Hi everyone, this is Deep Sharma. I'm here to present my study, I know it's a long name. So it's a 3M study, we call it a Columbus study. So I'm from Columbus, Indiana, at Columbus Regional Health. My co-authors are Elizabeth Sevens, Jordan Thompson, and then I would like to extend gratitude to CHESS to accept our work here. Disclosures include consultant for Merit Medical, speaker for Merit, past speaker for Boston Medical, and then research support from Philips. None of this has anything to do with my current talk. So again, I mean this room does not need any introduction about lung cancer epidemiology, but just to emphasize, we have about 1.8 million cases, new cases worldwide, and then an economic impact of about $188 billion. And then there's a new lung cancer case diagnosis every two minutes and 20 seconds in the United States. So our new USPSC criteria, which were launched and recommended in 2021, double the number of patients, at least for screening for lung cancer. And then one thing I always emphasize is that screening finds lung nodules, but it does not diagnose cancer. That's just the first step, right? So we have to have accurate biopsy to determine what those nodules are from. So traditionally, fiber optic bronchoscopy has had a very poor yield, about 20% for peripheral lung lesions. And then in last few years, accessing peripheral lung nodules has been one single area in the world of IP that has grown exponentially. And then I anticipate, and like everyone else, that these lung nodules, the numbers that we diagnose as the screening programs grow and mature is gonna get exponentially higher in future years. So the reason we need accurate diagnosis is because our survival in lung cancer varies so significantly between stage one and stage four. And then the various technologies that we have launched in few years from robotics and navigation and cone beam CT, all of that have improved our diagnostic yield, but it still remains suboptimal, especially for lesions less than two centimeters. So the data to support that statement lies in the acquired registry in 2016, which showed our yield was like 47%. The NAVIGATE study increased it to about 69% and PRECISE in 2021 showed the yield of about 82%, especially for nodules less than two centimeters. And then there's obviously this variability on how you define the diagnostic yield from very strict criteria in ACQUIRE and then INTERMEDIATE, NAVIGATE, and then a little bit liberal in PRECISE study. So my study was a Columbus study, which is a three-arm study. We first launched the study between a superdimensional, which is an EMN-based platform with the 2D fluoro, comparing that with the same super-deep platform with cone beam CT. And then the second part of the study was comparing the second arm, which is the super-deep plus cone beam to robotic plus cone beam. So it became a three-arm and then called it the Columbus. So it's a retrospective and prospective study, pragmatic design allowed for built-in randomization. These patients underwent bronchoscopy in our center with one single operator between 2017 to 2022. We had 234 patients. And then the second arm was bigger arm or biggest arm in the study. So we included only first 80 cases. And then we followed the benign diagnosis for about 12 months with serial CTs, PET scan, or second biopsy if needed, or response to treatment. So our current procedural flow, we used the CT patient to go pre-op and send his spirometry, nebulization, induced paralyzed, ventilated with high P protocol, do an area inspection. We do the robotic guidance, do an expert CT, do the 3D segmentation, 3D overlay, and then use the augmented fluoro to obtain like FNA, brushings, biopsies, and BL, and then end with the EVA staging and then extubated and discharged. So the variables between the two or different arms, our arm one was EMN plus 2D fluoro, arm two was EMN plus cone beam, and arm three was robotic plus cone beam. And then the ventilation protocol, which is the Pioneer's protocol, we applied that to arms two and three, which is high P, low tidal volume, low FiO2. But the constants between different arms were the same operator, which is myself. We have rapid onset for every case, general anesthesia and paralytics for every case. Radial E-BUS was used, and then very similar biopsy instruments, including the needle, which was arc point between 18 to 21, single or triple brush, forceps, and then the BL. So these are our results. So demographic-wise, our age, gender, size was very similar between all three arms. We did go a little bit more peripheral as our program matured and as our technology evolved throughout our program. As you can see, we went from 83 to 92% of our peripheral nodules, but it was not significantly different. And then obviously the size-wise, we went from 25 millimeter median to 26 and then 21. So these are our results. So our EMN and 2D fluoro, we had 78 lesions. EMN plus cone beam had 86, and then robotic plus cone beam had 70 nodules. And then our overall sensitivity was 77, then 91 and 92. And then our accuracy was 87 to 94 and 95. But if we stratified into smaller lesions, which is less than three centimeters, we went from a sensitivity from 70% to 93%, and an accuracy from 83 to 96. And then further subcategorized to lesions less than two centimeters, which is where we struggle in the pulmonary world to get accurate diagnosis. We went from 77% in arm one, 86% in arm two, and then 97.6 in arm three. And then if you look at the P values, there was significant difference in the arm one versus arm three between, for less than three and both less than two centimeter lesion. Procedure time also improved from 89 minutes. This is patient in to patient out. 89 minutes, 86, and then 77, which was again significantly different. But then the incidence of pneumothorax remains the same between 2.4, 2.3, and 2.9%. So our results concluded the use of combination of shape sensing, robotic navigation, which in our case was ION, and cone beam CT with augmented fluoro versus electromagnetic navigation with SuperD and CR. We were able to reach smaller lesions, more peripheral lesions, had higher sensitivity for malignancy and overall diagnostic accuracy. Reduce the procedure time without any change in complication rate. So the biggest strength of our study include that we have intermediate to strict definition for diagnostic yield. We had direct comparison between different diagnostic modalities and then minimal variables between study arms. But then our weakness include that we were a single center study with a single operator, and then we were unable to differentiate the impact of ventilation strategies from impact of the diagnostic modalities. And then our follow-up limited to one year, not two year. And this is our hybrid OR, or our pulmonary suite at my hospital. And that's all I have. My name is Alejandra Yuli-Mateos. I'm a research fellow from the Division of Pulmonary Allergy and Sleep Medicine at Mayo Clinic Florida, and I have nothing to disclose. The lesson objectives here are to review advances in bronchoscopy, describe new techniques for improving bronchoscopic diagnostic yield, and compare cryobiopsy to fine needle aspiration for diagnosis of peripheral pulmonary nodules. As we know, and it's widely known, lung cancer is the leading cause of cancer-related death in the United States and worldwide for men and women. It is estimated at over 127,000 deaths in the US alone for this year. Now new technologies, such as shape-sensing robotic-assisted bronchoscopy, in combination with three-dimensional fluoroscopy, have emerged to improve diagnostic yield for peripheral pulmonary nodules. Cryobiopsy, which uses frozen nitrous oxide for surrounding tissue sampling, has been successfully utilized in other settings, such as evaluation of lung transplant rejection using the CryoProbe 2.4 and 1.7 millimeters. Recently, the new 1.1 millimeter CryoProbe fits through the shape-sensing robotic-assisted bronchoscopy, and allows a 360-degree tissue acquisition, potentially improving tissue sampling and diagnostic yield. This is an image of how the bronchoscopy laboratory is set up in our institution, with the OR table in the center, the ion robot and robotic arm in the far left, the 3D fluoroscopy, C-arm on the right, and then our cryobiopsy machine in the near left. We conducted a retrospective study on patients who underwent shape-sensing robotic-assisted bronchoscopy for evaluation of peripheral pulmonary nodules, using fine-needle aspiration and cryobiopsy with a 1.1 millimeter CryoProbe. All procedures were performed using radial IBUS and 3D fluoroscopy, as well as IBUS TBNA for mediastinal and hilar staging. Sample adequacy was assessed by rapid on-site evaluation, and we eliminated non-diagnostic samples from malignancy sensitivity, but they were included in the overall diagnostic yield. The following is a short video of how we performed the procedure. So after we confirm the location on radial IBUS, as well as 3D fluoroscopy, we remove the robotic catheter and insert the CryoProbe through the robotic catheter. Then, visualizing on the 2D fluoroscopy, we take the sample, which we can see here in this image. And then we prepare the tissue sample for tissue handling for touch prep. Now, the first initial thought is the larger amount of tissue sample that we can gather compared to fine needle aspiration. This will go then to be evaluated by rapid on-site evaluation. And then finally, the tissue on formal for final histologic analysis. Now, we evaluated 60 nodules on 53 patients with an median age of 68 and a predominantly female population. The median size was 1.71 centimeters by 1.26, and the median distance to chest wall was 1.78 centimeters. Nodules were predominantly located in the upper lobes, most frequently in the right upper lobe, followed by the left upper lobe. Malignancy was found in 41 of our 60 nodules. The most common was adenocarcinoma, and the most frequent type of nodule was solid. The median duration of procedure was 62 minutes, including ibustibulinae for mediastinal staging. The most common or frequent needle size was 23. And we reported no complications that require intervention, abortion of procedure, or admission for observation. Now, the overall diagnostic yield for the procedure was 86.7, and sensitivity for malignancy was 92.7%. When comparing both diagnostic tools, the cryoprobe was diagnostic in, had a diagnostic yield of 86.7%, while fine needle aspiration, 76.5%. We found a statistically significant difference between them. Cryobiopsy was overall diagnostic in 12 lesions where fine needle aspiration was non-diagnostic, and specifically nine lesions, malignant lesions, where fine needle aspiration was non-diagnostic. Shape-sensing robotic-assisted bronchoscopy using the 1.1 millimeter cryoprobe resulted in a significantly higher diagnostic yield, with no reported complications for patients undergoing evaluation of peripheral pulmonary nodules. The use of the novel cryoprobe and concurrent hyaluronidin mediastinal staging could potentially enhance diagnostic accuracy and early detection and staging of lung cancer. Thank you. Thank you, everyone, for coming. My name is Pandit Patel. I'm a clinical assistant professor at Stanford University. I have no financial disclosures. Today, I'm presenting an interim analysis comparing the diagnostic accuracy and procedural outcomes for CT-guided transthoracic needle aspiration to co-mem CT-guided bronchoscopy for the diagnosis of peripheral pulmonary lesions or pulmonary nodules. Now, why do we care about this topic? As I'm sure everyone knows, we're discovering nodules more than ever because we're scanning everyone more than ever due to two main factors. One, lung cancer screening, which screens about nine million Americans a year, a third of whom are found to have nodules that require follow-up, and two, another five to seven million Americans that are scanned and found to have incidental nodules. Currently, three major methods for biopsy diagnosis exist, surgical resection, which has the best diagnostic yield, but is quite invasive, and two minimally invasive methods, CT-guided transthoracic needle aspiration, which is currently the gold standard for minimal invasive diagnosis, and bronchoscopy. The data for the latter two I present on the screen. On the top left, you can see for CT-guided TTNA, our diagnostic accuracy is fairly high, around 92%, but it has a non-trivial complication rate, around 20% for pneumothorax and 3% for hemorrhage, with a wide range of these metrics reported in the literature. In contrast, bronchoscopy, when done without co-mem CT, has a lower diagnostic accuracy, although in recent years, it has been improving towards 80%, but it has a better safety profile. And finally, there's some literature to suggest bronchoscopy, when combined with co-mem CT, can have a diagnostic accuracy comparable to CT-TTNA, but a better safety profile. Currently, no studies exist to date directly comparing any of these bronchoscopic methods to CT-TTNA for peripheral pulmonary lesion diagnosis. So we decided to study exactly that. The purpose of our study was to compare the diagnostic accuracy and procedural outcomes for CT-guided TTNA to co-mem CT-guided robotic bronchoscopy with the Monarch platform for peripheral pulmonary lesion diagnosis. This was a single-center retrospective review of adults over a two-year period, and I presented interim analysis with 50% data collection complete. Our primary outcome was diagnostic accuracy using a strict definition of either, one, a biopsy result yielding a specific benign or malignant pathology, two, organizing pneumonia that resolved on follow-up imaging with or without treatment, and three, a scar that persisted without change over six months of radiographic surveillance. Data was analyzed using RN-GMOV, and a two-tailed p-value less than 0.05 was considered statistically significant. We had a fairly decent sample size with just under 250 patients in each arm and were well-matched with regards to median age, BMI, gender, and ethnicity, with some differences in prior history of smoking and prior history of non-lung malignancies, but overall, well-matched. Overall diagnostic accuracy was found to be comparable, 87.1% in the comeme CT-Bronc arm, and 84.9% in the CT-TTNA arm with a p-value of 0.510. When diving deeper into the data and looking at nodule density, whether solid, part-solid, or ground glass, there was no difference. Blue bars here represent the comeme CT-Bronc arm, orange bars the CT-TTNA arm with p-values listed at the top. When looking at nodule size separated into 10-millimeter increments, again, there was no difference in diagnostic accuracy. And finally, when looking at low-bar location of the nodule, again, there was no difference. So whether we looked at nodule density, size, location, or overall diagnostic accuracy, there was no difference. However, with regards to complications, there was a difference. In the CT-TTNA arm, a little more than 40% of patients had mild to moderate bleeding, 30% had pneumothorax-needing observation, and just under 10% had pneumothorax-needing intervention. But in this comeme CT-Bronc arm, all three complications occurred less than 3% of the time. This supports the fact that comeme CT-Broncoscopy here was far safer than CT-TTNA. Median procedure times took about 69 minutes to complete a comeme CT-Bronc case, but only 32 minutes to complete a CT-TTNA case. This was statistically significant. However, 63% of patients in the Bronc arm also got E-Bus for nodal staging. That added significantly to the length of the case, as can be seen by the histograms on the right. In the top one, a median time for a comeme CT-Bronc case with E-Bus was 79 minutes. The same case without E-Bus was faster at around 55 minutes, although still longer than 32 minutes for CT-TTNA. Now, this is most likely because our IR colleagues have been doing CT-TTNA for decades and are well-experienced. A comeme CT was a new technology to our group, and this median time reflects our initial two years with working with the technology, where we were much slower as we learned it, and then we got much faster later. So in conclusion, comeme CT-Broncoscopy was slower, but the majority of cases also included E-Bus. When we look at just those peripheral permanent lesions diagnosed as primary lung malignancies, you can see there's a difference between the Bronc and TTNA arm, and in this cohort, everyone in the Bronc arm got E-Bus for nodal staging, but only 10% of patients in the TTNA arm did. This could be because our IR colleagues got more referrals for METs from non-lung malignancies that don't necessitate E-Bus, but it could also mean the patient needed a second procedure, a point we tease out in the second half of the table. In the Bronc arm, seven patients needed a second procedure, but almost double, or 12 patients needed a second procedure in the TTNA arm, and seven out of the 12 needed E-Bus as a second procedure, and two out of those seven, or 29%, had nodes positive for cancer. For our audience, this is an important point, because if your patient needs a second procedure that could potentially upstage their cancer, it could delay their care, affect their candidacy for surgical resection, and ultimately, their long-term survival. Strengths of our study included a large volume of patients, a retrospective design that reflected real-world referral patterns, curbing selection bias, and all slides were analyzed by the same pathology group, which helped reducing pathologist bias. Limitations, our experiences may differ at other centers due to inherent practice procedural variation. Results are not generalizable to non-comem-CT bronchoscopic methods, and we can't account for every preference or bias among referring providers. So in conclusion, this was the first study to directly compare the diagnostic accuracy of comem-CT guided bronchoscopy to CT-TTNA, which is the gold standard currently for the diagnosis of peripheral pulmonary lesions. We demonstrated comem-CT bronchoscopy was just as accurate as the current gold standard with no differences in nodule size, location, or density, and comem-CT bronchoscopy may have benefits, such as a better safety profile and the reduced need for subsequent procedures as it offers concurrent E-bus nodal staging. Further studies are still needed to clarify potential differences that could exist between these biopsy modalities and to understand when one might be favored over the other. So we hope everyone stays tuned for our final analysis, and a special thank you to our Stanford IP team of co-investigators, especially Drs. Hermit Beatty and Brian Schaller, who helped establish our comem-CT program, without whom this research would not have been possible. And thank you, everyone, for listening. I'm Debra Mee. I'm a third-year medical resident at Hopkins Bayview, and I'm here to talk about outcomes of atypical pathologic changes following robotic-assisted bronchoscopy. And I have no disclosures. Robotic-assisted bronchoscopy is an emerging technology with the potential to increase the diagnostic yield of peripheral nodular sampling. And despite technologic advances in guided bronchoscopy, like EMN, radial E-bus, comem-CT, the diagnostic yield of bronchoscopic approaches for peripheral nodules has been inconsistent, ranging from 40 to 70 or 60%. I think the slides are not updated, actually. Oh, okay. That's fine. Currently, there are a few studies reporting on diagnostic yield of robotic-assisted bronchoscopy which are limited by small sample size and varying definitions, with no standardized approach to the measurement of diagnostic yield. And within that, atypia is a controversial diagnostic category, and there is limited data reporting pathologic follow-up and outcomes in this subset population. By nature, the malignant category is generally diagnostic and specific, but non-malignant is highly variable. And as smaller nodules are sampled, the probability of malignancy is lower, leading to a higher incidence of non-diagnostic results. Atypia is considered non-diagnostic, but this category can be subdivided further into specific diagnoses as patients are followed over time. We... 156 patients underwent RAB at a single center from about November 2020 to January 2022, so a 15-month period, and we extracted patient demographics, nodule characteristics, pathology reports. All cases were performed using the ION system with shape-sensing technology, and we collected them either via FNA, transbronchial forceps biopsy, or both. Samples were considered diagnostic if there was confirmed malignancy or definitive benign finding. And atypia nonspecific benign findings were considered non-diagnostic. We stratified outcomes into two definitions, strict, which excluded atypia, and liberal, which included atypia, and we followed up these patients about a year after to see if they were diagnosed with or treated for malignancy. The mean nodule size was about 21 millimeters, mean distance to pleura, so center mass to pleura was 35 millimeters. 34% of the nodules were in the outer third of the lung, and about 30% had positive bronchus sign. Using a strict definition, atypia is non-diagnostic. The overall diagnostic yield of RAB was about 49%, and using the liberal definition, it was about 62%. And so that increased the yield about 13%. Malignant nodules of the 156 were 61, so 39%, and then within the non-diagnostic, a large proportion of them were atypia, so 21 out of 156, 13%. At the one-year follow-up of the atypia cases, about eight of them underwent repeat biopsy, and all eight of them had confirmed malignancy. 10 of them were presumed malignant and treated, and three of them were followed over time. There were no changes after a year. So overall, about 86% of the atypia cases were ultimately diagnosed and or treated for malignancy. The vast majority of cases with atypical pathologic changes following robotic-assisted bronchoscopy were positive for malignancy, and incorporation of atypia into the diagnostic yield increased the yield from about 49% to 62%. RAB findings of atypia has potential to shift estimates of diagnostic yield and may represent an actionable diagnosis. Prospective assessment with pathologic adjudication is needed to further understand the consequences of classifying atypia as an actionable diagnosis before consideration of classifying atypia as a diagnostic finding. All right, thank you very much. My name is David Wiese. I'll be talking to you about our study, Accessing Aerodipulmonary Lymph Nodes via Robotic-Assisted Bronchoscopy, a New Window of Opportunity. That's me, formerly of Memorial Sloan Kettering. I kept it on because that's where this study is, but currently at the West Palm Beach VA in Florida Atlantic University. I have no financial disclosures, so we're gonna take a quick start, talk about mediastinal lymph nodes. I hope everybody's ready to go over mediastinal staging and everything, but we're not gonna go over all that. Let's focus on, let's see if we can work this thing, right here, whoops, station five, station six. Not anything we go after with EBIS, not anything we go after on a regular basis. I'll take that back. Some people are very brave and will go after it with EBIS. We're used to station four, we're used to station seven, we're used to station 11 for a normal mediastinal staging. Station seven may be my favorite, but not stations five and six. Traditionally, via endobronchial ultrasounds, we go after, again, like I said, 12, 11, 10, seven, four, sometimes two, 3p. We don't go after five, we don't go after six, we don't go after eight and nine, at least for the bronchoscope. We'll change gears. This is a robotics talk. And everybody who's presented before has shown us the advantages and the improvement of technology that robotic bronchoscope has. A new technology for the sampling of pulmonary nodes, able to biopsy nodules with more accuracy and stability. For those who remember what super D felt like, just holding that scope for hours on end. You know, did I get it with radial, did I not? I think stability is a huge key here. And as what's been mentioned in a lot of the studies already, different tools for sampling. We talked about cryo already, we've talked about forceps, we've talked about needle aspiration. So can we kind of merge these two concepts? Can we use robotic-assisted bronchoscopy to access lymph nodes that were not previously accessible? Let's look at a case, at least one of the cases of our series. So we have a 29-year-old woman with a history of large B-cell lymphoma. Important to remember that diagnosis, who had completed her chemotherapy three months prior. She was referred to our service for new mediastinal findings of increased thickening along the aortic arch, and routine post-chemo surveillance. PET-CT was done, shows mild FDG avidity in that region, right adjacent to that aortic arch. After a bit of multidisciplinary review, we considered various options, including robotic-assisted bronchoscopy, percutaneous biopsy, or surgical biopsy approaches. I think at the session that we're in, you guys can take a guess on which approach we took. So just to set the scene before I play this video, I want you to focus here and here. I don't know if it's showing on here. Here and here. As the video plays, it goes by pretty quick. We identified our lesion via ultrasound, and placed a marker with a paper clip. All right, you see the EBIS, you see our paper clip, and you see that node coming in and out of the picture. Let's proceed to our robotic biopsy. Again, just to talk a few things before I play the video and narrate. It was kind of a long lesion, so we made multiple targets to go towards that one spot to make sure we had everything available. You'll see our first spin without any tools out of the robot. All right, you see our 3D imaging. And you can see our path right towards that lesion. See the mouse pointing at it. We move the paper clip out of the way so we can stick our tool, which in this case was a needle. And we repeat a spin. And you can see the tip of that needle is right in our spot. All right, here's a few stills if you didn't see it on the video. On the left, you have the EBIS. It's identifying where our spot is. You have our paper clip, giving us a rough estimate of where we wanna be. On the right, you have the robot with the needle out in what we believe to be the proper position. And you take a look at our 3D imaging. You have the path on the left and your tool and lesion on the right. So results of this case. Well, FNA reveals a clonal T-cell lymphocyte population, which is likely reactive. Remember, our patient had a B-cell lymphoma. Discussing with her lymphoma team, decision was made to monitor and follow-up imaging in three months. No treatment was given. And then the repeat PET-CT in three months revealed improved thickening and a resolution of FDG avidity. Now, that was one case of our five-case series. And that's case five that you see at the bottom. There are four other cases. Three other station six biopsies, all diagnostic for malignancy. Station five biopsy that was non-diagnostic. So it's not perfect, just like nothing is perfect, but right around an 80% yield in this small series, which is consistent with the data for robotic bronchoscopy. So what kind of things do you want to consider when you're going to do this more non-traditional biopsy using the robot? Well, let's start with provider and institutional experience. If you're starting off with your robot service, you're not going after these right away. You're not going after these without cone beam CT or a variation of it. I think case selection here is very important. You're not going after these without cone beam CT. Case selection here is very important. Multidisciplinary decision-making is very important. And obviously the importance of support staff. You need skilled nursing, skilled respiratory therapy if they're there for your cases. The importance of anesthesiology and their familiarity with the procedure, and especially with our ventilation protocols. And then obviously pathology, on-site pathology is extremely helpful as been discussed in some of the earlier cases. So conclusions to the study. Mediastinal stations five and six can be accessed with robotic-assisted bronchoscopy with cone beam CT. And this is a new area where the pulmonologist usually hasn't been going for these. It stresses the importance of multidisciplinary decision-making. And then obviously we're going to need a bit more experience in this five case series to optimize and standardize this technique. A special thank you to my co-authors, especially Dr. Chawla for the videos and some of the images. Thank you. Hi, I am Jim Dugan. At the time of this study with Dr. Vasarala and Dr. Bonifidel, who are the lead authors here, I was a third-year resident. I'm now a first-year pulmonary critical care fellow at the Cleveland Clinic. I was at Case Western Reserve University and University Hospitals at the time. No authors have any relevant disclosures. This was a non-funded, non-sponsored study. So as you all know, suspicious lung nodules can be sampled via guided bronchoscopy if standard bronchoscopy is not sufficient. But there's a paucity of data comparing the two guided bronchoscopy techniques, which are ENB and robotic-assisted. Contemporary data suggests that there may be some trends towards targeting smaller lesions with robotic-assisted bronchoscopy, but they have not been directly compared. The most recent Fleschner guidelines do not recommend sampling lesions less than eight millimeters and this means that technological capability should not translate to procedure necessity. So, excuse me, we hypothesized that robotic-assisted bronchoscopy at our center would change the characteristics of the lesions that we target with robotic-assisted bronchoscopy. With the incorporation of this new technology, smaller and perhaps more difficult to reach lesions were being sampled. So what we did was a single center retrospective study looking at adults who had guided bronchoscopy between June of 2020 and December of 2021. The reason we chose this time period was in May of 2021, the centers shifted from predominantly using electromagnetic bronchoscopy and adapted them on our platform. The exclusion criteria was any conversion between the two platforms, which did not happen. So what we looked at was lesion characteristics, location, both in the lobe and the lung centrality, the size, whether there was the radiographic appearance where that was with a bronchus sign, the characteristic of the lesion itself, and then the radiocrobe view. We used descriptive analysis with means, standard deviations, median, and interquartile ranges for continuous variables and proportions for categorical variables. I don't need to read the stats here, my least favorite part. So the primary outcome of this study was looking at the difference in mean lesion size for each guided bronchoscopy group. The secondary outcomes were those that I just mentioned, the lesion characteristics, location, the radiographic appearance, the centrality, and whether they had a radioprobe EBIS view, and whether there was a bronchus sign on the CT. So we screened 300 patients. This was retrospective, as I mentioned. We performed data collection on 100 patients and 50 were in each arm. Once 50 in each arm were met, we stopped data collection. We used a common procedure code to identify patients who underwent guided bronchoscopy, and there was no patient contact and no effect on management decisions in the bronchoscopy suite. So looking at our demographics here, this is a busy table, so I apologize. But essentially, both groups between the platforms were similar, mean age around 70. I think importantly, the history of malignancy was also similar between groups. And kind of combining their comorbidities and looking at their overall disease severity with ASA grading, most patients were in the two to three grade, which was similar between groups, as I mentioned. A majority of patients did not have a PET scan within 30 days prior to the bronchoscopy, and almost everyone had no prior tissue diagnosis. So looking at our primary outcome, which is lesion size, you can see that the electromagnetic arm, the mean lesion size was 20.2 millimeters, and the robotic arm was 18.4, which did not meet statistical significance. There was a standard deviation of about eight millimeters. Most of the lesions that were sampled were in the 10 to 30 range, although there were a handful that were less than 10 millimeters, which I will show on this slide here. So looking at the subgroup analysis, the robotic-assisted arm did have a slight trend towards targeting smaller lesions, less than 10 millimeters. However, this, again, did not meet statistical significance and the numbers are small. So looking at our secondary outcomes, most lesions that were sampled were in the upper lobes, no difference in location that was targeted between the platforms. The radiographic appearances were also similar, as were lung centrality, so not targeting more peripheral lesions between the platforms. Similarly, no difference between the radioprobe views. I feel like I'm repeating myself a little bit, but largely the secondary outcomes were similar between the platforms. So what we concluded was that, although in this study, looking at 100 patients total, there was a greater proportion of targeting smaller lesions, less than 10 millimeters in the robotic arm. This was not statistically significant, and future studies looking at larger sample sizes and further delineating this question would be beneficial. Limitations of the study, this was a single center, a retrospective study with, again, limited sample size, and the prior speakers all discussed diagnostic yield. We did not look at diagnostic yield in this study, so that would be another future consideration. Thank you.

robotic-assisted bronchoscopy

diagnosing lung nodules

biopsy accuracy

cone beam CT

cryobiopsy

fine needle aspiration

CT-guided bronchoscopy