Glutamatergic synaptic input to glioma cells drives brain tumour progression.

A network of communicating tumour cells that is connected by tumour microtubes mediates the progression of incurable gliomas. Moreover, neuronal activity can foster malignant behaviour of glioma cells by non-synaptic paracrine and autocrine mechanisms. Here we report a direct communication channel between neurons and glioma cells in different disease models and human tumours: functional bona fide chemical synapses between presynaptic neurons and postsynaptic glioma cells. These neurogliomal synapses show a typical synaptic ultrastructure, are located on tumour microtubes, and produce postsynaptic currents that are mediated by glutamate receptors of the AMPA subtype. Neuronal activity including epileptic conditions generates synchronised calcium transients in tumour-microtube-connected glioma networks. Glioma-cell-specific genetic perturbation of AMPA receptors reduces calcium-related invasiveness of tumour-microtube-positive tumour cells and glioma growth. Invasion and growth are also reduced by anaesthesia and the AMPA receptor antagonist perampanel, respectively. These findings reveal a biologically relevant direct synaptic communication between neurons and glioma cells with potential clinical implications.

Effects of endoluminal vacuum sponge therapy on the perfusion of gastric conduit in a porcine model for esophagectomy.

BACKGROUND: After esophagectomy, the postoperative rate of anastomotic leakage is up to 30% and is the main driver of postoperative morbidity. Contemporary management includes endoluminal vacuum sponge therapy (EndoVAC) with good success rates. Vacuum therapy improves tissue perfusion in superficial wounds, but this has not been shown for gastric conduits. This study aimed to assess gastric conduit perfusion with EndoVAC in a porcine model for esophagectomy. MATERIAL AND METHODS: A porcine model (n = 18) was used with gastric conduit formation and induction of ischemia at the cranial end of the gastric conduit with measurement of tissue perfusion over time. In three experimental groups EndoVAC therapy was then used in the gastric conduit (- 40, - 125, and - 200 mmHg). Changes in tissue perfusion and tissue edema were assessed using hyperspectral imaging. The study was approved by local authorities (Project License G-333/19, G-67/22). RESULTS: Induction of ischemia led to significant reduction of tissue oxygenation from 65.1 ± 2.5% to 44.7 ± 5.5% (p < 0.01). After EndoVAC therapy with - 125 mmHg a significant increase in tissue oxygenation to 61.9 ± 5.5% was seen after 60 min and stayed stable after 120 min (62.9 ± 9.4%, p < 0.01 vs tissue ischemia). A similar improvement was seen with EndoVAC therapy at - 200 mmHg. A nonsignificant increase in oxygenation levels was also seen after therapy with - 40 mmHg, from 46.3 ± 3.4% to 52.5 ± 4.3% and 53.9 ± 8.1% after 60 and 120 min respectively (p > 0.05). An increase in tissue edema was observed after 60 and 120 min of EndoVAC therapy with - 200 mmHg but not with - 40 and - 125 mmHg. CONCLUSIONS: EndoVAC therapy with a pressure of - 125 mmHg significantly increased tissue perfusion of ischemic gastric conduit. With better understanding of underlying physiology the optimal use of EndoVAC therapy can be determined including a possible preemptive use for gastric conduits with impaired arterial perfusion or venous congestion.

Robotic-assisted versus laparoscopic bowel anastomoses: randomized crossover in vivo experimental study.

BACKGROUND: Initial learning curves are potentially shorter in robotic-assisted surgery (RAS) than in conventional laparoscopic surgery (LS). There is little evidence to support this claim. Furthermore, there is limited evidence how skills from LS transfer to RAS. METHODS: A randomized controlled, assessor blinded crossover study to compare how RAS naïve surgeons (n = 40) performed linear-stapled side-to-side bowel anastomoses in an in vivo porcine model with LS and RAS. Technique was rated using the validated anastomosis objective structured assessment of skills (A-OSATS) score and the conventional OSATS score. Skill transfer from LS to RAS was measured by comparing the RAS performance of LS novices and LS experienced surgeons. Mental and physical workload was measured with the NASA-task load index (NASA-Tlx) and the Borg-scale. OUTCOMES: In the overall cohort, there were no differences between RAS and LS for surgical performance (A-OSATS, time, OSATS). Surgeons that were naïve in both LS and RAS had significantly higher A-OSATS scores in RAS (Mean (Standard deviation (SD)): LS: 48.0 ± 12.1; RAS: 52.0 ± 7.5); p = 0.044) mainly deriving from better bowel positioning (LS: 8.7 ± 1.4; RAS: 9.3 ± 1.0; p = 0.045) and closure of enterotomy (LS: 12.8 ± 5.5; RAS: 15.6 ± 4.7; p = 0.010). There was no statistically significant difference in how LS novices and LS experienced surgeons performed in RAS [Mean (SD): novices: 48.9 ± 9.0; experienced surgeons: 55.9 ± 11.0; p = 0.540]. Mental and physical demand was significantly higher after LS. CONCLUSION: The initial performance was improved for RAS versus LS for linear stapled bowel anastomosis, whereas workload was higher for LS. There was limited transfer of skills from LS to RAS.

The Rise of Facial Palsy on Social Media Over the Last 5 Years.

BACKGROUND: Social media (SoMe) has become a powerful platform for distributing health information. Facial palsy (FP) results in functional and social impairment and lowers quality of life. Social media may help to raise awareness of FP sequalae. This study aims to determine the FP information growth on SoMe platforms and parameters that influence user engagement on FP content. METHODS: Five commonly used SoMe platforms (Facebook, Instagram, TikTok, Twitter, and Reddit) were analyzed. Data on 18 FP hashtags and their social interaction parameters (posts, likes, reaches, comments, shares, language, and country of origin) over the past 5 years (July 31, 2016, to July 31, 2021) were collected. In-depth account analysis was performed on the 5 most popular Instagram profiles associated with FP. RESULTS: The annual growth curve was positive on each platform. Facial Palsy Awareness Week 2021 trended best on TikTok. Facebook accumulated 315,411 likes and 1,922,678 reaches on 8356 posts. On Instagram, 24,968 posts gathered 4,904,124 likes and 9,215,852 reaches. TikTok users interacted on 3565 posts, accumulating 4,304,155 likes and 4,200,368 reaches. The implementation of reels ( P <0.001) and the profile host interacting with their followers by liking ( P <0.001) and replying ( P <0.001) to users' comments significantly increased the engagement rate. CONCLUSIONS: Facial palsy is of increasing interest on SoMe. Facial palsy surgeons may post reels, interact with their community, and engage into FPAW to promote user engagement.

[Status Quo of Surgical Navigation]. / Update zur Navigation im OP-Saal.

Surgical navigation, also referred to as computer-assisted or image-guided surgery, is a technique that employs a variety of methods - such as 3D imaging, tracking systems, specialised software, and robotics to support surgeons during surgical interventions. These emerging technologies aim not only to enhance the accuracy and precision of surgical procedures, but also to enable less invasive approaches, with the objective of reducing complications and improving operative outcomes for patients. By harnessing the integration of emerging digital technologies, surgical navigation holds the promise of assisting complex procedures across various medical disciplines. In recent years, the field of surgical navigation has witnessed significant advances. Abdominal surgical navigation, particularly endoscopy, laparoscopic, and robot-assisted surgery, is currently undergoing a phase of rapid evolution. Emphases include image-guided navigation, instrument tracking, and the potential integration of augmented and mixed reality (AR, MR). This article will comprehensively delve into the latest developments in surgical navigation, spanning state-of-the-art intraoperative technologies like hyperspectral and fluorescent imaging, to the integration of preoperative radiological imaging within the intraoperative setting.

A systematic review and meta-analysis of randomized controlled trials comparing laparoscopic and open liver resection.

BACKGROUND/PURPOSE: The dissemination of laparoscopic liver resection (LLR) has been based on non-randomized studies and reviews of these. Aim of this study was to evaluate if the randomized evidence comparing LLR to open liver resection (OLR) supports these findings. METHODS: A prospectively registered (reviewregistry866) systematic review and meta-analysis following Cochrane and PRISMA guidelines comparing LLR to OLR for benign and malignant diseases was performed via Medline, Web of Science, CENTRAL up to 31.12.2020. The main outcome was postoperative complications. Risk of bias was assessed with the Cochrane Risk of Bias tool 2.0, certainty of evidence was assessed using the GRADE approach. RESULTS: The search yielded 2080 results. 13 RCTs assessing mostly minor liver resections with 1457 patients were included. There were reduced odds of experiencing any complication (Odds ratio (OR) [95% confidence interval (CI)]: 0·42 [0·30, 0·58]) and severe complications (OR[CI]: 0·51 [0·31, 0·84]) for patients undergoing LLR. LOS was shorter (Mean difference (MD) [CI]: -2·90 [-3·88, -1·92] days), blood loss was lower (MD: [CI]: -115·41 [-146·08, -84·75] ml), and functional recovery was better for LLR. All other outcomes showed no significant differences. CONCLUSIONS: LLR shows significant postoperative benefits. RCTs assessing long-term outcomes and major resections are needed.

[Robotic pancreatic surgery]. / Robotische Pankreaschirurgie.

Robotic operations as a further development of conventional laparoscopic surgery have been introduced for nearly all interventions in visceral surgery during the last decade. They also currently have a high importance and acceptance in pancreatic surgery despite a relevant learning curve and high associated costs. Standard procedures, such as robotic distal pancreatectomy (RDP) and partial pancreatoduodenectomy (RPD) are most frequently performed, whereas extended resections, e.g., vascular reconstructions of the portal vein, are still limited to a small number of centers worldwide. Potential advantages of robotic pancreatic surgery compared to open surgery include, in particular, less blood loss and a faster postoperative recovery of the patients leading to a shorter hospital stay. Compared to conventional laparoscopic surgery, robotic approaches offer advantages with respect to better visualization and three-dimensional dexterity of the instruments; however, the currently published literature comprises only retrospective or prospective observational studies and randomized controlled results are not yet available but first study results in this respect are expected within the next 2-3 years.

Endotracheal Intubation via Tracheotomy and Subsequent Thoracotomy in Rats for Non-Survival Applications.

Endotracheal intubation and subsequent ventilation are often basic requirements for translational research in rat models for various interventions that require controlled or high ventilation pressures or access to the thoracic cavity and organs. Conventional endoorotracheal intubation using the anatomically existing route through the mouth is well suited for survival experiments. However, this procedure poses some challenges, including generally higher levels of the required experience and technical skill, more advanced equipment, and greater time effort with relevant intubation failure rates and complications such as tracheal perforation, temporary systemic hypooxygenation, and relevant aerial leakage. This manuscript, therefore, presents a detailed step-by-step protocol for endotracheal intubation through tracheotomy in non-survival rat models when guaranteed intubation success, constant oxygenation levels, high ventilation pressures, or open thoracotomy are required. The protocol emphasizes the importance of meticulous surgical technique to ensure consistent and reliable outcomes, especially for researchers who are inexperienced or lack routine in the technique of endoorotracheal intubation via direct laryngoscopy. This procedure is, therefore, expected to minimize animal suffering and unnecessary animal losses.

Clinical Research in Renal Transplantation: A Bibliometric Perspective on a Half-century of Innovation and Progress.

BACKGROUND: Groundbreaking biomedical research has transformed renal transplantation (RT) into a widespread clinical procedure that represents the mainstay of treatment for end-stage kidney failure today. Here, we aimed to provide a comprehensive bibliometric perspective on the last half-century of innovation in clinical RT. METHODS: The Web of Science Core Collection was used for a comprehensive screening yielding 123 303 research items during a 50-y period (January 1973-October 2022). The final data set of the 200 most-cited articles was selected on the basis of a citation-based strategy aiming to minimize bias. RESULTS: Studies on clinical and immunological outcomes (n = 63 and 48), registry-based epi research (n = 38), and randomized controlled trials (n = 35) dominated the data set. Lead US authors have signed 110 of 200 articles. The overall level of evidence was high, with 84% of level1 and -2 reports. Highest numbers of these articles were published in New England Journal of Medicine , Transplantation , and American Journal of Transplantation. Increasing trend was observed in the number of female authors in the postmillennial era (26% versus 7%). CONCLUSIONS: This study highlights important trends in RT research of the past half-century. This bibliometric perspective identifies the most intensively researched areas and shift of research interests over time; however, it also describes important imbalances in distribution of academic prolificacy based on topic, geographical aspects, and gender.

Test-time augmentation with synthetic data addresses distribution shifts in spectral imaging.

PURPOSE: Surgical scene segmentation is crucial for providing context-aware surgical assistance. Recent studies highlight the significant advantages of hyperspectral imaging (HSI) over traditional RGB data in enhancing segmentation performance. Nevertheless, the current hyperspectral imaging (HSI) datasets remain limited and do not capture the full range of tissue variations encountered clinically. METHODS: Based on a total of 615 hyperspectral images from a total of 16 pigs, featuring porcine organs in different perfusion states, we carry out an exploration of distribution shifts in spectral imaging caused by perfusion alterations. We further introduce a novel strategy to mitigate such distribution shifts, utilizing synthetic data for test-time augmentation. RESULTS: The effect of perfusion changes on state-of-the-art (SOA) segmentation networks depended on the organ and the specific perfusion alteration induced. In the case of the kidney, we observed a performance decline of up to 93% when applying a state-of-the-art (SOA) network under ischemic conditions. Our method improved on the state-of-the-art (SOA) by up to 4.6 times. CONCLUSION: Given its potential wide-ranging relevance to diverse pathologies, our approach may serve as a pivotal tool to enhance neural network generalization within the realm of spectral imaging.

Multiscale and multimodal imaging for three-dimensional vascular and histomorphological organ structure analysis of the pancreas.

Exocrine and endocrine pancreas are interconnected anatomically and functionally, with vasculature facilitating bidirectional communication. Our understanding of this network remains limited, largely due to two-dimensional histology and missing combination with three-dimensional imaging. In this study, a multiscale 3D-imaging process was used to analyze a porcine pancreas. Clinical computed tomography, digital volume tomography, micro-computed tomography and Synchrotron-based propagation-based imaging were applied consecutively. Fields of view correlated inversely with attainable resolution from a whole organism level down to capillary structures with a voxel edge length of 2.0 µm. Segmented vascular networks from 3D-imaging data were correlated with tissue sections stained by immunohistochemistry and revealed highly vascularized regions to be intra-islet capillaries of islets of Langerhans. Generated 3D-datasets allowed for three-dimensional qualitative and quantitative organ and vessel structure analysis. Beyond this study, the method shows potential for application across a wide range of patho-morphology analyses and might possibly provide microstructural blueprints for biotissue engineering.

Biophotonics-Intraoperative Guidance During Partial Nephrectomy: A Systematic Review and Meta-analysis.

CONTEXT: Partial nephrectomy (PN) with intraoperative guidance by biophotonics has the potential to improve surgical outcomes due to higher precision. However, its value remains unclear since high-level evidence is lacking. OBJECTIVE: To provide a comprehensive analysis of biophotonic techniques used for intraoperative real-time assistance during PN. EVIDENCE ACQUISITION: We performed a comprehensive database search based on the PICO criteria, including studies published before October 2022. Two independent reviewers screened the titles and abstracts followed by full-text screening of eligible studies. For a quantitative analysis, a meta-analysis was conducted. EVIDENCE SYNTHESIS: In total, 35 studies were identified for the qualitative analysis, including 27 studies on near-infrared fluorescence (NIRF) imaging using indocyanine green, four studies on hyperspectral imaging, two studies on folate-targeted molecular imaging, and one study each on optical coherence tomography and 5-aminolevulinic acid. The meta-analysis investigated seven studies on selective arterial clamping using NIRF. There was a significantly shorter warm ischemia time in the NIRF-PN group (mean difference [MD]: -2.9; 95% confidence interval [CI]: -5.6, -0.1; p = 0.04). No differences were noted regarding transfusions (odds ratio [OR]: 0.5; 95% CI: 0.2, 1.7; p = 0.27), positive surgical margins (OR: 0.7; 95% CI: 0.2, 2.0; p = 0.46), or major complications (OR: 0.4; 95% CI: 0.1, 1.2; p = 0.08). In the NIRF-PN group, functional results were favorable at short-term follow-up (MD of glomerular filtration rate decline: 7.6; 95% CI: 4.6, 10.5; p < 0.01), but leveled off at long-term follow-up (MD: 7.0; 95% CI: -2.8, 16.9; p = 0.16). Remarkably, these findings were not confirmed by the included randomized controlled trial. CONCLUSIONS: Biophotonics comprises a heterogeneous group of imaging modalities that serve intraoperative decision-making and guidance. Implementation into clinical practice and cost effectiveness are the limitations that should be addressed by future research. PATIENT SUMMARY: We reviewed the application of biophotonics during partial removal of the kidney in patients with kidney cancer. Our results suggest that these techniques support the surgeon in successfully performing the challenging steps of the procedure.

Microcirculatory tissue oxygenation correlates with kidney function after transcatheter aortic valve implantation-Results from a prospective observational study.

Introduction: Kidney dysfunction is common in patients with aortic stenosis (AS) and correction of the aortic valve by transcatheter aortic valve implantation (TAVI) often affects kidney function. This may be due to microcirculatory changes. Methods: We evaluated skin microcirculation with a hyperspectral imaging (HSI) system, and compared tissue oxygenation (StO2), near-infrared perfusion index (NIR), tissue hemoglobin index (THI) and tissue water index (TWI) in 40 patients undergoing TAVI versus 20 control patients. HSI parameters were measured before TAVI (t1), directly after TAVI (t2), and on postinterventional day 3 (t3). The primary outcome was the correlation of tissue oxygenation (StO2) to the creatinine level after TAVI. Results: We performed 116 HSI image recordings in patients undergoing TAVI for the treatment of severe aortic stenosis and 20 HSI image recordings in control patients. Patients with AS had a lower THI at the palm (p = 0.034) and a higher TWI at the fingertips (p = 0.003) in comparison to control patients. TAVI led to an increase of TWI, but had no uniform enduring effect on StO2 and THI. Tissue oxygenation StO2 at both measurement sites correlated negatively with creatinine levels after TAVI at t2 (palm: ρ = -0.415; p = 0.009; fingertip: ρ = -0.519; p < 0.001) and t3 (palm: ρ = -0.427; p = 0.008; fingertip: ρ = -0.398; p = 0.013). Patients with higher THI at t3 reported higher physical capacity and general health scores 120 days after TAVI. Conclusion: HSI is a promising technique for periinterventional monitoring of tissue oxygenation and microcirculatory perfusion quality, which are related to kidney function, physical capacity, and clinical outcomes after TAVI. Clinical trial registration: https://drks.de/search/de/trial, identifier DRKS00024765.

HeiPorSPECTRAL - the Heidelberg Porcine HyperSPECTRAL Imaging Dataset of 20 Physiological Organs.

Hyperspectral Imaging (HSI) is a relatively new medical imaging modality that exploits an area of diagnostic potential formerly untouched. Although exploratory translational and clinical studies exist, no surgical HSI datasets are openly accessible to the general scientific community. To address this bottleneck, this publication releases HeiPorSPECTRAL ( https://www.heiporspectral.org ; https://doi.org/10.5281/zenodo.7737674 ), the first annotated high-quality standardized surgical HSI dataset. It comprises 5,758 spectral images acquired with the TIVITA® Tissue and annotated with 20 physiological porcine organs from 8 pigs per organ distributed over a total number of 11 pigs. Each HSI image features a resolution of 480 × 640 pixels acquired over the 500-1000 nm wavelength range. The acquisition protocol has been designed such that the variability of organ spectra as a function of several parameters including the camera angle and the individual can be assessed. A comprehensive technical validation confirmed both the quality of the raw data and the annotations. We envision potential reuse within this dataset, but also its reuse as baseline data for future research questions outside this dataset. Measurement(s) Spectral Reflectance Technology Type(s) Hyperspectral Imaging Sample Characteristic - Organism Sus scrofa.

ICG-augmented hyperspectral imaging for visualization of intestinal perfusion compared to conventional ICG fluorescence imaging: an experimental study.

BACKGROUND: Small bowel malperfusion (SBM) can cause high morbidity and severe surgical consequences. However, there is no standardized objective measuring tool for the quantification of SBM. Indocyanine green (ICG) imaging can be used for visualization, but lacks standardization and objectivity. Hyperspectral imaging (HSI) as a newly emerging technology in medicine might present advantages over conventional ICG fluorescence or in combination with it. METHODS: HSI baseline data from physiological small bowel, avascular small bowel and small bowel after intravenous application of ICG was recorded in a total number of 54 in-vivo pig models. Visualizations of avascular small bowel after mesotomy were compared between HSI only (1), ICG-augmented HSI (IA-HSI) (2), clinical evaluation through the eyes of the surgeon (3) and conventional ICG imaging (4). The primary research focus was the localization of resection borders as suggested by each of the four methods. Distances between these borders were measured and histological samples were obtained from the regions in between in order to quantify necrotic changes 6 h after mesotomy for every region. RESULTS: StO2 images (1) were capable of visualizing areas of physiological perfusion and areas of clearly impaired perfusion. However, exact borders where physiological perfusion started to decrease could not be clearly identified. Instead, IA-HSI (2) suggested a sharp-resection line where StO2 values started to decrease. Clinical evaluation (3) suggested a resection line 23 mm (±7 mm) and conventional ICG imaging (4) even suggested a resection line 53 mm (±13 mm) closer towards the malperfused region. Histopathological evaluation of the region that was sufficiently perfused only according to conventional ICG (R3) already revealed a significant increase in pre-necrotic changes in 27% (±9%) of surface area. Therefore, conventional ICG seems less sensitive than IA-HSI with regards to detection of insufficient tissue perfusion. CONCLUSIONS: In this experimental animal study, IA-HSI (2) was superior for the visualization of segmental SBM compared to conventional HSI imaging (1), clinical evaluation (3) or conventional ICG imaging (4) regarding histopathological safety. ICG application caused visual artifacts in the StO2 values of the HSI camera as values significantly increase. This is caused by optical properties of systemic ICG and does not resemble a true increase in oxygenation levels. However, this empirical finding can be used to visualize segmental SBM utilizing ICG as contrast agent in an approach for IA-HSI. Clinical applicability and relevance will have to be explored in clinical trials. LEVEL OF EVIDENCE: Not applicable. Translational animal science. Original article.

Robust deep learning-based semantic organ segmentation in hyperspectral images.

Semantic image segmentation is an important prerequisite for context-awareness and autonomous robotics in surgery. The state of the art has focused on conventional RGB video data acquired during minimally invasive surgery, but full-scene semantic segmentation based on spectral imaging data and obtained during open surgery has received almost no attention to date. To address this gap in the literature, we are investigating the following research questions based on hyperspectral imaging (HSI) data of pigs acquired in an open surgery setting: (1) What is an adequate representation of HSI data for neural network-based fully automated organ segmentation, especially with respect to the spatial granularity of the data (pixels vs. superpixels vs. patches vs. full images)? (2) Is there a benefit of using HSI data compared to other modalities, namely RGB data and processed HSI data (e.g. tissue parameters like oxygenation), when performing semantic organ segmentation? According to a comprehensive validation study based on 506 HSI images from 20 pigs, annotated with a total of 19 classes, deep learning-based segmentation performance increases - consistently across modalities - with the spatial context of the input data. Unprocessed HSI data offers an advantage over RGB data or processed data from the camera provider, with the advantage increasing with decreasing size of the input to the neural network. Maximum performance (HSI applied to whole images) yielded a mean DSC of 0.90 ((standard deviation (SD)) 0.04), which is in the range of the inter-rater variability (DSC of 0.89 ((standard deviation (SD)) 0.07)). We conclude that HSI could become a powerful image modality for fully-automatic surgical scene understanding with many advantages over traditional imaging, including the ability to recover additional functional tissue information. Our code and pre-trained models are available at https://github.com/IMSY-DKFZ/htc.

[Artificial intelligence and hyperspectral imaging for image-guided assistance in minimally invasive surgery]. / Künstliche Intelligenz und hyperspektrale Bildgebung zur bildgestützten Assistenz in der minimal-invasiven Chirurgie.

BACKGROUND: Intraoperative imaging assists surgeons during minimally invasive procedures. Hyperspectral imaging (HSI) is a noninvasive and noncontact optical technique with great diagnostic potential in medicine. The combination with artificial intelligence (AI) approaches to analyze HSI data is called intelligent HSI in this article. OBJECTIVE: What are the medical applications and advantages of intelligent HSI for minimally invasive visceral surgery? MATERIAL AND METHODS: Within various clinical studies HSI data from multiple in vivo tissue types and oncological resections were acquired using an HSI camera system. Different AI algorithms were evaluated for detection and discrimination of organs, risk structures and tumors. RESULTS: In an experimental animal study 20 different organs could be differentiated with high precision (> 95%) using AI. In vivo, the parathyroid glands could be discriminated from surrounding tissue with an F1 score of 47% and sensitivity of 75%, and the bile duct with an F1 score of 79% and sensitivity of 90%. Furthermore, ex vivo tumor tissue could be successfully detected with an area under the receiver operating characteristic (ROC) curve (AUC) larger than 0.91. DISCUSSION: This study demonstrates that intelligent HSI can automatically and accurately detect different tissue types. Despite great progress in the last decade intelligent HSI still has limitations. Thus, accurate AI algorithms that are easier to understand for the user and an extensive standardized and continuously growing database are needed. Further clinical studies should support the various medical applications and lead to the adoption of intelligent HSI in the clinical routine practice.

Implementation of hyperspectral imaging in a trauma resuscitation room: a randomized controlled trial.

BACKGROUND: Hyperspectral imaging (HSI) is a novel imaging technology with the ability to assess microcirculatory impairment. We aimed to assess feasibility of performing HSI, a noninvasive, contactless method to assess microcirculatory alterations, during trauma resuscitation care. METHODS: This randomized controlled clinical trial was conducted in a dedicated trauma resuscitation room of a level one trauma center. We included adult patients who were admitted to the trauma resuscitation room. Patients were allocated in a 1:1 ratio to the HSI group (intervention) or control group. In addition to the standard of care, patients in the intervention group had two hyperspectral recordings (HSR) of their hand palm taken. Primary outcomes were the treatment duration of the primary survey (until end of ABCDE-evaluation, ultrasound and evaluation by the trauma team) and the total resuscitation room care (until transport to definitive care) as well as the ability to perform measurements from all HSR. Secondary outcomes were analyses from the intervention group compared to HSI measurements of 26 healthy volunteers including an analysis based on the ISS (Injury severity score) (< 16 vs. ≥ 16). Care givers, and those assessing the outcomes were blinded to group assignment. RESULTS: Our final analysis included 51 patients, with 25 and 26 allocated to the control and intervention group, respectively. There was a statistically significant shorter median duration of the primary survey in the control group (03:22 min [Q1-Q3 03:00-03:51]) compared to the intervention group (03:59 min [Q1-Q3 03:29-04:35]) with a difference of -37 s (95% CI -66 to -12). Total resuscitation room care was longer in the control group, but without significance: 60 s (95% CI -60 to 180). From 52 HSI, we were able to perform hyperspectral measurements on all images, with significant differences between injured patients and healthy volunteers. CONCLUSION: HSI proved to be feasible during resuscitation room care and can provide valuable information on the microcirculatory state. Trial registration DRKS DRKS00024047- www.drks.de . Registered on 13th April 2021.

Spectral organ fingerprints for machine learning-based intraoperative tissue classification with hyperspectral imaging in a porcine model.

Visual discrimination of tissue during surgery can be challenging since different tissues appear similar to the human eye. Hyperspectral imaging (HSI) removes this limitation by associating each pixel with high-dimensional spectral information. While previous work has shown its general potential to discriminate tissue, clinical translation has been limited due to the method's current lack of robustness and generalizability. Specifically, the scientific community is lacking a comprehensive spectral tissue atlas, and it is unknown whether variability in spectral reflectance is primarily explained by tissue type rather than the recorded individual or specific acquisition conditions. The contribution of this work is threefold: (1) Based on an annotated medical HSI data set (9059 images from 46 pigs), we present a tissue atlas featuring spectral fingerprints of 20 different porcine organs and tissue types. (2) Using the principle of mixed model analysis, we show that the greatest source of variability related to HSI images is the organ under observation. (3) We show that HSI-based fully-automatic tissue differentiation of 20 organ classes with deep neural networks is possible with high accuracy (> 95%). We conclude from our study that automatic tissue discrimination based on HSI data is feasible and could thus aid in intraoperative decisionmaking and pave the way for context-aware computer-assisted surgery systems and autonomous robotics.