Unedited Videos Research Articles

A novel approach for long-term secure storage of domain independent videos

Long-term protection of multimedia contents is a complex task, especially when the video has critical elements. It demands sophisticated technology to ensure confidentiality. In this paper, we propose a blended approach which uses proactive visual cryptography scheme along with video summarization techniques to circumvent the aforementioned issues. Proactive visual cryptography is used to protect digital data by updating periodically or renewing the shares, which are stored in different servers. And, video summarization schemes are useful in various scenarios where memory is a major concern. We use a domain independent scheme for summarizing videos and is applicable to both edited and unedited videos. In our scheme, the visual continuity of the raw video is preserved even after summarization. The original video can be reconstructed through the shares using auxiliary data, which was generated during video summarization phase. The mathematical studies and experimental results demonstrate the applicability of our proposed method.

Initial surgical performance in robot-assisted radical prostatectomy is associated with clinical outcomes and learning curves.

The association between surgical performance ratings and clinical outcomes in robotic surgery is poorly understood. Additionally, no studies have reported on the relationship between the surgeon's initial case-skill evaluation and the learning curve in robot-assisted surgery. We evaluated whether an objective surgical technique evaluation score for initial robot-assisted radical prostatectomy (RARP) was associated with clinical outcomes and surgeons' learning curves. Six surgeons who were trained in and started to perform RARP at our institution were included. Anonymized, unedited videos of each surgeon's 10th RARP case were evaluated by three reviewers, using modified Objective Structured Assessment of Technical Skill (OSATS) scores. We then divided the surgeons into two groups on the basis of these OSATS scores. We retrospectively compared the clinical outcomes and learning curves of the console time of the two groups for consecutive RARPs, performed from March 2018 to July 2023. We analyzed 258 RARPs (43 cases/surgeon), including 129 cases performed by high-OSATS score surgeons (18.2-19.3 points) and 129 cases performed by low-OSATS score surgeons (11.9-16.0 points). Overall, the high-OSATS score group had significantly shorter operation and console times than the low-OSATS score group did (both P < 0.01) and their patients' rate of continence recovery by 3months post-RARP was significantly higher (P = 0.03). However, complications, blood loss, and positive margins did not differ between the groups (P = 0.08, P = 0.51, and P = 0.90, respectively). The high-OSATS score group had a significantly shorter console time than the low-OSATS score group did after the 11-20 cases. The OSATS score in early RARP cases can predict subsequent surgical outcomes and surgeons' learning curves.

MDJ: A multi-scale difference joint keyframe extraction algorithm for infrared surveillance video action recognition

Many action recognition methods require significant computational resources to achieve good results on unedited videos. However, their performance on infrared videos, which contain less information, is often unsatisfactory. In this paper, we propose a multi-scale difference joint key frame extraction algorithm for action recognition in infrared surveillance videos. To evaluate our algorithm, we have created a self-built dataset comprising 1200 unedited infrared surveillance videos categorized into 10 action categories. Our algorithm extracts key frames by jointly analyzing the global and local differences between adjacent frames. Experimental analysis demonstrates that by using only 10 frames, our algorithm improves the accuracy of generic action recognition algorithms by more than 10% on both our self-built dataset and the Infrared-Visible dataset. Moreover, our proposed method achieves high recognition accuracy with minimal computational cost, even when using a small number of frames. It outperforms state-of-the-art methods by 1.82% on the Infrared-Visible dataset and 1.35% on the InfAR dataset. These results highlight the effectiveness of our algorithm as a preprocessing module to significantly enhance the accuracy of action recognition algorithms prior to employing generic models.

Knitting Ladies Online

Knitting Ladies Online

A novel anatomical description of the esophagus: the supracarinal mesoesophagus.

During thoracoscopic esophageal resection, while performing the supracarinal lymphadenectomy along the left recurrent laryngeal nerve (LRLN) from the aortic arch to the thoracic apex, we observed a not previously described bilayered fascia-like structure, serving as prolongation of the already known mesoesophagus. We retrospectively evaluated 70 consecutively unedited videos of thoracoscopic interventions on esophageal resections for cancer, in order to determine the validity of this finding and to describe its utility for performing a systematic and more accurate dissection of the LRLN and its adequate lymphadenectomy. After mobilization of the upper esophagus from the trachea and tilting the esophagus by means of two ribbons, a bilayered fascia was observed between the esophagus and the left subclavian artery in 63 of the 70 patients included in this study. By opening the right layer, the left recurrent nerve became visualized and could be dissected free in its whole trajectory. Vessels and branches of the LRLN were divided between miniclips. Mobilizing the esophagus to the right, the base of this fascia could be found at the left subclavian artery. After dissecting and clipping the thoracic duct, complete lymphadenectomy of 2 and 4L stations could be performed. Mobilizing the esophagus in distal direction, the fascia continued at the level of the aortic arch, where it had to be divided in order to mobilize the esophagus from the left bronchus. Here, a lymphadenectomy of the aorta-pulmonary window lymph nodes (station 8) can be performed. It seems that from there the fascia continued without interruption with the previously described mesoesophagus between the thoracic aorta and the esophagus. Here we described the concept of the supracarinal mesoesophagus on the left side. Applying the description of the mesoesophagus will create a better understanding of the supracarinal anatomy, leading to a more adequate and reproducible surgery.

MP25-15 INTERNATIONAL EXPERTS CONSENSUS ON PERFORMANCE METRICS FOR A TRANSURETHRAL RESECTION OF BLADDER

MP25-15 INTERNATIONAL EXPERTS CONSENSUS ON PERFORMANCE METRICS FOR A TRANSURETHRAL RESECTION OF BLADDER

Serbian National Training Programme for minimally invasive colorectal surgery (LapSerb): short-term clinical outcomes of over 1400 colorectal resections.

The Serbian National Training Programme for minimally invasive colorectal surgery (LapSerb) was introduced to implement laparoscopic colorectal surgery across Serbia. The programme aimed to accelerate training of established colorectal surgeons through a competency-based programme. This involved knowledge assessment, workshops, live operating, and competency-based assessment of unedited videos. The aim of this study is to report the outcomes of laparoscopic colorectal resection performed by LapSerb certified surgeons. LapSerb prospectively maintained multicentred database was analysed for laparoscopic colorectal resections from January 2015 to February 2021. Data collected included patient demographics, indications for surgery, perioperative data, and 30-day outcomes. A total of 1456 laparoscopic colectomies by 24 certified surgeons were included in the final analysis. Mean age was 67 (± 12) years old and male to female ratio was 1:1.5. 83.1% of the colectomies were malignant, mainly due to adenocarcinoma. Anterior resection was the most common procedure with 699 (48%) cases, followed by right and left colectomies with 357 (24.5%) and 303 (21%) procedure respectively. 4.8% of patients required conversion to open surgery. Thirty-day readmission and reoperation rates were 2.3% and 4.7%, respectively. Overall mortality in all cases was 1.1% and R0 resections were achieved in 97.8% of malignant colectomies. The LapSerb programme successfully and safely established laparoscopic colorectal surgery across the country with comparable and acceptable short-term clinical outcomes.

International Expert Consensus on Metric-based Characterization of Robot-assisted Partial Nephrectomy

BackgroundRobot-assisted partial nephrectomy (RAPN) training usually takes place in vivo, and methods vary across countries/institutions. No common system exists to objectively assess trainee ability to perform RAPN at predetermined performance levels prior to in vivo practice. The identification of objective performance metrics for RAPN training is a crucial starting point to improve training and surgical outcomes. ObjectiveWe sought to identify objective performance metrics that best characterize a reference approach to RAPN, and obtain face and content validity from procedure experts through a modified Delphi meeting. Design, setting, and participantsDuring a series of online meetings, a core metrics team of three RAPN experts and a senior behavioral scientist performed a detailed task deconstruction of a transperitoneal left-sided RAPN procedure. Outcome measurements and statistical analysisBased on published guidelines, manufacturers’ instructions, and unedited videos of RAPN, the team identified performance metrics that constitute an optimal approach for training purposes. The metrics were then subjected to an in-person modified international Delphi panel meeting with 19 expert surgeons. Results and limitationsEleven procedure phases, with 64 procedure steps, 43 errors, and 39 critical errors, were identified. After the modified Delphi process, the international expert panel added 13 metrics (two steps), six were deleted, and three were modified; 100% panel consensus on the resulting metrics was obtained. Limitations are that the metrics are applicable only to left-sided RAPN cases and some might have been excluded. ConclusionsPerformance metrics that accurately characterize RAPN procedure were developed by a core group of experts. The metrics were then presented to and endorsed by an international panel of very experienced peers. Reliable and valid metrics underpin effective, quality-assured, structured surgical training for RAPN. Patient summaryWe organize a meeting among robot-assisted partial nephrectomy (RAPN) experts to identify and reach consensus on objective performance metrics for RAPN training. The metrics are a crucial starting point to improve and quality assure surgical training and patients’ clinical outcomes.

An Efficient Framework for Video Documentation of Bladder Lesions for Cystoscopy: A Proof-of-Concept Study.

Processing full-length cystoscopy videos is challenging for documentation and research purposes. We therefore designed a surgeon-guided framework to extract short video clips with bladder lesions for more efficient content navigation and extraction. Screenshots of bladder lesions were captured during transurethral resection of bladder tumor, then manually labeled according to case identification, date, lesion location, imaging modality, and pathology. The framework used the screenshot to search for and extract a corresponding 10-seconds video clip. Each video clip included a one-second space holder with a QR barcode informing the video content. The success of the framework was measured by the secondary use of these short clips and the reduction of storage volume required for video materials. From 86 cases, the framework successfully generated 249 video clips from 230 screenshots, with 14 erroneous video clips from 8 screenshots excluded. The HIPPA-compliant barcodes provided information of video contents with a 100% data completeness. A web-based educational gallery was curated with various diagnostic categories and annotated frame sequences. Compared with the unedited videos, the informative short video clips reduced the storage volume by 99.5%. In conclusion, our framework expedites the generation of visual contents with surgeon's instruction for cystoscopy and potential incorporation of video data towards applications including clinical documentation, education, and research.

Basic motion behavior recognition of single dairy cow based on improved Rexnet 3D network

• Rexnet was extended to Rexnet 3D for basic motion behavior recognition of dairy cows. • Sliding window mechanism were used to verify the effectiveness of untrimmed videos. • Rexnet 3D could be used for recognizing the basic motion behavior of dairy cows. Basic motion behaviors of dairy cows (lying, standing and walking) contain abundant health information. However, to monitor these behaviors accurately and efficiently by using camera-based methods remains challenging. In this study, time dimension was added to extend the Rexnet to Rank eXpansion Network 3D algorithm (Rexnet 3D) network to realize non-contact and automatic recognition of dairy cows’ basic motion behaviors. A total of 406 videos containing 256,500 frames of dairy cows in different scenes and postures were collected and tested. Six different networks including Resnet101, Mobilev2, Mobilev3, Shufflev2, C3D and S3D were selected to validate the performance of the proposed network. The average number of frames recognized per second (ANFR), FLOPs and Networks Weight were used to evaluate the performance of the method. Experimental results showed that behavior recognition accuracy, ANFR, FLOPs and Networks Weight of the proposed method were 95.00%, 76.52 fps, 15.80 G and 14.30 M, respectively. Meanwhile, a total of 10 unedited videos with 1,722 s in total (43,065 frames) were collected to verify the effectiveness in real environment. The results showed that the accuracy of the proposed method was higher when the sampling interval was five frames and the accuracy of the specific three basic motion behaviors was 91.02%, and the ANFR was 101.02 fps. In general, these findings indicated that it was feasible to use a designed network for accurate recognition of cows’ basic motion behaviors. The network could provide necessary reference for improving the precision breeding and health welfare of dairy cows.

Surgical skill and oncological outcome of laparoscopic radical hysterectomy: JGOG1081s-A1, an ancillary analysis of the Japanese Gynecologic Oncology Group Study JGOG1081

ObjectivesWe investigated whether surgical skill and procedure were related to oncological outcomes in cervical cancer patients who underwent Laparoscopic Radical Hysterectomy (LRH). MethodsWe previously assessed data of LRH from 251 patients with FIGO stage (2009) IA2, IB1and IIA1 cervical cancer collected for JGOG 1081s study. 1) The JGOG 1081s cohort study was re-examined to refine the surgical details and extend the follow-up period as chart review. 2) Unedited videos for recurrent cases and matched non-recurrent control cases were newly compared by experts for various surgical skills and surgical procedures using the modified Objective Structured Assessment of Technical Skills (OSATS) tool, without awareness of the recurrence status as video review. ResultsAfter a median follow-up of 46 months, tumors had recurred in 31 of the 251 patients. The five-year Recurrence-Free Survival rate was 86.9% (81.8–90.6) and five-year Overall Survival rate was 93.7% (87.5–96.8). Multivariate analysis from chart reviews found that an experience with LRH of less than 20 cases per institution was an independent prognostic factor for recurrence (Hazard Ratio (HR) 2.49, 95%CI 1.12–5.53, p = 0.025). For the surgical video review, we compared 23 videos of recurrent cases with 23 background-matched non-recurrent controls. Lower modified OSATS scores from the video review were consistently trended to have a higher risk of recurrence. ConclusionsOur new study has found that LRH surgical experience and skill trended to have better oncological outcomes.

Application of a novel surgical difficulty grading system during laparoscopic cholecystectomy.

Prevention of bile duct injury and vasculo-biliary injury while performing laparoscopic cholecystectomy (LC) is an unsolved problem. Clarifying the surgical difficulty using intraoperative findings can greatly contribute to the pursuit of best practices for acute cholecystitis. In this study, multiple evaluators assessed surgical difficulty items in unedited videos and then constructed a proposed surgical difficulty grading. We previously assembled a library of typical video clips of the intraoperative findings for all LC surgical difficulty items in acute cholecystitis. Fifty-one experts on LC assessed unedited surgical videos. Inter-rater agreement was assessed by Fleiss's κ and Gwet's agreement coefficient (AC). Except for one item ("edematous change"), κ or AC exceeded 0.5, so the typical videos were judged to be applicable. The conceivable surgical difficulty gradings were analyzed. According to the assessment of difficulty factors, we created a surgical difficulty grading system (agreement probability=0.923, κ=0.712, 90% CI: 0.587-0.837; AC2 =0.870, 90% CI: 0.768-0.972). The previously published video clip library and our novel surgical difficulty grading system should serve as a universal objective tool to assess surgical difficulty in LC.

The Editing Density of Moving Images Influences Viewers' Time Perception: The Mediating Role of Eye Movements.

The present study examined whether cinematographic editing density affects viewers' perception of time. As a second aim, based on embodied models that conceive time perception as strictly connected to the movement, we tested the hypothesis that the editing density of moving images also affects viewers' eye movements and that these later mediate the effect of editing density on viewers' temporal judgments. Seventy participants watched nine video clips edited by manipulating the number of cuts (slow- and fast-paced editing against a master shot, unedited condition). For each editing density, multiple video clips were created, representing three different kinds of routine actions. The participants' eye movements were recorded while watching the video, and the participants were asked to report duration judgments and subjective passage of time judgments after watching each clip. The results showed that participants subjectively perceived that time flew more while watching fast-paced edited videos than slow-paced or unedited videos; by contrast, concerning duration judgments, participants overestimated the duration of fast-paced videos compared to the master-shot videos. Both the slow- and the fast-paced editing generated shorter fixations than the master shot, and the fast-paced editing led to shorter fixations than the slow-paced editing. Finally, compared to the unedited condition, editing led to an overestimation of durations through increased eye mobility. These findings suggest that the editing density of moving images by increasing the number of cuts effectively altered viewers' experience of time and add further evidence to prior research showing that performed eye movement is associated with temporal judgments.

Performance Assessment for Total Laparoscopic Hysterectomy in the Operating Room: Validity Evidence of a Procedure-specific Rating Scale

The technical conduct of total laparoscopic hysterectomy (LH) is critical to surgical outcomes. This study explored the validity evidence of an objective scale specific to the assessment of technical skills (H-OSATS) for 7 tasks of an LH with salpingo-oophorectomy procedure performed in the operating room. Observational cohort study. Two academic hospitals in Marseille and Montpellier, France. Three groups of operators (novice, intermediate, and experienced surgeons) were video recorded during their live performances of LH on a simple case. For each group, a dozen unedited videos were obtained for the following tasks: division of the round ligament, division of the infundibulopelvic ligament, creation of the bladder flap, opening of the posterior peritoneum, division of the uterine vessels, colpotomy, and closure of the vault. Two qualified raters blindly assessed each video using the H-OSATS rating scale. Inter-rater reliability and test-retest reliability were calculated as measures of internal structure. In a separate round of evaluations, the raters provided a global competent/noncompetent decision for each performance. As a measure of consequential validity, a pass/fail score was set for each task using the contrasting group method. Three tasks (creation of the bladder flap, colpotomy, and closure of the vault) displayed sound validity evidence: a meaningful total score difference among the 3 groups of experience as well as between the intermediate and experienced surgeons, reliability outcomes of >0.7, and a pass/fail score with a theoretical false-positive rate of <10%. The validity evidence of the H-OSATS rating scale differed for separate evaluations of the 7 tasks. Three tasks (i.e., creation of the bladder flap, colpotomy, and closure of the vault) revealed sound validity evidence, including at the level of the attending surgeon, whereas other tasks were more consistent with low-stakes formative evaluation standards.

Live Case Demonstrations are Essential for the Success of Pediatric Urology Meetings in India.

Purpose:The purpose of the study was to survey the opinion of delegates attending national pediatric urology meeting regarding live case demonstrations (LCDs) to see whether these can be replaced with taped case demonstrations (TCDs) in future.Methods:A questionnaire-based survey was conducted at the end of the 3-day annual conference and live operative workshop in pediatric urology. Apart from general data such as age of the respondent and type of practice setting, four key questions on LCDs with yes/no responses included: would you have attended this meeting if there were no LCDs? Are unedited videos (TCDs) as effective as LCDs for teaching? Do you think LCDs should be continued? Would you allow your child to be operated in LCD by an expert? For question 3, the outcomes were compared between junior surgeons (<45-year-old) and senior surgeons (>45-year-old).Results:On analysis, 88/140 (62%) respondents (95% confidence interval [CI]: 54%–70%) said that they would not have attended the meeting if there were no LCDs; 70/139 (50%) respondents (95% CI: 42%–58%) felt that TCDs may be an effective alternative to LCDs; 129/144 (90%) respondents (95% CI: 83%–94%) felt that LCDs should be continued. For question 4, 101/129 (79%) said they would allow their child to be operated in LCD by an expert, while 28 (21%) did not agree for their child to be operated in LCD. There was no significant difference between junior and senior surgeons regarding support for LCDs (P = 0.15).Conclusions:In the Indian scenario, LCDs were favored as an essential part of pediatric urology meeting, attract more participation, and are likely to be important for the success of the meeting.

European expert consensus on a structured approach to training robotic-assisted low anterior resection using performance metrics.

AimThe aim was to develop and operationally define ‘performance metrics’ that characterize a reference approach to robotic‐assisted low anterior resection (RA‐LAR) and to obtain face and content validity through a consensus meeting.MethodThree senior colorectal surgeons with robotic experience and a senior behavioural scientist formed the Metrics Group. We used published guidelines, training materials, manufacturers’ instructions and unedited videos of RA‐LAR to deconstruct the operation into defined, measurable components – performance metrics (i.e. procedure phases, steps, errors and critical errors). The performance metrics were then subjected to detailed critique by 18 expert colorectal surgeons in a modified Delphi process.ResultsPerformance metrics for RA‐LAR had 15 procedure phases, 128 steps, 89 errors and 117 critical errors in women, 88 errors and 118 critical errors in men. After the modified Delphi process the final performance metrics consisted of 14 procedure phases, 129 steps, 88 errors and 115 critical errors in women, 87 errors and 116 critical errors in men. After discussion by the Delphi panel, all procedure phases received unanimous consensus apart from phase I (patient positioning and preparation, 83%) and phase IV (docking, 94%).ConclusionA robotic rectal operation can be broken down into procedure phases, steps, with errors and critical errors, known as performance metrics. The face and content of these metrics have been validated by a large group of expert robotic colorectal surgeons from Europe. We consider the metrics essential for the development of a structured training curriculum and standardized procedural assessment for RA‐LAR.

Development and validation of an endoscopic submucosal dissection video assessment tool.

Despite a need for assessment of endoscopic submucosal dissection (ESD) skills in order to track progress and determine competence, there is no structured measure of assessing competency in ESD performance. The present study aims to develop and examine validity evidence for an assessment tool to evaluate the recorded performance of ESD for gastric neoplasms. The ESD video assessment tool (EVAT) was systematically developed by ESD experienced endoscopists. The EVAT consists of a 25-item global rating scale and 3-item checklist to assess competencies required to perform ESD. Five unedited videos were each evaluated by 2-blinded experienced ESD endoscopists to assess inter-rater reliability using intraclass correlation coefficients (ICC). Seventeen unedited videos in total were rated by 3 blinded experienced ESD endoscopists. Validity evidence for relationship to other variables was examined by comparing scores of inexperienced (fellows) and experienced endoscopists (attending staff), and by evaluating the relationship between the EVAT scores and ESD case experience. Internal consistency was evaluated using Cronbach's alpha. The inter-rater reliability for the total score was high at 0.87 (95% confidence interval 0.11 to 0.99). The total score [median, interquartile range (IQR)] was significantly different between the inexperienced (71, 63-77) and experienced group (95, 91-97) (P = 0.005). The total scores demonstrated high correlation with the number of ESD cases (Spearman's ρ = 0.79, P < 0.01). The internal consistency was 0.97. This study provides preliminary validity evidence for the assessment of video-recorded ESD performances for gastric neoplasms using EVAT.

Surgeon Quality Control and Standardization of D2 Lymphadenectomy for Gastric Cancer: A Prospective Multicenter Observational Study (KLASS-02-QC).

To qualify surgeons to participate in a randomized trial comparing laparoscopic and open distal D2 gastrectomy for advanced gastric cancer. No studies have sought to qualify surgeons for a randomized trial comparing laparoscopic and open D2 gastrectomy for advanced gastric cancer. We conducted a multicenter prospective observational study evaluating unedited videos of laparoscopic and open D2 gastrectomy performed by 27 surgeons. Surgeons performed 3 of each laparoscopic and open distal gastrectomies with D2 lymphadenectomy for gastric cancer. Five peers reviewed each unedited video using a video assessment form. Based on experts' review of videos, a separate review committee decided surgeons as "Qualified" or "Not-qualified." Twelve surgeons (44.4%) were qualified on initial evaluation whereas the other 15 surgeons were not. Another 9 surgeons were finally qualified after re-evaluation. The median score for Qualified was significantly higher than Not-qualified (P < 0.001).Significant differences between Qualified and Not-qualified were noted both in operation type and in all evaluation area of surgical skill, perigastric, and extra-perigastric lymphadenectomy, although the inter-rater variability of the assessment score was low (kappa = 0.285). However, Not-qualified surgeons' scores improved upon re-evaluation of resubmitted videos.When compared laparoscopy with open surgery, median scores were similar between the 2 groups (P = 0.680). However, open gastrectomy scores for surgical skills were significantly higher than for laparoscopic surgery (P = 0.016). Our surgeon quality control study for gastrectomy represents a milestone in surgical standardization for surgical clinical trials. Our methods could also serve as a system for educating surgeons and assessing surgical proficiency.

Find Objects and Focus on Highlights: Mining Object Semantics for Video Highlight Detection via Graph Neural Networks

With the increasing prevalence of portable computing devices, browsing unedited videos is time-consuming and tedious. Video highlight detection has the potential to significantly ease this situation, which discoveries moments of user's major or special interest in a video. Existing methods suffer from two problems. Firstly, most existing approaches only focus on learning holistic visual representations of videos but ignore object semantics for inferring video highlights. Secondly, current state-of-the-art approaches often adopt the pairwise ranking-based strategy, which cannot enjoy the global information to infer highlights. Therefore, we propose a novel video highlight framework, named VH-GNN, to construct an object-aware graph and model the relationships between objects from a global view. To reduce computational cost, we decompose the whole graph into two types of graphs: a spatial graph to capture the complex interactions of object within each frame, and a temporal graph to obtain object-aware representation of each frame and capture the global information. In addition, we optimize the framework via a proposed multi-stage loss, where the first stage aims to determine the highlight-probability and the second stage leverage the relationships between frames and focus on hard examples from the former stage. Extensive experiments on two standard datasets strongly evidence that VH-GNN obtains significant performance compared with state-of-the-arts.

Efficient Bronchoscopic Video Summarization.

Bronchoscopy enables many minimally invasive chest procedures for diseases such as lung cancer and asthma. Guided by the bronchoscope's video stream, a physician can navigate the complex three-dimensional (3-D) airway tree to collect tissue samples or administer a disease treatment. Unfortunately, physicians currently discard procedural video because of the overwhelming amount of data generated. Hence, they must rely on memory and anecdotal snapshots to document a procedure. We propose a robust automatic method for summarizing an endobronchial video stream. Inspired by the multimedia concept of the video summary and by research in other endoscopy domains, our method consists of three main steps: 1) shot segmentation, 2) motion analysis, and 3) keyframe selection. Overall, the method derives a true hierarchical decomposition, consisting of a shot set and constituent keyframe set, for a given procedural video. No other method to our knowledge gives such a structured summary for the raw, unscripted, unedited videos arising in endoscopy. Results show that our method more efficiently covers the observed endobronchial regions than other keyframe-selection approaches and is robust to parameter variations. Over a wide range of video sequences, our method required on average only 6.5% of available video frames to achieve a video coverage = 92.7%. We also demonstrate how the derived video summary facilitates direct fusion with a patient's 3-D chest computed-tomography scan in a system under development, thereby enabling efficient video browsing and retrieval through the complex airway tree.