Collaborative Solutions using X-ray Imaging Analysis with Josephine of CJEL

Episode Overview

Episode Topic: In this episode of Skeleton Crew – The Rad Tech Show, we delve into the groundbreaking world of imaging analysis with Josephine Jiyoun Arns, the Managing Director of CJEL Digital Imaging. Explore the multifaceted applications of imaging technology beyond conventional radiology, uncovering its transformative impact on research across various industries. From fluid movement in rocks to the biodegradation of plastic materials, Josephine provides insights into the diverse projects undertaken by CJEL, shedding light on the dynamic nature of their work.

Lessons You’ll Learn: Listeners are in for an educational treat as Josephine shares invaluable lessons derived from her extensive experience in imaging analysis. Discover the crucial factors beyond image capture, including the significance of sample preparation, meticulous image acquisition, and the intricate process of analysis. Uncover the challenges faced in the research field, such as funding constraints, and grasp the immense potential for students and professionals to enhance their careers through hands-on projects and advanced technology training.

About Our Guests: Meet Josephine Jiyoun Arns, a visionary leader and the Managing Director of CJEL Digital Imaging. With a background in chemical and resources engineering, Josephine has dedicated over two decades to pushing the boundaries of imaging analysis. Her expertise extends to customizing software, optimizing research processes, and collaborating with young talents to drive innovation. Join us in exploring Josephine’s journey and the impactful work she spearheads at CJEL, revolutionizing the intersection of imaging and research.

Topics Covered: Dive into the rich tapestry of topics covered in this episode, ranging from the advancements in imaging technology to the intricate details of specific research projects. Unearth the fascinating applications of 2D, 3D, and 4D imaging in diverse fields, including the analysis of bone strength, muscle structure, and even the biodegradation of plastic materials. Gain insights into the challenges and triumphs of projects, such as the time-lapse imaging of chemical reactions and the development of customized software for efficient data analysis. Explore the intersection of artificial intelligence and imaging, paving the way for a future where technology complements human expertise seamlessly.

Our Guest: Josephine Jiyoun Arns – Leading the Way in Imaging Technology

Episode Transcript

Josephine Jiyoun Arns: People are not quite aware about what can be done with imaging analysis. You think that imaging alone is just answer, just perfect to a right answer. But the fact is that really important that how to prepare sample, how to read and controlling taking image itself, but also how you analyze it there.

Jennifer Callahan: Welcome to the Skeleton Crew. I’m your host, Jen Callahan, a technologist with ten-plus years of experience. In each episode, we will explore the fast-paced, ever-changing and suburbs completely crazy field of radiology. We will speak to technologists from all different modalities about their careers and education, the educators and leaders who are shaping the field today, and the business executives whose innovations are paving the future of radiology. This episode is brought to you by X-raytechnicianschools.com. If you’re considering a career in X-ray, visit X-raytechnicianschools.com to explore schools and to get honest information on career paths, salaries, and degree options. Hey, everybody. Welcome back to another episode of The Skeleton Crew. I’m your host, Jen Callahan. I have with me Josephine. She’s joining me from Australia, and she’s coming in with a different perspective of using X-ray and CAT scan. And she’s working for a company. She’s a managing director of CJEL Digital Imaging. So Josephine, thank you so much for being here with me tonight or today, I should say, because it’s the daytime there.

Josephine Jiyoun Arns: Yeah. It’s quite interesting to see you got a long sleeve. I’ve got a short summer here.

Jennifer Callahan: Let’s jump into this. So she’s going to bring a perspective to us that we’re not really looking at radiation from a perspective in health care mainly what her research does can be applied into health care. But their research company is using it in other research capabilities. So Josephine, if you could maybe give a broad overview of what CJEL is and what your company does?

Josephine Jiyoun Arns: Yeah, we work with two different folds. One part is that extensively relates with the research so that rather than just taking an image, we’re looking at perspective of uncertainty, how we make it a more accurate prediction out of image itself. Mathematically, we estimate the clarity of what clients want to know further than just image visualization. And then the second part we does that with young people. We help, young people when they graduate university, when they got uncertain and unclear about where their career will develop, we let them put it into the research project together with us. We train them. We work this research project to help them to develop their career, and we actually had a couple of successful PhDs along the line. Yeah.

Jennifer Callahan: All right. So really two-fold joint venture going on with CJEL. I was on your website earlier and I like that CJEL stands for communicate, join, educate and Level Up. Am I right?

Josephine Jiyoun Arns: Yeah. That’s right. That’s so many. Exactly. We found that people open when you the benefit of working with the imaging industry. You visualize, you get a lot of insight, inspiration out of that and you can imagine. These are all commercial benefits. But more than that, to see what is going to happen without breaking samples is fantastic. And 3D, 4D imaging and seeing the changes in the material from there, like without any intervention. It’s just great. So people quite happy with that. We also found it’s great for high school students to understand about science insight. So I don’t know whether you saw the image with Tim Tam, one of the chocolate Australia famous with and inside the Tim Tam, we opened the talks about this Tim Tam to the high school student. When we want to promote the research and science, then we’re saying that like, do you know why Tim Tam and two different Tim Tam taste different? Which one tastes good? How that surface affecting an instructor? So we just showing these 3D images out of that. And they just are so excited.

Jennifer Callahan: The difference between the two that you said the surface of them. It has something to do with the taste.

Josephine Jiyoun Arns: Yeah. When you go to unlovable surfaces in there because sometimes the merit is outside the inside crunches in there when you see a lot of cracks in there, and people, when you bite it inside of the correct particle loading on the tongue, it makes it taste fantastic. So that’s one of the thing hacking community process. But then we have enabled some like biotech process picture is the product people now looking for the environmental concern looking at promote the soil quality. So when you put it back on the ground, not only helping the carbon, capturing it, also helps the pH and the temperature, also the humidity inside of soil and improves the quality. But the manufacturers of biochar, they like to know how good all the time. So rather than just taking the image, we do time lapse imaging doing that while continually dropping the water. So we had to set up special way from the acquisition and analysis. And then we have another side. People want to know how much gas, how much oil interacts inside the barrel. Then we take the rock image and we do dynamic imaging, but also we just mathematical calculate so that they know that what kinds of contamination can happen, what kinds of recovery can happen without actual experiment. They like one of the success story. We so happy with it. It’s just like really, really well worked so much harder to repeat it again, again to get high-quality images. And out of that we were suddenly able to. So prediction without so much hard, intensive days and days work from the users.

Jennifer Callahan: So the time-lapse images that you’re doing, what’s the space between the imaging?

Josephine Jiyoun Arns: Oh that depends on. So the bilateral ones they are the reactions we first prior analysis with. We open engaging with the biologists in the cases or the material science engineering people. And then we identify how the reaction is happening open. The reaction is faster within 24 hours after it is slow. So we got from the prior like every 15 minutes. And then we then slow down to two hour, three hours, six hours, 12 hours and 24 hours, 36 hours, something like that. But then we had another one like cement hydration. That was, I can’t say the company name, but there was a big company wants to know how the hydration happening and we wanted to know really mechanics out of that. So then we just had to like recreate a terabyte using the supercomputer facility. We just had to look very fast imaging every 15 minutes or it’s just so difficult. It was really challenging because three staff together worked overnight to make sure we can able to. We were so excited about this. We haven’t published it yet. We just preparing that. But it will come out. Yeah.

Jennifer Callahan: So what are you using for this imaging then? You said 2D, 3D and 4D imaging. What type of machine are you using? Would it be like a CAT scan machine that you might possibly find in a hospital, or like an outpatient health care setting, or has it been modified? Yeah, we.

Josephine Jiyoun Arns: Doing modified ones. We do bench-based one. We did actually involve our group in closely working with a new and unstable collaboration that had the spin up the company in previously, and they were very successful in the resources company area to spin up the company and develop the software. Now CGI works particularly for go beyond that to work. We have initial work also done with the medical work area, but in different steps. We work. For example, we looked at the bone. How do we dance of the bone is important in a person. If you implement part of the bone to another structure, where is the rice bowl? Those kinds of spatial analysis. We have done that too. And muscle is another thing we are interested about. Because if you are looking at muscle structure carefully, it will also look at the strengths of the bone together. It will be better for the patient as well. Doctor, to understand what this patient needs to be done to be truly able to change their life.

Jennifer Callahan: For sure, just to give a little insight of what she’s talking about. Bone strength has a lot to do with muscle strength.

Josephine Jiyoun Arns: Exactly.

Jennifer Callahan: So if you take this and you relate it to say like a DEXA scan, which is looking at bone density, generally it’s that type of scan is for a woman who is postmenopausal. But it’s not only for women that’s looking at bone strength. So it’s being correlated with that. So you’re looking at a person who might have really strong muscles. And I guess really they should have strong bones.

Josephine Jiyoun Arns: Yeah there’s a lot of resources there. But what do you do is that it’s similar to the one of the publications we’ve done in the fiber. So we’re looking at the direction of how the muscle lay down and more than just the strength. So we’re looking at it. So, our benefit is we can see the 3D structure to do more mathematical calculations. So we can more accurately predict that rather than just visual aspect.

Jennifer Callahan: Okay. And then when you were talking about looking at the bone and then inserting a powder type or something, would that be something possibly? Maybe along the lines of, say, a person who might have had like a kyphoplasty done for their to one of their vertebrae in their spine. It’s not powder bone, but cement is basically being injected into a bone to fix a fracture in the vertebrae. Would that be something along the lines of something that you’re talking about?

Josephine Jiyoun Arns: That’s possible. It’s just that at the moment we are targeting a lot more execution aspects. So because we’ve just found that people are not quite aware about what can be done with imaging analysis, you think that imaging alone is just answer, just perfect to a right answer, but it to fact is that really important that how to prepare sample, how to read and controlling taking image itself, but also how you analyze it that we actually do some samples during two different acquisition to get clarity of material identification. There are a lot of things in there and I find really great to work with a lot of bright students that they see that what can be done more than just normal or just they doing research, it is very complements existing research. For example, we are the case. People understand the rocks, how the fluid moving into there. There’s a typical experiment that people do MRCp from the MRCp. They don’t really understand why the heck is raw, quite different homogeneous rock, but from visually when you assimilate it and then also take the image, we can clearly understand why that is happening. And people see only micro nano. It’s just another challenge because our effect we got to be generalized about bigger sample. But at the same time, we need to understand clearly the nano and microscale as well. So like really combined research to be able to us understand nature and help people.

Jennifer Callahan: It’s interesting that you say that fluid goes into a rock, which is something that I would have never thought I think of a rock. I just think of a solid mass that nothing is going in or coming out. It’s interesting to hear that you say that they can be heterogeneous, that it’s not just one form of matter, that something else can exist inside of it. I honestly would have never known that or even thought of that.

Josephine Jiyoun Arns: That was my PhD that understanding how the fluid moves inside of rocks, how they are interconnected, and extract the geometry connectivity issues. Then you understand the contamination. It goes to, of course application of soil. But it’s more a lot more in a different way because the underground is not visible. You got to be a lot of working on heterogeneity and the homogeneity. Lots of sample is just I don’t know when it’s finishing research, but we can always go progress. This I know. So we know a lot more than what we need. What did know the 25 years ago and you.

Jennifer Callahan: Started your journey in this over 20 years ago, you said did your PhD 25 years ago.

Josephine Jiyoun Arns: Yeah, I started my background originally was chemical engineering, then I did with the resources engineering, then the end of resources engineering. We worked out about the ambiguity we have continuously. So we thought what might be if we develop ourselves explicitly with really high resolution, and then whether we can really come up with a true answer that we have to do this, this experiment in the lab. And it worked.

Jennifer Callahan: So the students that you have come into Kjelle, you said that you have high school students, but do they range from high school all the way up to students who are doing post graduate work?

Josephine Jiyoun Arns: We mainly works the postgraduate and graduate of undergraduate students who are interested about the imaging, who interested about some of our AI, machine learning, deep learning, those directions. So we particularly see who likes to do rather than just doing a like when they’re thinking about PhD, for example, if they come to us, we training them. What kind of other capacity could have, for example, we helping them to understand Unix supercomputing and we helping them to understand that huge data set handling. This is something very everyone needs these days. And then they identify potential supervisor and work together so they can publish before even they can start. It means that they really want to do it so they can really designing. And some of them actually went to the company at the end. They know, but they are good at it.

Jennifer Callahan: So when they come on and they’re coming in to do some research, or they joining a project that’s already in the midst of being researched, or are they coming in with their own idea of most of them?

Josephine Jiyoun Arns: Yes. Some of them have an idea by their potential supervisor. Most of. Them. Actually, we have a given project working with it so that we can give them a choice to see what they’re good at it, and then we can, along the line modify for them. Also, sometimes the company demanding, like if I got clients they want this particular, then I get consent to involving the young people so they can potentially have a staff to work together. So they’re very happy with it.

Jennifer Callahan: Okay. What’s the general length of time of one of your research studies that you do?

Josephine Jiyoun Arns: That the really depends on the research, because some company don’t want to reveal anything and just have to be silently work very, very fast as well, then we always then involving them extra research staff or collaboration so that we make ensure that we’re giving a clear answers. And those can be it can take about two weeks to a year more than a year. But there is a process to if the case was a student who published it, if we involving just conference paper, not a big problem, but if the publication is only if we are allowed to publish, then in that case, then we have to went through normal processing of reviewing and things like that. It can take like few years to admit there, because reviewing and actually everyone to allow we get allowance to publish. Not always easy.

Jennifer Callahan: So just thinking about the equipment that you’re using to do your research, I can only assume, just like the equipment that we use in radiology, that your 2D, 3D, 4D imaging equipment has made quite advancements since you began your journey through research 25 years ago. Is it considerable the change in technology that you’re using from then to now? A lot.

Josephine Jiyoun Arns: The, for example, detector has developed hugely, so we can take a lot faster acquisition. We found that rather than just buying the company existing ones, we if we modify, we can find even upgrading system. It works as well. So that detector speeds a lot faster like clearer got automatic noise options. It happened. And also data processing as well. It’s like before we had to use a lot of supercomputing facilities, but these days we can buy just low cost workstations to processing it. It wasn’t possible in the past. Has it changed or software aspect as well? What I find when we started, everyone seemed like we want new, but we want new kinds of because we were intensively putting our effort to developing fast processing software and parallelization, and we had a lot of researchers involved in working together. So back then was seems like a very new. But suddenly these days I’ve noticed a lot of people doing. It’s like when you see that someone pass the one minute, like those kinds of barriers, then everyone can do it suddenly. Yeah, yeah, it’s a good thing. Yeah.

Jennifer Callahan: Here the processing times are huge. I think across the board for anybody, even to this point. Now that I feel like you expect computers and technology to move so fast, and when it stalls and it stops moving, you’re are why are you not moving faster? Because you become so impatient. Because you’ve just expected that things are just progressing like that. So it’s got.

Josephine Jiyoun Arns: To be faster because, for example, one of the projects with us is that now we are going to heading on to the moving towards the direction to customize software. For example, if they want to be only analyzing bonds in a 2D. Yes, we just put it a lot of input into it that already analyzed and we trained it. We just give them trained images, particularly for their sample. They are there for them. It will be like a factor of 100 times faster just applying and get an answer immediately. That’s what we are going to. Moving on. We have already done it. Some case studies and we do more. It will get more cost-effective for everyone.

Jennifer Callahan: So talking about the advancements and the lapse of time and having to do things quicker, what other challenges do you face within doing research for these different companies that are asking you to do the image analysis?

Josephine Jiyoun Arns: The main difficulty is really is fundings, because we got to be somehow fairly pay this even though students, they graduated because we have a certain time, the trainee after the teacher, they have to pay and we also have to pay ourselves as well. And then we got to also satisfy our clients. So the funding is a big issue. But if the client is happy with the initial research and they put on it, we keep going on that. Yeah, people.

Jennifer Callahan: And then obviously opportunistic for all parties involved, everybody’s learning something new. And then the client exactly is getting the answer of what they’re looking for. Hopefully it’s a positive answer of what they’re looking for.

Josephine Jiyoun Arns: It’s more than just a simple because they would get a staff potential staff trained, people working.

Speaker3: Together so.

Jennifer Callahan: Well, we were talking about I. Have you been in the midst of any potential or current research studies involving artificial intelligence related to radiology or not even related to radiology?

Josephine Jiyoun Arns: It’s not always easy to, from the visually of radiography image itself to identify all the materials, but if we train them certain aspects of future it will able to detect automatically. So it’s always more information. Allow us to get more close to the right answer and remove the uncertainty. And it’s very interesting to see where the challenging goes on. But I’m actually thinking that people worry about their job losing more and more. It’s true answer that it is true that people losing their jobs, some of people, but some of people will gain the opportunity because of this. Say that. Imagine that rather than a normal just having a radiologist degree and work with that. If you have additional research capability, you would be able to be keep going on with your career further developing always. That’s why I found that our company will be even better for later on cases and once we can involving more students to testing our software to more getting clearer, we can then part of software and commercialize from the even normal clinics. That’s what we hope. But rather than we like to really focus on what we are good at it right now, and then we do step by step rather than just hushing it.

Jennifer Callahan: So it’s really interesting how you think of imaging is just related to sometimes radiology or just maybe in my mind I’m thinking that because that’s just really where my head is the majority of the time. So it’s interesting to hear how imaging analysis can be used in other senses to do research of what you’re talking about. What does CJEl have in the works, and the exciting new research that the company is doing?

Josephine Jiyoun Arns: Yeah, the current image that we are finalizing it, we are just processing our responding the radio combine, one of the nature journals we are going to getting into was biodegradation of plastic materials. So they got more natural plastic which can degrade, which can be degraded faster. We have analyzed that how the 3D image analyzing the process of degradations. And we just fascinated can see it because often the researchers can see with a high resolution only 2D and they open the bracket when you move out. Then they can totally see how the degradation happening in nature while we just took it as it is 3D. So with an initial study finished, we just about to publish images. So excited! I’m happy to share that images with you. And then the future would like to more and more working that all the existing image analysis we have done that we make it optimized to make it clarify so that everyone can use more easily and do further enhanced research. So we can more customize in their own equipment because they have their own example. You can provide us, we can work continually with the young people. And I think the really future that our professional fieldwork with the research because we will not be overtaken by optimization.

Jennifer Callahan: Even how you were talking about the A, I’ve had this conversation with many different people who have come along that they work doing AI and incorporating it into the workflow, but it’s not to take over anyone’s job. It’s just there to be an extra aid, maybe an extra pair of try to help you to help the human eye, basically, or to make the workflow go better. It’s not there to take over your job because honestly, I don’t think computers can take over our jobs at all. They function with the input that we put into it, right? Yeah.

Josephine Jiyoun Arns: Yeah, exactly. Also faster. And also in our area as well because humans created errors, right? We are creative at the same time error. We need repeated work and collections. We can remove all errors through the AI. Yeah there’s a lot of advantages. And if you have enhancing research capacity then means you are not going to be dominated. This is another advantage to our company.

Jennifer Callahan: Well, Josephine, thank you so much.

Jennifer Callahan: For taking the time over there in the day. And then in the evening here, um, at our conversation and the different ways that image analysis can go. So thank you again, everybody. This is Josephine joining me from Australia from Cgil. I’m doing image analysis on many different things, not just on bodies. And this is Skeleton Crew. So make sure you check us out for past and future episodes. And make sure you check out this one on Spotify, Apple Podcast, and then also to on YouTube. So all right guys this is Jen and Josephine. Thank you.

Josephine Jiyoun Arns: You’re welcome. It’s great talking with you.

Jennifer Callahan: You’ve been listening to the Skeleton Crew, brought to you by X-raytech.Org, the Rad Tech career resource. Join us on the next episode to explore the present and the future of the Rad Tech career and the field of radiology.