Send Multiplex Us Currency with airSlate SignNow

Send multiplex image

right um grafton from the eastern seaboard and good morning to um all of you who are in other time zones um my name is sesha tekor and i'm the technical sales specialist for high content screening solutions and i welcome you all to the connective science um high content screening user group week this is the fourth stock uh of this series and with me are my colleagues brent sampson and uh dr neil durso and we as a team support the north american high content product portfolio our user group meetings have historically been for our user and prospect base and more importantly these events have largely focused on highlighting some of the excellent work one of which you'll see today by uh scientists utilizing the high content screening analysis platforms and i personally hope that this week you will see the value uh if you've already seen the other three talks you would have definitely seen the true value and part of the technology and today's talk is going to highlight that as well this is going to be for both your needs and workflow our user group meetings have always been a great forum to bring all of us together as a community and share the value of thermo fisher salamix high content solutions across the board today to round off this week of high content speakers i'm proud to introduce a good friend of mine dr michael johnson not many people can gloat about this but michael johnson has been the forbes 30 under 30 honorary for both 2017 and 2019 and that's a that speaks volumes to begin with uh he's uh the ceo and co-founder of a company called visicollink based out of new jersey it's a bio-imaging company that's spun out of rutgers university uh and it started back in 2016 and michael co-founded this company along with his phd candidate colleagues tom viliani and nick crider michael to say the least has a diverse background with experiences both in technical and business considerations of basically running biotech business which is a strong four day um he did his undergrad at muhlenberg college and after that he spent three years in various roles in johnson johnson where he concurrently pursued his phd in applied microbiology at rodgers his research background has largely focused on a wide range of applications and projects from ranging from remote sensing research with nasa which is which is just brilliant and to building light sheet microscopes and uh producing biofuels as well so he's got a very diverse background and prior to launching this call michael worked on several other biotech entrepreneurial pursuits and is very passionate about translating cutting-edge research into life-changing commercial technologies today he's going to be talking about multiplexing imaging and analysis of slides using high content imaging and um i look forward to that this talk michael uh for all the attendees i believe you all will all be in mute during the seminar but uh feel free to type in your questions um on the chat box there is one right at the bottom right uh the speaker can see these questions and so can the panelists i.e brent and i um so they might be able to the speaker might be able to respond to these questions almost immediately but the goal is for us as panelists to kind of collate these questions and bring it up as a q a session after the talk this kind of keeps the flow going but michael please feel free to jump in and answer any questions immediately as you go along um so without further ado i give you dr johnson uh michael the stage is all yours excellent well thank you for the introduction let me just share my screen here cool okay my screen yeah just to uh i didn't see victoria i didn't see you were in the panel as well uh i we have victoria tane in the panel as well and she's one of our colleagues from the western uh coast actually mr imaging analysis of slides using high content imaging but i was also going to focus on some of the other 3d cell culture assays and work that we've done our company we've been a user of the cx-7 for a few years now um we absolutely love the instruments and i would say this um it's not a testimonial i'm not being paid for this but the machine has been incredibly reliable for us we're running assays every single day we're running extended time points uh we need machine that's very very reliable and the cx-7 has fit that bill for us so we've been very happy with it i don't think we've had any hardware or software issues and uh over about three years now so overall we're super happy with the uh the machine and running and imaging business having a reliable piece of equipment is absolutely everything so to uh to jump into it just a little bit of background on our company just to understand where we're coming from with our feedback here the focus of our company is transforming tissues into insights and that using uh do imaging other tools that we've developed but ultimately we're focused as a contract research organization to help pharma companies address their most challenging drug discovery research questions looking at pk looking at talks looking at safety looking at efficacy we're trying to help pharma companies generate terabytes of data and distill that data down into useful insights that allow part of the drug discovery and process the company today we offer services in a couple key areas our expertise is imaging image analysis and advanced cell culture so we do 2d and 3d cell culture assays we do what we call hole mount imaging which is using tissue clearing and 3d imaging on large tissues digital path 3d cell culture assays and then a bunch of custom image and image analysis projects for clients now to dive right into the focus of this talk here is immuno oncology and the second focus i'll talk about is nash these are two disease areas that we do work in a lot and where three culture models and advanced imaging tools are particularly helpful for a couple of reasons so in the iowa space cancer cells can escape immunosurveillance by avoiding acute inflammation and tumor cytoactivity they can promote chronic inflammation which can encourage tumor growth and our objectives of io are strengthened or to modify the immune system to target these cascades through immunotherapies and there's two different types of immunoecology therapy approaches today there's passive and there's active i won't dive into these in a lot of detail here but the main focus of why we're using high content imaging or using different types of multiplex analysis is that there are a lot of different types of immune cell subtypes of players and an immune system space and we need to image a lot of different types of cells all at the same time which is where this concept of multiplex slide imaging and tissue imaging comes to and where cx7 provides a lot of value one of the common questions that we're looking at in this space is immune cell infiltrations we're looking at tumor infiltrating immune cells as it's associated with better prognosis for cancer patients typical analysis involves sectioning of tumor samples and processing for ihc to label t cells but with in vitro assays we're able to actually monitor them in 2d formats using boyden chambers or chasses and then lastly here three cell culture models which are a relatively new approach for studying immune cell filtration allow us to use a more relevant model to assess how we're actually invading into tumoral cells so again a traditional technique for looking at immune cells and immune profiling is using immunohistochemistry problem here being we're able to look at one single slice from one single tissue at a time usually for one or two so we're not of data that we're interested in for many different subtypes of immune cells multiplexing approaches that are out there and available today there's fluorescence there's a couple different fluorescence-based approaches for multiplexing and studying this this io landscape there's the fluorescence approaches are rely upon fluorescent probes tagged to antibodies there is the um the scissif fluorescence bleaching approach there's co-detection by indexing or fluorescent immunohisto pcr this fluorescent termite mediated application and then lastly there's antibody stripping which is what i'll talk about today in this segment on multiplex slide labeling technology which i won't dive into in great detail but it's a really cool technology is scitoff or imaging mass cytometry this is a way to use mass cytometry to actually create imaging data it has the ability to generate lots of markers from a single slide but it's very slow it's quite expensive and you can only image a small area from a slide at a time so these are the different approaches that we have to studying slides and to assess the immune profile of slides that we're looking at now the immune imaging mass cytometry approach which i just left off there we can image through a blade we have metal tag antibodies and we're able to get about 37 plex imaging from a single slide but it's a really slow approach for actually acquiring data from slides so we're able to use it to look at immune cells we're looking at how they into high proliferating areas or areas of interest on our slides we're able to answer many questions by looking at many different types of cells at the exact same time and we're able to look at a lot of different labels i'm just showing three at a time here just because chuck that's tough to show many different labels from a single slide as you run out of colors after three or four colors but the approach allows us to get a lot of data from a single slide and we can go up to about 40 markers with imaging mass cytometry the uh the problem with imaging mass cytometry though and a lot of the other techniques that are out there for multiplexing are that they are quite expensive they require proprietary reagents they require expensive imaging equipment that you don't have in your lab or the antibodies that are available are only available for certain targets that you might not be interested in technology that we've been working on for a while now is something called easyplex and this is an easy to use multiplexing approach that's compatible with any imager that's out there and i know the cx7 is typically um slated for cell-based assays but we've put it through the paces for almost every type of application you can think of from imaging large one millimeter thick brain slices to whole amounts of retinas to organoids and this is just one other example of how we've put the cx7 to use as a slide scanner as a slide scanner it's actually relatively comparable to some of the other slide scanners that are out there as far as time goes we're able to actually image whole slides for multiple channels so with this easyplex approach that we've developed we're able to wash antibodies off of slides and in between panels of four to five uh labels at a time we wash off and then relabel and re-image and through this process we're able to image many labels from a single slide all using the uh the cx7 lzr that we have in our lab um so just a little more background which we're uh we're labeling the slide we're imaging we're removing that um that set of uh labels we're washing them with new antibodies and new labels and then re-imaging again um and we're able to show that we can actually remove those antibodies completely there's no noise left behind we're able to get 8 12 15 plex from a single slide using our cx7 in our lab [Music] so the approach really simply just to give an overview of what we're actually doing is that we're labeling slides with four antibodies plus dappy two to three antibodies are typically conjugated because we have only so many species to work with ideally we're using secondary labeling for only our most expensive antibodies to reduce costs while multiplex thing is really cool it provides you a lot of great data it is actually pretty expensive per slide the main driver of that costs are your primary antibodies so always looking at ways to reduce that if you choose your antibodies appropriately you can definitely reduce the overall cost of doing multiplex imaging we then image these slides with our 67 lzr using 3d printed slide holders i know thermal also offers slide holders that you can buy as well through them we 3d print some of our own for custom size slides and different formats that we have here and then we remove the labels with um easyplex reagent that we've developed and we repeat the process of labeling and imaging again and again and again and on the back end one of the challenges with multiplex labeling which is a stumbling block for a lot of researchers is co-registration so you get one panel you relabel and you re-image with the second panel of labels and a problem always with imaging is those two panels don't align perfectly so we at visit call we've developed software for code so aligning those two panels with one another using the dappy channel so adapting channel will be the same between those two slides and we're able to use that to co-register all the other biomarkers together using this elastic co-registration process that we've developed in the next few weeks you'll see this coming out this will be an open source tool that we're giving out for free so if folks use this approach it's a tool through image j that anybody can access and freely use but it's a piece that allows multiplexing to be adopted by any lab you won't need special equipment you won't need special tools you can use the antibodies you have today and if you have a cx7 lcr you can be up and running with multiplexing right away you don't need an expensive piece of equipment to get up to speed so um be on the lookout for that that'll be coming out the next few weeks here we're launching this reagent and also this um software in the next few weeks but it pairs seamlessly with the cx7 for routine slide scanning and multiplexing which is really cool and it really drops the barrier to adoption you've seen this all before but we have two image data sets with dappy we have panels associated with each daffy channel quite fit exactly with one another so we're taking those dappy channels we're aligning them with an elastic co-registration approach and then all of our other channels are going to line up appropriately as well and then we can actually do our analysis after the fact of all the data that we've generated and one other example um this is again plex panel we're looking at colon here we're using our 7lcr for the data generation but we're getting really good slide scanning uh from this you see a little bit of tiling here but overall the quality of data is fantastic from the cx7 and uh relatively fast compared to even uh slide scanners are designed to do just slide scanning so it's a pretty cool tool i don't think a lot of folks are using high content imagers to do slide scanning but you have the ability to do it and you can do it pretty easily pretty easily and seamlessly with the cx7 uh but why do want to use multiplexing we can generate cool data but what do you actually do with it when we work with our pharma clients we're typically looking at a couple different problems in this case here this is data generated with the cx7 we're doing five labeling of a lung tumor and different channels that we're imaging with here from this data look at different types of segmentation approaches we can look at just the tumor area we can look at just the stromal area we can look at how t cells are invading into our tunal area and then if the animals are dosed with different kinds of amendments uh different treatments we can see how those different modulations change what's happening with our immune cells and how they're interacting with one another and we can take this data and of course quantify it so we can look at our cd3 positive cells and our stroma versus our tumor we can look at normalized population of cells by region so our cd45 positive cells pd1 positive cells where are they um what are they doing how they interacting with each other um we can get total cat you can look at every combination of markers you can think of whether they're in the stroma other than the tumor area how close the boundary of the tumor area they are and we can start to understand how this immune immune system is responding to our treatment how it's responding to disease and how we can modulate the immune system for the response that we're we're looking for and just to show you another um endpoint example here a common question that we're asked is how are our t cells infiltrating into our tumor area so we can look at all of our individual t cells we can look at what they're labeling for whether it's cd3 or cd3 and pd1 and we can see how these different our two ml area which allows us to better understand what's happening and how we're modulating that immune response so multiplexing provides us with an enormous amount of data from a single slide slides from immuno oncology studies are precious they are expensive we do not have a lot of tissue material and this approach allows us to generate a huge amount of data and the process itself is also non-destructive so we can come back with other assays we can come back with traditional h e i c afterwards rnac qpcr and look for other types of data from these tissues that we've looked at with multiplex slide labeling so this is a really powerful tool we're able to answer a whole bunch of questions that traditionally we're not able to answer with standard immunohistochemistry or standard imaging and we can really dig deep into what's happening with the immune system and just to show one last example here we can look at masks so using this approach we can look for a mask of where our t cells and we can overlay that on top of the data that we're generating which allows us to have a much clearer picture of what's happening on our slides and sometimes we'll actually share this data with our pathologists who might be characterizing these slides using a traditional qualitative pathology approach we can give them this data to amend that and they can answer more complex questions by adding this data on top of their pre-existing approach for analyzing these slides so just overall a really cool way to generate data we're able to generate meaningful quantitative data from slides using this approach now the second topic which is also in line with multiplexing so generating lots of imaging data and lots of channel data from a single piece of tissue or in this case a 3d cell culture model is an initiative that we launched a few years ago called open liver in the 3d cell culture space it's a new space i mean at this point we've probably only gone from academia to industry in about five or six years and now 3d cell culture models are being used by pharma companies and biotech companies all over the place but the adoption has been rapid and the field is very new and we don't yet have standardization like we see in a lot of other fields in cell culture we don't have a case standard assays in this space so something we set out to do is to put out a couple standard 3d cell culture models that researchers could see how they're made they can see how they're procured see how they're validated and really open source them so all researchers have access to them if you want to learn more about it you can check out our website we closely refer to this as our open liver initiative and the need for in vitro liver models is that um our in vivo models our vivo animal models don't always tell us the true picture of what's going on or what's going to happen once we take a compound from our natural models to of course animal models and because of that there's this gap in the liver and vitro space for better or predictive in vitro models if we had a model that allowed us to jump ahead and not have to go through so many steps of animal studies and better predict what's going to happen we can earlier on and drug this dick that a drug is going to fail and kill it and not waste a whole lot of money pushing it forward into animal studies needlessly so there's a big gap today i won't go through these charts in a whole lot of detail here but there's a big gap today of predicting what's going to happen in vivo during in vitro assays and 3d cell culture models compared to their counterparts 2d cell culture models provide a lot of benefits both in the spatial context but then also in the fact they facilitate a lot of formation of morphologies that we don't see in 2d like bioconiculi we don't see the complex phenotypes in 2d that we see in 3d or better able to mimic in 3d what's actually happening in vivo compared to traditional 2d adherent cell culture systems now the open liver that uh model that we've developed is comprised of a couple different components um so you see here we have our hepatocytes these can be hepa rgs they can be primary human epidecites uh we have our liver sinusoidal endothelial cells or couple cells or cellulite cells we've glandiocytes a number of other key components of our liver models here and we designed this model to have similar composition to actual in vivo liver tissue and you see on the bottom here these images these are all done using our cx7 we're able to look at the composition of these models using um our imaging based approach and a lot of the work that we're doing we're generating these models ultra low attachment u-bottom plates we're able to generate these models pretty easily using that construct now other features of this model we can go glycogen storage transfer approach we have our sips that are functional in this model we're able to characterize both from a compositional standpoint and then also a functional standpoint the show that is actually mimicking what we're seeing in vivo and we can look at a bunch of different types of assays using this model so we can take really simple assays like drug-induced liver injury we can acetaminophen which is always your your standard go to and look at increasing concentrations of acetaminophen and we start to see our mitochondrial health get worse over time and our viability of course is getting worse over time as well so we can multiplex a couple different other markers to this assay so apoptosis if it's applicable we can look at cell proliferation we're able to look at many different markers at the same time using this assay approach and using these models another one is and we can look at indirect and mediated response here so we can look at viability look at with and without lps and we can see the difference between those two amendments here and again in this case using high content imaging again with the cx7 lzr system and we're able to get a lot of imaging based endpoints from these models but i would say that these endpoints aren't um now they are becoming of imaging there's a lot of folks that are doing similar endpoints without imaging the reason that we hear call rely upon imaging for a lot of our endpoints is that we're able to start multiplexing endpoints to answer multiple questions at the same time if you're looking at just viability you can use a dissolution based atp assay which is really models but as we look at diseases like nash and non-alcoholic fatty liver disease we're starting to look at really complex phenotypes that are tough to characterize using simple plate reader assays or simple wide field imaging assays where it starts to lend itself to high content imaging especially confocal analysis during the nash cascade we're looking at fibrosis we're looking at collagen deposition these are features and phenotypes that are really challenging to characterize with a lot of the traditional assays that are out there where multiplex imaging and vocal imaging become particularly important so for something like steatosis very simple part of the nash cascade this is not that complicated this is an assay that we can do with very simple 3d cell culture models we can use hefty twos we can use our open liver model but in this case we're doing a very simple imaging based approach and we're looking at just lipid accumulation but for more complicated readouts such like fibrosis we're looking at collagen deposition throughout these models collagen deposition is something that we can't truly appreciate if we're looking at a 3d cell culture model in 2d or if we're doing a dissolution based assay it's something we really need the ability to image these models with confocal microscopy something that we employ at vizical is tissue clearing so we use tissue clearing which allows us to take our high content and actually image the entire depth of these 3d cell culture models so we can look at collagen deposition throughout the depth of the model instead of just looking at the exterior a problem always with imaging based endpoints is that if you're only getting light to penetrate one two or three cell layers deep into one of these models it really limits your ability to characterize the model in its entirety we're ultimately switching from 2d to because 3d allows us to have spatial context but we're not able to get photons into our models we're not actually characterizing the entirety of the model so we see with things like nash looking at immune cell infiltration looking at spatial dose response curves the detail and the spatial context of these models really matters and unless you use tissue clearing you're just not going to see it you're only going to see the outside the model and the outside few cell layers which don't depict the entire population of cells within your 3d cell culture model so the example i'm just showing here is we're looking at one of these open liver models uh we've already so nothing's really happening for um college and deposition but if we look at tgf beta so we're stimulating this nash cascade we start to see a whole lot of cell death we start to see a whole lot of collagen deposition where if we take our other control here which is our um beta stimulation our alk5 inhibitor and we're ameliorating that collagen deposition to some degree we're seeing less cell death we're actually seeing that nash cascade reversed to some degree here or ameliorated at the very least and this is something again that you couldn't really appreciate without imaging so in this case this nash assay is totally dependent upon the ability images models in 3d to use tissue and also a really reproducible and robust system something that we've run into a lot with some of our other imagers is for assays like this where we're imaging models over extended period of time that imager cannot go down it cannot break it cannot have a problem we have to have that imager up and running all the time especially for a business like ours where we're doing contract research for pharma companies and timing is essential so again this assay just speaks volumes to the importance of high content imaging tissue clearing and also the capability that we have the cx7 of using lasers for excitation lasers allow us to actually get much better depth than leds in a lot of cases and this really allows us to get this assay with a really strong assay window and be able to provide our clients with a nash assay that means a truly unmet need today now with that i will um i'll wrap i think one of the last comments i want to have was that you know overall we've been super happy with our cx7 system and we've built out a lot of software over the last few years a stack on top of that system but i know and i'm sure sesh has talked about this a number of times each individual on the phone but they've built out a great suite of software on top of the cx-7 especially as new applications and 3d cell culture becomes more pertinent and those have been very helpful and i'm sure sasha has dived into a lot of detail with all of you but overall we've just been super happy with our cx7 system today awesome thank you so much michael really appreciate it there are a few questions on the from the attendees um i think let me read them out to you okay um i think one of the questions is from kalina and she asks do you use u-bottom plates for your um steroids i guess right i mean it doesn't specify spheroids but kalina i'm assuming it's for these feed-outs the liver steroids i i um i can answer that so we use um we use the ultra low attachment plates for 3d cell culture work that we do they just allow for very easy aggregation one of the challenges we have seen is with some don't use lasers there's a bit of challenge with actually focusing at the bottom of those and getting good imaging data and with leds in particular with the u-bottom plates you will have some image artifacts you'll have this weird halo effect around the bottom of the model that we've seen some of those factors get subtracted out and uh post-processing uh but overall with the lasers on our cx-7lc i'm hurting with the new bottom place we don't have any issues with focusing or imaging oh that's awesome so you don't have to pretty much take it uh from a u bottom and transfer it to a flat bottom right that's the question that she's been trying to ask there's projects where they'll send us place that are incompatible with imaging for one reason or another and i mean it's a nightmare anybody who's transferred steroids before knows how tedious it is and all of a sudden for an asset that should be high throughput you're adding hours of processing time and you're losing steroids in the process and it's it's a nightmare that's not a good way to automate it so yeah we just do all of our processing and imaging right in those u-bottom plates that's awesome that's fantastic um i'm not sure if there are any additional questions but i have a few questions of my own i hope you don't mind me asking um i really um i'm curious about that co-registration algorithm that you guys built um particularly um based on the dappy channel right i think are you saying that it uses the dappy to uh as a fiduciary mark right um exactly how far can you go because i i'm starting to see a few um of our scientist base um trying to investigate in complex issues like like a brain slice right but look for like 50 different markers but so they might want to do it like six or seven times right um so is it it does it lose its accuracy after a certain time or have you seen it be consistent across say if you stack up six or seven slices oh i see you're saying you're talking actual serial section right no the same section wiping it out and then restaining got it got it um so long as you're not damaging the epitopes and the tissue itself the processing you should be able to do it theoretically indefinitely however what we've seen that happens is you will start to degrade the proteins during the process with ours we've done five six rounds of real issues depending on how they're fixed and how they're processed but so long as the tissue isn't shrinking or expanding dramatically as you're processing it it really shouldn't be that much of an issue um but then again one of the other issues that you'll see is by that fourth round or that fifth round the tissue becomes more porous and the signal intensity will actually increase so as you get if you label the same marker five times in a row the fourth time or third time you might have a higher signal intensity than the first time so some other factors at play but those adapting signals they should be in the exact same spot almost every time and if not that algorithm is stretching the um the image dynamically to put them back together um so as long as you're preserving the tissue you shouldn't have any issues with that even six or seven times through okay that's that's fantastic i'd love to support it to use at some point um any uh any thoughts about 3d analysis i don't know whether you were there at next nick radius talk yesterday but he highlighted the utility of amira which is our 3d volumetric software that's that's built specifically for the cx7 workflow right um i i see a true value in uh in utilizing something like this alongside with amira so that you could address um the analysis piece in a three-dimensional space right have you ever thought of uh thought of that or any comments on that sure so we we got in the space um about four or five years ago and initially we built out a huge suite of software for our own internal use there's not not a lot of software existed at the time and those that did were pretty expensive and not that extensive so we built out a lot of software and i would say a lot of software that we've built is actually inherent in the software that nick had talked about um so i would say we kind of ball i wish we'd have to spend three years building our own software um the software that you guys have does answer a lot of the questions that we get asked by our researchers today it just didn't exist when we originally had launched our services but i would say yeah those special questions um you know to be answered in that software sometimes it's just it's trying to come up with clever solutions on how to answer reasonable questions something as simple as collagen deposition measure do you measure intensity with volume a combination thereof um you know of course algae on the outside of the cell or outside of a model will be more intense not on the outside even if they're the same quantity or volume so i think there's some complicated questions behind a lot of research questions that software definitely ends itself to be used to answer those questions i just wish it existed a few years ago yeah i mean it's still worth a shot i think we can have an offline conversation about how we can uh help you with the average software uh to see if there is some cross comparisons there right so excellent excellent that's awesome michael i uh i really appreciate this um i'm i'm just wondering if there are any additional questions from the panelists from the attendees um if not this was really nice uh it's it shows the versatility of the cx7 platform uh for not just imaging which we knew that it was capable of but to see the volume of stuff that you showed on tissue imaging i hope our attendees and our user base benefit out of that so i think it's it's very valuable information um one more question i think johnny wang has a question i'm just going to read out can you do serial staining tech for to be organized i don't know what that means but um i use uh johnny can you elaborate a little more on that on your uh comment box um are you saying um serial checking of 3d organized is that that what you're referring to michael can you see the the question that was on the right side um sorry i'm scrolling through right now um so i'm trying to find it over here sorry johnny he's asking can you repeat staining for 3d model oh i see what he means yeah yeah so you use um referrals to multiple rounds of labeling for one model it's something um we have tried a little bit but i would say we haven't extensively done it would be very cool um wash out the labels and cut back in some 3d cell culture models that is getting labels in is very challenging i would imagine getting them out and then back in will be further challenging some models like neuronal organoids are really hard to get labels to penetrate i mean it can take you a week to get your label to penetrate 50 or 100 microns into it um so i would it's feasible it just might be incredibly time intensive to do and by time intent i just mean it might take a long period of time label wash out and relabel again i know um the imaging mass cytometry people have gone up to 40 labels with a single organoid but you have that's really expensive to do um it's not really theoretically it's probably possible we just don't have a lot of experience with it i hope that answers your question johnny this is the um stairs and i wanted to thank you for uh such a great talk uh one of the questions i had was uh on the thickness of the tissues that you're imaging with the cx7 lzr um what sort of uh distance and spread in the tissues have you seen yeah the the the largest depth that we've ever done is on one of the mac tech uh full thickness epidural models um it's about millimeter in thickness that's the deepest we've ever gone um practically i would say most of what we're doing is 500 microns or less than 500 microns you start to get all sorts of problems with refractive index mismatch um and we've tried um using dipping objectives on other systems but with speaking clearing uh the different objectives don't actually buy that much value without tissue clearing they do but i always say so is this spot um it's also cable dependent some antibody labels just will not go that deep so usually for about 500 microns to a thousand microns is a sweet spot so in that range you can do like whole out retina so you can do one your mouse brain sections um but beyond that i would say you don't get good data even towards a thousand the quality and the intensity has declined quite bad thank you