analyses with the POT method.

Background: Yoga asanas help to increase the wholesome development of an individual. At the same time,the execution of the asanas gives complete benefits to the practitioners.Aim: The purpose of the study was to find out the application of video analysis for enhancing kinematicparameters and techniques of yoga performers.Methods and Material: For this study, three men yoga performers selected from SRMV Maruthi Collegeof Physical Education, Coimbatore, Tamil Nadu, India. The age of the participants was 24 years. He hada good experience in performing yoga including difficult classical asanas. For this study angle of joints,movement of body parts and position of balance, holding time were selected as dependent variables. SportsInsight motion analysis cameras and software were used to find out the angle of joints, movement of bodyparts and body balance. The baseline kinematic variables were tested before intervention. At the time ofintervention, the execution of the asana (virksasana) was recorded by the researcher through the sportsmotion analysis camera and software.Statistical analysis used: Descriptive statistics were used to analyse the performance of the yoga performers.Results: The captured video was showed to the participant to understand his execution of asana. Theresearcher gave the necessary guidance to correct the mistakes and provide opportunities for self-evaluation.After the 10 days of training, post-test was conducted and data were collected from the participant.Conclusion: It is concluded that the application of video analysis produced a significant difference inangle of joints, movement of body parts among the yoga performers. It is concluded that the application ofvideo analysis produced significant improvement in the position of balance, holding time among the yogaperformers.

Advances in video technology are being incorporated into today’s healthcare practice. For example, colonoscopy is an important screening tool for colorectal cancer. Colonoscopy allows for the inspection of the entire colon and provides the ability to perform a number of therapeutic operations during a single procedure. During a colonoscopic procedure, a tiny video camera at the tip of the endoscope generates a video signal of the internal mucosa of the colon. The video data are displayed on a monitor for real-time analysis by the endoscopist. Other endoscopic procedures include upper gastrointestinal endoscopy, enteroscopy, bronchoscopy, cystoscopy, and laparoscopy. However, a significant number of out-of-focus frames are included in this type of videos since current endoscopes are equipped with a single, wide-angle lens that cannot be focused. The out-of-focus frames do not hold any useful information. To reduce the burdens of the further processes such as computer-aided image processing or human expert’s examinations, these frames need to be removed. We call an out-of-focus frame as non-informative frame and an in-focus frame as informative frame. We propose a new technique to classify the video frames into two classes, informative and non-informative frames using a combination of Discrete Fourier Transform (DFT), Texture Analysis, and K-Means Clustering. The proposed technique can evaluate the frames without any reference image, and does not need any predefined threshold value. Our experimental studies indicate that it achieves over 96% of four different performance metrics (i.e. precision, sensitivity, specificity, and accuracy).

The performance of existing image-based training models in evaluating bowel preparation on colonoscopy videos was relatively low, and only a few models used external data to prove their generalization. Therefore, this study attempted to develop a more precise and stable AI system for assessing bowel preparation of colonoscopy video. We proposed a system named ViENDO to assess the bowel preparation quality, including two CNNs. First, Information-Net was used to identify and filter out colonoscopy video frames unsuitable for Boston bowel preparation scale (BBPS) scoring. Second, BBPS-Net was trained and tested with 5,566 suitable short video clips through three-dimensional (3D) convolutional neural network (CNN) technology to detect BBPS-based insufficient bowel preparation. Then, ViENDO was applied to complete withdrawal colonoscopy videos from multiple centers to predict BBPS segment scores in clinical settings. We also conducted a human-machine contest to compare its performance with endoscopists. In video clips, BBPS-Net for determining inadequate bowel preparation generated an area under the curve of up to 0.98 and accuracy of 95.2%. When applied to full-length withdrawal colonoscopy videos, ViENDO assessed bowel cleanliness with an accuracy of 93.8% in the internal test set and 91.7% in the external dataset. The human-machine contest demonstrated that the accuracy of ViENDO was slightly superior compared to most endoscopists, though no statistical significance was found. The 3D-CNN-based AI model showed good performance in evaluating full-length bowel preparation on colonoscopy video. It has the potential as a substitute for endoscopists to provide BBPS-based assessments during daily clinical practice.

This paper describes a novel system for real-time video texture analysis. The system utilizes hardware to extract second-order statistical features from video frames. These features are based on the Gray Level Co-occurrence Matrix (GLCM) and describe the textural content of the video frames. They can be used in a variety of video analysis and pattern recognition applications, such as remote sensing, industrial and medical. The hardware is implemented on a Virtex-XCV2000E-6 FPGA programmed in VHDL. It is based on an architecture that exploits the symmetry and the sparseness of the GLCM and calculates the features using integer and fixed point arithmetic. Moreover, it integrates an efficient algorithm for fast and accurate logarithm approximation, required in feature calculations. The software handles the video frame transfers from/to the hardware and executes only complementary floating point operations. The performance of the proposed system was experimentally evaluated using standard test video clips. The system was implemented and tested and its performance reached 133 and 532 fps for the analysis of CIF and QCIF video frames respectively. Compared to the state of the art GLCM feature extraction systems, the proposed system provides more efficient use of the memory bandwidth and the FPGA resources, in addition to higher processing throughput, that results in real time operation. Furthermore, its fundamental units can be used in any hardware application that requires sparse matrix representation or accurate and efficient logarithm estimation.

A new threat in encrypted databases, called volume leakage was reported by Kellaris et al. and Grubbs et al., where the volume of data associated with each key in a multi-map is leaked. An existing method addresses this problem by adding noise entries to the multi-map according to differential privacy so that adversaries cannot infer the volume of data. However, it assumes that the data is static and therefore does not support any update operations, such as insertion, deletion, etc. To this problem, this paper proposes a method that enables data insertion in the encrypted multi-map with volume hiding. The basic idea is to use local differential privacy instead of differential privacy in adding noise entries when performing insertions. Besides, we employ cuckoo hashing to perturb the place of insertion, thereby allowing insertions of new items without leaking the volume of entries.However, our proposed method is not suitable for large databases. We consider the use of cloud database in healthcare as a use case. The cloud database is necessary for healthcare in cases such as when multiple medical institutions want to share medical data or when home healthcare providers want to access medical data from outside of hospitals. There are privacy issues if the number of patients for each disease is known. Furthermore, when inserting data of a new patient, we want to hide which disease the patient is suffering from.We conduct experiments to assess the feasibility of the proposed method, and it presents a reasonable performance.

Objective: To evaluate the therapeutic effects of slotted manuduction on knee extension apparatus-induced traumatic adhesion in multiple medical institutions. Methods: A total of 160 patients with knee extension apparatus induced traumatic adhesion who were admitted in our hospital from January 2017 to December 2017 were selected, and randomly divided into two groups: experimental group and control group, (n=80). There were 44 males and 36 females in the experimental group, with mean age of (45.95±9.48) years old, which was treated with slotted manuduction; and 43 males and 37 females in the control group, with mean age of (46.52±9.07) years old, which received traditional treatment. The treatment was performed once per day for 30 days, with 30 minutes once a time. Informed consent was obtained from each patient. Main outcome measures: The degree of knee joint activities, score of knee joint actively flexural behavior function, and the grading of joint function prior to and after treatment, as well as therapeutic effects were compared in two groups. Results: The significant effective rate was 90.00% in the experimental group and 76.25% in the control group (P<0.05), the former group more effective than the latter group. Comparing the degree of knee joint activities, score of knee joint function, score of knee joint actively flexural behavior function, and the grading score of joint function, the experimental group was better than the control group (P<0.01). Conclusion: Slotted manuduction has an extract and safe effect on treating knee extension apparatus induced traumatic adhesion, which has high operational and promotional value.

Federated semi-supervised representation augmentation with cross-institutional knowledge transfer for healthcare collaboration

An allocation model for emergency medical services based on integrated healthcare during an epidemic

CAROTID – A web-based platform for optimal personalized management of atherosclerotic patients

Recent advancements in deep learning methods bring computer-assistance a step closer to fulfilling promises of safer surgical procedures. However, the generalizability of such methods is often dependent on training on diverse datasets from multiple medical institutions, which is a restrictive requirement considering the sensitive nature of medical data. Recently proposed collaborative learning methods such as Federated Learning (FL) allow for training on remote datasets without the need to explicitly share data. Even so, data annotation still represents a bottleneck, particularly in medicine and surgery where clinical expertise is often required. With these constraints in mind, we propose FedCy, a federated semi-supervised learning (FSSL) method that combines FL and self-supervised learning to exploit a decentralized dataset of both labeled and unlabeled videos, thereby improving performance on the task of surgical phase recognition. By leveraging temporal patterns in the labeled data, FedCy helps guide unsupervised training on unlabeled data towards learning task-specific features for phase recognition. We demonstrate significant performance gains over state-of-the-art FSSL methods on the task of automatic recognition of surgical phases using a newly collected multi-institutional dataset of laparoscopic cholecystectomy videos. Furthermore, we demonstrate that our approach also learns more generalizable features when tested on data from an unseen domain.