Airway label prediction in video bronchoscopy: capturing temporal dependencies utilizing anatomical knowledge.

PURPOSE: Navigation guidance is a key requirement for a multitude of lung interventions using video bronchoscopy. State-of-the-art solutions focus on lung biopsies using electromagnetic tracking and intraoperative image registration w.r.t. preoperative CT scans for guidance. The requirement of patient-specific CT scans hampers the utilization of navigation guidance for other applications such as intensive care units. METHODS: This paper addresses bronchoscope tracking by solely incorporating video data. In contrast to state-of-the-art approaches, we entirely omit the use of electromagnetic tracking and patient-specific CT scans to avoid changes in clinical workflows and additional hardware requirements in intensive care units. Guidance is enabled by means of topological bronchoscope localization w.r.t. a generic airway model. Particularly, we take maximally advantage of anatomical constraints of airway trees being sequentially traversed. This is realized by incorporating sequences of CNN-based airway likelihoods into a hidden Markov model. RESULTS: Our approach is evaluated based on multiple experiments inside a lung phantom model. With the consideration of temporal context and use of anatomical knowledge for regularization, we are able to improve the accuracy up to to 0.98 compared to 0.81 (weighted F1: 0.98 compared to 0.81) for a classification based on individual frames. CONCLUSION: We combine CNN-based single image classification of airway segments with anatomical constraints and temporal HMM-based inference for the first time. Our approach shows first promising results in vision-based guidance for bronchoscopy interventions in the absence of electromagnetic tracking and patient-specific CT scans.

Feasibility and safety of a three-dimensional anatomic map-guided transseptal puncture for left-sided catheter ablation procedures.

AIMS: Transseptal puncture (TP) for left-sided catheter ablation procedures is routinely performed under fluoroscopic or echocardiographic guidance [transoesophageal echocardiography (TEE) or intracardiac echocardiography (ICE)], although three-dimensional (3D) mapping systems are readily available in most electrophysiology laboratories. Here, we sought to assess the feasibility and safety of a right atrial (RA) 3D map-guided TP. METHODS AND RESULTS: In 104 patients, 3D RA mapping was performed to identify the fossa ovalis (FO) using the protrusion technique. The radiofrequency transseptal needle was visualized and navigated to the desired potential FO-TP site. Thereafter, the interventionalist was unblinded to TEE and the potential FO-TP site was reassessed regarding its convenience and safety. After TP, the exact TP site was documented using a 17-segment-FO model. Reliable identification of the FO was feasible in 102 patients (98%). In these, 114 3D map-guided TP attempts were performed, of which 96 (84%) patients demonstrated a good position and 18 (16%) an adequate position after TEE unblinding. An out-of-FO or dangerous position did not occur. A successful 3D map-guided TP was performed in 110 attempts (97%). Four attempts (3%) with adequate positions were aborted in order to seek a more convenient TP site. The median time from RA mapping until the end of the TP process was 13 (12-17) min. No TP-related complications occurred. Ninety-eight TP sites (85.1%) were in the central portion or in the inner loop of the FO. CONCLUSION: A 3D map-guided TP is feasible and safe. It may assist to decrease radiation exposure and the need for TEE/ICE during left-sided catheter ablation procedures.

3D mapping for the identification of the fossa ovalis in left atrial ablation procedures: a pilot study of a first step towards an electroanatomic-guided transseptal puncture.

AIMS: Transseptal puncture (TP) for left atrial (LA) catheter ablation procedures is routinely performed under fluoroscopic guidance. To decrease radiation exposure and increase safety alternative techniques are desirable. The aim of this study was to assess whether right atrial (RA) electroanatomic 3D mapping can reliably identify the fossa ovalis (FO) in preparation of TP. METHODS AND RESULTS: Between May 2019 and August 2019, electroanatomic RA mapping was performed before TP in 61 patients with paroxysmal or persistent atrial fibrillation. Three electroanatomic methods for FO identification, mapping catheter-induced FO protrusion, electroanatomic-guided analysis, and voltage mapping, were evaluated and compared with transoesophageal echocardiography (TOE). Mapping catheter-induced FO protrusion was feasible in 60 patients (98%) with a mean displacement of 6.8 ± 2.5 mm, confirmed by TOE, and proofed to be the most valuable and easiest marker for FO identification. Electroanatomic-guided analysis localized the FO midpoint consistently in the lower half (43 ± 7%) and posterior (18.2 ± 4.4 mm) to a line between coronary sinus and vena cava superior. Analysis of RA voltage maps during sinus rhythm (n = 40, low-voltage cut-off value 1.0 and 1.5 mV) allowed secure FO recognition in 33% and 18%, only. A step-by-step approach, combining FO protrusion (first step) with anatomy criteria in case of protrusion failure (second step) would have allowed for the correct localization of a TP site within the FO in all patients. CONCLUSION: Right atrial electroanatomic 3D mapping prior to TP proofed to be a simple tool for FO identification and may potentially be of use in the safe and radiation-free performance of TP prior to LA ablation procedures.

Evaluation of a new ultralow-dose radiation protocol for electrophysiological device implantation: A near-zero fluoroscopy approach for device implantation.

BACKGROUND: Radiation is one of the main hazards of electrophysiological device implantation, and insertion of cardiac resynchronization therapy (CRT) devices in particular is associated with high radiation doses. OBJECTIVE: The purpose of this study was to evaluate the impact of a new ultralow-dose radiation protocol on radiation doses, success rate, and safety of electrophysiological device implantations. METHODS: In 2018, we established a new ultralow-dose radiation protocol (reduced pulse width, increased thickness of minimum copper filters, reduced detector entrance dose, reduced pulse rate, optimized image postprocessing settings) for de novo device implantation at our hospital. A total of 1173 patients (11% single-chamber devices, 69% dual-chamber devices, 20% CRT devices) were analyzed. Five hundred twelve patients (44%) in the ultralow-dose group were compared to 661 patients (66%) treated during 2017 with a conventional low-dose protocol. RESULTS: With the ultralow-dose radiation protocol, effective doses could be reduced by 59% (median 0.25 [interquartile range: 0.11-0.63] vs median 0.10 [interquartile range: 0.03-0.28] mSv; P <.0001) per procedure without a significant change in procedure time (P = .5). This dose reduction could be achieved without decreasing procedure success (P = 1) or increasing complication rate (P = .8). Male gender, higher body mass index, increased procedure and fluoroscopy times, and use of the conventional radiation protocol were independent predictors of higher radiation doses in multivariate regression analysis. CONCLUSION: By establishing a new ultralow-dose radiation protocol, we could significantly decrease radiation exposure, reaching the lowest radiation doses for electrophysiological device implantation reported to date.

REduction of THRomboembolic EVents during Ablation using the laserballoon: The RETHREVA registry.

INTRODUCTION: Cerebral events (CEs), including silent (SCEs), are a known complication of left atrial catheter ablation (LACA) in patients with atrial fibrillation. The aim of this prospective registry was to gain more information about CEs during laserballoon LACA and to reduce the risk of their occurrence. METHODS AND RESULTS: We enrolled 74 patients (age 61 ± 11 years; 74% male; CHA2 DS2 -VASc 1.9 ± 1.4). Cerebral MRI (1.5 Tesla) was performed to detect CEs. ASPItest identified aspirin-resistant patients (ARPs). All bleeding complications were recorded. Due to an initial high CE rate, we evolved our clinical procedure step-by-step arriving at an optimized protocol: -Group 1: heparin after single transseptal puncture (TP), activated clotting time (ACT) > 300 seconds (CE: 64.3%). -Group 2: heparin after double TP, ACT > 300 seconds; 2a without (CE: 45.5%, RRR: -29.2%) and 2b with additional intravenous aspirin (CE: 36.4%, RRR: -43.4%; excluding ARP: 30%, RRR: -53.3%). -Group 3: heparin before double TP, ACT > 350 seconds; 3a without (CE: 54.5%, RRR: -15.2%) and 3b with aspirin (CE: 18.5%, RRR: -71.2%; excluding ARP: 8.7%, RRR: -86.5%). Larger LA diameter > 44 mm (OR: 1.149, P  =  0.005) and no aspirin use (OR: 4.308, P  =  0.008) were CE risk factors in multivariate logistic regression. In those patients receiving aspirin, aspirin resistance (OR: 22.4, P  =  0.011) was an exceptionally strong risk factor. CONCLUSION: These data support the use of intravenous aspirin including monitoring of aspirin resistance in addition to ACT-guided heparin. An optimized protocol of heparin before TP, double TP, and intravenous aspirin in non-ARP resulted in a significantly lowered CE incidence and severity.