Daniel K. Sodickson, "The Future of Seeing: How Imaging is Changing the World" (Columbia UP, 2025)

Summary

---

Episode Overview

Podcast: New Books Network
Host: Gregory McNiff
Guest: Daniel K. Sodickson, author of The Future of Seeing: How Imaging is Changing the World (Columbia University Press, 2025)
Date: October 3, 2025

This episode explores Daniel K. Sodickson’s ambitious new book, which unpacks the science, history, and profound implications of imaging—how we see, how we’ve built machines to see beyond human limits, and how imaging is poised to transform healthcare, science, and society in the age of AI. Covering foundational biology, pivotal inventions like X-rays and MRI, advances in astronomy and photonics, and the promise (and perils) of AI-driven imaging, Sodickson presents an accessible but profound narrative about how imaging changes our world and, potentially, ourselves.

---

Main Themes and Purpose

• How imaging extends natural human senses and underpins modern science and medicine.
• The crucial history and breakthroughs in imaging—from eyes to X-rays, MRIs, and beyond.
• The convergence between biological vision and technological imaging.
• The transformative role of AI, information theory, and the future democratization (“everywhere scanners”) of imaging.
• Reflections on the ethical and societal dilemmas posed by ever more powerful imaging technology.

---

Key Discussion Points & Insights

1. Why Write The Future of Seeing? [02:48–04:03]

• Sodickson aims to “give imaging back to everyone,” highlighting how foundational “seeing” is across daily life, medicine, and science. The field’s complexity, he argues, unfairly places imaging “in the domain of specialized experts.”
• Imaging illuminates our lives, “but it’s seldom under the spotlight itself” [03:20].
• The timing is crucial in the AI era: “Imaging has shaped science and healthcare… it’s going to do so even more in an even more transformative way in the era of AI.” [03:47]

2. Defining Imaging [04:03–04:42]

• “I like to think of images as spatially organized information… a map… of what is where in the world around us or within us.” —Daniel Sodickson [04:26]

3. Evolution of Seeing and Human Biology [05:43–10:38]

• The journey from simple sensing to the development of “eyes” as a competitive biological advantage (Cambrian explosion).
• The split between “compound eyes and camera eyes” as divergent evolutionary paths, with humans inheriting the latter.
• The retina’s rods, cones, and “opsins” that convert light into brain signals.
• Stereopsis—seeing in stereo (two eyes, two vantage points) is crucial for perceiving depth and became a recurring design principle in man-made imaging:

  “If you can see something from two different perspectives, you can tell a lot more about it than if you just see it from one.” [09:23]

4. Why Light? [10:57–11:57]

• Light’s unmatched speed, ubiquity, and the richness of the information it conveys make it “the ideal evolutionary means for conveying information.”
• “In some ways, light is the ideal spy.” [11:52]

5. Man-Made Imaging: X-Rays and Tomography [11:57–19:49]

• The serendipitous discovery of X-rays by Wilhelm Roentgen, which allowed the first glimpses inside the living body—a truly “revelatory” moment in science and society.
• The global explosion of “X-ray fever”: “Within a year… there was an amazing outpouring of interest. Over a thousand scientific papers… 50 books.” [15:32]
• The advent of tomography: taking many “projections” from different angles to build 3D images (CT, MRI, PET, ultrasound).

  “Don’t just look at something one way, look at it in lots of different ways—ended up being a revolutionary concept…” [10:04]

6. MRI: Physics, Projections, and Personalities [19:49–28:32]

• How MRI makes us “living, breathing radio transmitters” by aligning atomic nuclei (hydrogen protons) in a magnetic field and detecting their radio signals.
• Recognition of Nobel winners, and the “bitter, personal… all out fights” that sometimes follow [20:20], revealing the human drama of scientific discovery.
• Technical clarifications: projection gathering, the mathematics of Radon transforms, and the development of “spin warp imaging.”
• The massive clinical impact: “If you can see it, then you can understand it and ideally you can cure it.” [27:16]

7. Imaging Quality and the Tyranny of Perfection [28:22–31:42; 55:05–57:16]

• Optimization for “signal-to-noise ratio”—the difference between clarity and randomness in an image.
• The tension between maximizing image quality vs. prioritizing actionable information emerges as a central theme:

  “If we keep focusing on getting the very best quality in each pixel… we’re going to basically end up with exactly the same type of devices we’ve been optimizing for years. I think instead… what information do we need…? How can we use all the other information… to clarify the picture?” [55:39]

• Example: The “clunk” heard in MRI machines is due to shifts in magnetic gradients collecting different projections. [28:32–30:23]

8. Imaging Across Scales: Medicine, Astronomy, and “From Cells to Galaxies” [31:42–36:41]

• Radio astronomy’s use of imaging to understand cosmic phenomena (e.g., the Crab Nebula).
• Striking mathematical kinship: “We were describing… how an MRI machine works or how a radio telescope worked… essentially the exact same equations on the page.” [34:50]

  “It was like discovering a long lost sibling… A radio telescope is basically the same as an MRI machine from the outside in.” [36:07]

9. Surpassing Limits: Photonics, Diffraction, Interferometry [36:41–40:59]

• Overcoming physics’ “diffraction limit” in microscopy by manipulating light or using multiple views.
• Astronomers’ use of “interferometry” and distributed arrays to effectively create Earth-sized telescopes.

10. Speeding Up Imaging: Parallel Imaging & Compressed Sensing [41:29–44:51]

• Sodickson’s own breakthrough: parallel MRI, making imaging dramatically faster by acquiring many projections at once.
• “If I’m just going to compress my image… why am I spending so much time gathering [all the data] in the first place?” [43:18]
• Compressed sensing: acquire only the minimum, random, most “information-rich” samples possible.

11. The Future: Imaging Differently & Reimagining Access [44:51–49:27; 57:16–61:07]

• To democratize imaging globally, we need to move beyond the “bigger, better, more” paradigm and deploy smarter, leaner, AI-supported methodologies.
• “Rather than just saying… big or small, we need to start thinking: how do we coordinate information differently? How do we apply AI to get us what we want… without always focusing on a perfect image right now?” [45:17]
• Vision for the next decade: earlier, more accessible, and more personalized diagnostics through networked, miniaturized “everywhere scanners.”

  “The more we see you, the faster and better we can image you and the better we can predict the trajectory of your health.” [58:49]

Everywhere Scanners: Types 1, 2, and 3 [59:09–61:00]

• Progressive layering of baseline imaging, low-cost future images, and even wearables/embedded sensors to mark personal health change.
• “Maybe just the information from the sensors is enough to tell us if there is a change from your baseline… The more we see you, the better we know what’s normal for you, and the cheaper and cheaper the technology we need in order to detect change.” [60:22]

12. The Rise of AI: Self-Supervised Learning and Sense of Change [47:05–51:22]

• Move beyond classic “labeled” training; self-supervised learning can exploit the massive troves of unlabelled data, similar to how children gain a world model.
• Medical imaging’s future is not just accuracy but tracking relevant “change”: “If you see something that’s a little bit suspicious, you better report it… But a really simple expedient is to look back at a previous image and see, hey, is this finding the same as it was last time? If so, it’s your normal.” [50:53]

13. Learning from Bats: Beyond Human Vision [51:22–55:05]

• Drawing inspiration from nature—bats as exquisitely-evolved imagers—urges us to develop more continuous, “on the fly,” adaptive imaging.
• New MR techniques (e.g., fingerprinting, multitasking) move toward dynamic, real-time, “streaming” acquisition.

14. Societal and Ethical Crossroads: Utopia, Dystopia, and Evolution [61:10–65:09]

• Imaging technology offers both utopian (early disease detection, expanded senses) and dystopian (surveillance, privacy erosion) futures.
• “Every time our species has expanded its vision, we’ve also expanded our minds… but we need to be very vigilant and not just give away our privacy without thought…” [61:51]
• The next imaging revolution may look something like a new stage of evolution: integrating artificial sensors and neural interfaces directly into human experience.

15. Human Stories and Gratitude [65:09–67:17]

• Personal reflections: collaboration, recognition, and highlighting a diversity of scientists (including a shoutout to his brother, emergency radiologist Aaron Sodickson [65:35]).
• Book’s acknowledgement of the importance of community, mentors, and family.

---

Notable Quotes & Memorable Moments

• “Imaging illuminates our lives… but it’s seldom under the spotlight itself.” —Daniel Sodickson [03:20]
• “If you can see it, then you can understand it and ideally you can cure it.” —Daniel Sodickson [27:16]
• “Inner space and outer space are way more connected than we might realize… essentially the exact same equations.” —Daniel Sodickson [34:58]
• “If we keep focusing on getting the very best quality in each pixel… we’re going to basically end up with exactly the same type of devices we’ve been optimizing for years.” —Daniel Sodickson [55:39]
• “The more we see you, the better we know what’s normal for you, and the cheaper and cheaper the technology we need in order to detect change.” —Daniel Sodickson [60:22]
• “Every time our species has expanded its vision, we’ve also expanded our minds.” —Daniel Sodickson [61:32]

---

Key Timestamps

| Timestamp | Segment / Topic | |-------------|---------------------------------------------------| | 02:48–04:03 | Why the book? Societal impact of imaging | | 05:43–07:39 | Evolution of seeing & human vision | | 09:15–10:38 | Stereopsis & multiple perspectives | | 11:57–16:54 | X-ray discovery and cultural impact | | 17:20–24:28 | Tomography, MRI fundamentals, innovations | | 28:22–30:23 | MRI noises & sounds | | 31:42–36:41 | Imaging in astronomy, “cells to galaxies” insight | | 36:41–40:59 | Photonics, diffraction, interferometry | | 41:29–44:51 | Parallel imaging & compressed sensing | | 47:05–49:27 | Self-supervised AI learning & implications | | 51:22–55:05 | Bats, “imaging on the fly,” inspiration for MR | | 55:05–57:16 | Tyranny of quality vs. actionable information | | 57:26–61:07 | Next 5-10 years: everywhere scanners, equity | | 61:10–63:30 | Utopia/dystopia—the ethics of powerful imaging | | 63:30–65:09 | Imaging as evolution; “new senses” | | 65:09–66:46 | Personal stories; the collaboration of imaging |

---

Tone and Language

Sodickson’s language is lucid, wide-ranging, humble, and sometimes playful—he delights in “space age names like spin warp imaging” [25:33] and marvels at the apparent “magic” of MRI even as an expert.

There’s a deep sense of optimism and wonder, balanced by sober attention to privacy, equity, and societal impact. The episode is highly accessible, inviting lay listeners to appreciate science’s beauty and social implications.

---

Final Reflection

Daniel Sodickson’s appearance on the New Books Network offers an engaging journey through how imaging transforms not just medicine but our very understanding of the world—from the inner workings of our bodies to the far reaches of space, and, potentially, to new senses and ethical frontiers. With clarity, authority, and warmth, he makes a compelling case for why the “future of seeing” is a story everyone should know.

Transcript

[A]

When did making plans get this complicated? It's time to streamline with WhatsApp, the secure messaging app that brings the whole group together. Use polls to settle dinner plans, send event invites and pin messages so no one forgets mom 60th and never miss a meme or milestone. All protected with end to end encryption. It's time for WhatsApp message privately with everyone. Learn [e-mail] this episode is brought to you by Indeed. When your computer breaks, you don't wait for it to magically start working again. You fix the problem. So why wait to hire the people your company desperately needs? Use Indeed's sponsored jobs to hire top talent fast and even better, you only pay for results. There's no need to wait. Speed up your hiring with a $75 sponsored job credit@ Indeed.com podcast. Terms and conditions apply. The holidays have arrived at the Home Depot and we're here to help bring the excitement with decor for every part of your home. Check out our wide assortment of easy to assemble pre lit trees so you can spend less time setting up and more time celebrating. And bring your holiday spirit outdoors with unique decor like one of our Santa inflatables. Whatever your style, find the right pieces at the right prices this holiday season at the Home Depot.

[B]

Welcome to the New Books Network welcome to the New Books Network. I'm your host Gregory McNiff and I'm excited to be joined by Daniel Sodigson, the author of the Future of How Imaging is Changing Our World. The book was published by Columbia University Press in the United States in October of 2025. Daniel Sodigson is a physicist in medicine who has devoted his career to developing new ways of seeing. He is Chief of Innovation and Radiology at NYU Grossman School of Medicine, past President and Gold Medalist of the International Society for Magnetic Resonance in Medicine and a fellow of the U.S. national Academy of Inventors. I selected the Future Scene because of attempts and succeeds to explain how modern day imaging works in a conceptual sense as well as paints a very thoughtful map of where we're heading. Literally. The book is eye opening. I wish I could get it into the hands of every college student to develop a better appreciation for this mysterious field that has such a large and beneficial impact on our daily life. I should also note it has a wonderful color insert for the diagrams that Dan and the illustrations Dan discusses throughout the book. Hello Dan, thank you for joining me today to discuss your book.

[A]

Thank you Greg. It's a real pleasure to be talking with you today.

[B]

Dan, why did you write the Future of Seeing? And who is the target reader?

[A]

Well, Greg, we really are all creatures of imaging. We use our eyes and our other senses to navigate the world. We have these powerful phones we carry around in our pockets, and we consume countless hours of video as we go about our daily lives. But somehow, in the modern world, imaging has become the domain of specialized experts who build and use these remarkable imaging devices like radio telescopes and MRI machines. So I guess I really wanted to give imaging back to everyone. Imaging illuminates our lives, right? We use it to illustrate everything, but it's seldom under the spotlight itself. And so I thought it was about time that people heard the remarkable story of how we first came to see and how we learned to augment our natural vision in all these extraordinary ways. And then also there's the question of impact. And imaging has shaped science and healthcare and daily life for centuries, and it's going to do so even more in an even more transformative way in the era of AI. So I thought it was important people had a chance to take stock of how modern imaging is really changing the way we see.

[B]

Yeah, you hit on some themes, obviously, that run throughout the book. Other themes include the relationship between our natural biology and development of imaging, and, as you said, how imaging influences how we see ourselves in our environments. But before we unpack those themes, could you just define what is imaging?

[A]

Yeah, imaging and images are so commonplace in our life that we really don't often stop to think about what they actually are. I like to think of images as spatially organized information. So a. A map, if you will, of what is where in the world around us or within us.

[B]

Perfect. And, Dan, in the beginning, when you're defining this field, you use a few terms, including field of view and spatial resolution, which I think requires discrimination. Could you talk about why these concepts are important for imaging?

[A]

Well, sure. I mean, you know, we all interact with images just. Just as. As kind of pictures in our daily life. If you're an imaging professional like I am, you learn to characterize different features of images. So one of the main features of an image is what does it capture? How much of the world, how big is the area that you can see? That's the field of view. And then the spatial resolution is how fine detail can you discriminate. So can you see two different points in the image that are an inch away from each other, A millimeter away from each other, a fraction of a millimeter away from each other? That's the spatial resolution.

[B]

You open the book Discussing how we human beings evolve to see. Could you talk a little bit about that process?

[A]

Absolutely. And it's a fascinating process because if you think about it, seeing and sensing are absolutely fundamental to life. So initially there were just kind of single celled organisms, right, floating around in the early oceans. But at some point, these organisms learned how to sense their surroundings, how to point themselves towards nutrients they needed, or how to get away from other creatures that thought they looked like food. And I think that's the origin of sensing the world. Then in the Cambrian explosion, which was this remarkable kind of explosion of biodiversity in the early oceans, creatures developed a mechanism of essentially telling here from there using light. In other words, eyes. And eyes were such a remarkable competitive advantage that some people actually credit them with fueling the Cambrian explosion in the first place. And before you knew it, all of these different creatures had progressively more complex eyes. And, you know, then it was off to the races then, basically, you know, new variants on ways of seeing. A couple of different fundamental design principles. The compound eye versus the camera eye, which is what we came to inherit. And fast forward a little bit and boom, we've got our human camera eyes.

[B]

Perfect. And the way we see, Dan, is I think you talk about opsins, convert light into electrical signal in our brains. Is that how we see images or how we process images?

[A]

Right. Well, so first of all, we need to capture light, right, because light is what we're using to create these visual images. So we have this fascinating apparatus in the eye which includes a lens to focus and collect light. But then that light shines on actually a pretty impressive detector array because we need to detect the light and turn it into a signal that our brains can use. And that detector array, the retina, has a bunch of different types of cells, rods and cones. Inside those cells there are light detectors, and those are called opsons. And interestingly enough, they evolved relatively similarly in lots of different species. So it's clearly a clever trick that nature learned to reuse. And basically what they do is they receive light, they absorb it, and then they undergo a chemical confirmation change, which launches a signaling pathway that basically turns into our perception of light.

[B]

Okay, I want to unpack that answer. That's a great. Yeah, no, I, I feel like I'm getting a, a crash course from a university professor in how we see. And it's excellent. Dan, you've referenced seeing, and obviously in your book you talk about the brain being part of this. A term you use is stereopsis, and I hope I'm pronouncing that Right. But you introduce that in the early part of the book, and then it does seem like it's a recurring theme. And as we get into the technology side, we can talk about that. But could you define that term and why it's important for our ability to see and process?

[A]

Absolutely. Well, stereopsis really means seeing in stereo, seeing from more than one position. And I don't know, Greg, I mean, you know, stop and think about it for a second. Um, have you wondered ever why we have two eyes rather than, you know, one? Why, why aren't we Cyclops is walking around, or for that matter, 500. And in fact, there are species that have thousands of eyes. But I think one of the key ideas that nature discovered is if you can see something from two different perspectives, you can tell a lot more about it than if you just see it from one. And so stereopsis is, you know, according to kind of standard theory, part of what allows us to see depth. So, you know, we look at things from the vantage point of our two spatially separated eyes, and we can sort out from various cues how far away something is from us and various other features of it. And that idea, don't just look at something one way, look at it in lots of different ways, ended up being sort of a revolutionary concept for artificially designed imaging. And it was used over and over again in countless imaging devices that we humans eventually made.

[B]

Yeah, that's a very nice recurring theme in your book, that relationship between the biology and the man made imaging technologies. I absolutely want to get into that before we move on to the, the evolutionary discussion. You say, quote, light is the ideal evolutionary means for conveying information. Could you, could you expand on that?

[A]

Yeah, yeah, yeah. Well, there are a number of reasons for that. First of all, light is fast. And you know, you can say, well, how fast is fast? Light is actually the fastest thing in the universe. We know this from Einstein. Nothing can travel faster than the speed of light. So in other words, the information light has, it reports to us really quickly. Second of all, light is everywhere in our little corner of the universe. We happen to occupy this nice cozy region of a solar system right near a big old sun. And so we have lots of sources of light bouncing off things everywhere. And then third, light interacts with matter in very informative ways. So it bounces off surfaces of things, it's emitted from hot things, things. It bends through transparent materials like glass, which allows us to make lenses. And it also has color, different frequencies, which says a lot about the properties of objects. So in some Ways light is the ideal spy.

[B]

Excellent. I want to move to sort of the man made imaging machines. And you provide a nice history of the telescope and its implications for the Copernican revolution as well as the microscope. But where the book, I think really starts to resonate or really clicks is the X ray machine. Could you talk about the background for the X ray machine? Who was involved and what was so unique about it?

[A]

Absolutely. And the discovery of X rays is one of the sort of canonical stories of singular, you know, discovery in the history of science. I mean, many discoveries take lots of different people working together over time. This was an accidental discovery by one guy in his lab, which within a year had revolutionized the world. So the story is that Wilhelm Conrad Roentgen, who was a physicist working in a lab in Germany, was playing around with cathode ray tube, something that sort of spits out electrons. It was something physicists were fascinated by at the time. And he also was a photographer. He photographed his experiments, but he also did it for fun on vacations and so on. And he noticed one time when he had, when, when he was shining this cathode ray on some things that there was a screen, you know, that he had up a sort of glowing screen that had started glowing even though there was nothing he could see shining on it. And rather than just passing it off as some accident, he, being a rigorous scientist, decided he was going to test under what conditions he could get that glow. He worked on it for, you know, some feverish time, you know, barely leaving the lab, and eventually realized that this was reproducible and he could get that screen to glow whenever he shone the tube, pointed the tube in the right direction. He also found that he could capture that glow on film like you used in cameras. And he, he looked at what could block that glow, and it turns out that it went right through many things. And it took something very dense like lead to block it. And so then he wrote this discovery up and he called these mystery rays that produced this effect X rays. Like, you know, X marks the spot or X the unknown. He actually, by the way, then, you know, grabbed his, his wife, Annaberta Ludwig, and asked her to put her hand in front of the tube. And he captured the first X ray image ever of the inside of a hand. He then wrote it up and sent it around to colleagues. And that picture of a hand became one of the most famous pictures ever in the world.

[B]

Yeah, that's absolutely. That was a fascinating read. I think she has her wedding ring on it and talked About I need to go back and talk about seeing death or something. When she first saw the paper, I.

[A]

Mean, imagine these were people who had never seen bones in a living body. You know, I mean, the inside of the body was, except for some doctors who did dissections and, you know, and, and so on. You know, this was something that nobody had ever seen before. And all of a sudden she's staring at the bones in her hand. To her it was like seeing her own skeleton. But it turned out to everybody else it was equally revelatory. And in, in just one year after Roentgen made this discovery, there was an amazing outpouring of interest. Over a thousand scientific papers. Back before there were all that many scientists, something like 50 books. It's something that, you know, other authors have called X ray fever that kind of took over the world.

[B]

Yeah, I think was it Edison, who basically, after reading the paper, had an X ray machine up and running within a few days or something like that. It really, I mean, like you said, it does seem to have dramatically changed the world. One of those, you know, equivalent to Einstein space time insights, just revolutionary, a hundred percent.

[A]

It was one of these scientific discoveries that people could grasp instantly, in part because, you know, what you see is what you get. Right. I mean, this produced actual pictures and the mechanisms to do it weren't even all that complicated. It was something that had been basically hiding in plain sight. And what it did is it revealed to everybody what was previously invisible and people just couldn't get enough.

[B]

Yeah, no, it really, it's a fascinating chapter in your book. And I want to move on because you cover four sort of technologies that I think the modern person is familiar with, given their own, you know, experience and our healthcare system, X rays. And then you move on to mris. And I actually found the concept behind mris amazing. I'll let you describe it, but how it works is just, it's just fascinating.

[A]

Absolutely. Well, if you don't mind, I'm going to take one quick step back to get from the X ray to something that we call tomography, which is how you make slices of things. So X rays were amazing. They could see through previously opaque things, but they had this one sort of funny feature, which is you could see all kinds of detail, say left to right, but along the direction the X ray was traveling, you couldn't tell what was where at all because the X rays just blasted their way right through and all you saw was the net effect. So that's something we call a projection, basically an X ray is you squashed along the direction the X rays are traveling. So it takes your 3D body and you know, compresses it into two dimensions, which is fine. And you know, for doctors, for example, who wanted to find where maybe a stray bullet was that was lodged in a patient, it was really useful. But there was a one whole direction where they couldn't tell where it was. And so what people really wanted was not just to see the projection, they wanted to see a whole slice like where exactly everything was in 3D. And it turns out in order to do that, there was a trick that as we were discussing nature had figured out long ago, which is take many different views. So take your X ray machine and don't just point it one way and squash the body in one direction. Point it from lots of different angles. And what you got was these squashed projections from many different angles. And it turns out that some brilliant people around the 1960s to 1970s figured out how to take lots of different projections and work out mathematically what structures inside must have produced them. This allowed them to basically make slices through the body, any direction you like, without ever making a single cut. The key tomographic modalities that came out of that were cat scanning, CT, which uses X rays in lots of different angles, MRI, which uses magnets to make projections, PET, which uses antimatter to make those projections, believe it or not, and ultrasound, which uses high frequency sound. So that's the kind of, that's the kind of context for mri.

[B]

Yeah, that is truly amazing. And you talked about those breakthroughs, particularly on the mathematics side. Just as a side note, I think there the Nobel was awarded 79 to Alan Cormack and Godfrey Hounsfield. Does seem like some of these breakthroughs involved a little tension or a little personal back and forth. Is that. Was that a one off or. I noticed that throughout some of these discoveries, a little back and forth then and even in later decades. How do you.

[A]

Decide? Yeah, what you describe as a little back and forth was actually bitter personal ad hominem controversies and, you know, all out fights, sometimes beginning, interestingly enough, with the invention of X rays. So the first Nobel prize in physics went to Roentgen for his discovery, because it was so seminal. But there was a guy named Philip Thenart who had lent Roentgen one of these tubes that he was using and who argued that he had actually seen some of these same types of shadow pictures or glowing phosphorescent screens. And he believed that he deserved credit. And so when Roentgen got the Prize. Leonard was furious and told anyone who would listen to that, you know, he should have gotten it. And then, interestingly enough, Leonard eventually did get a Nobel prize of his own, and he spent the majority of his Nobel speech complaining about how he should have gotten the X ray prize. So basically, I think the lesson here is scientists are people too. They work together. They, they, they create knowledge over generations, and boy, do they want credit when something is really influential.

[B]

I know, totally. And I should say we'll get to this, but you talk about a number of your colleagues you work with, and it seems like a very collaborative environment.

[A]

Absolutely, absolutely. Science advances collaboratively and there are personality conflicts. And, you know, the house is never as clean as you might imagine if you don't live in it.

[B]

Totally, totally. I want to just drill down specifically on the MRI technology because there was a study later on and I'll talk about that. But Dan, help me out here. You talk about atomic nuclei like those in hydrogen, which make up as a component in water, is placed in a strong magnetic field and exposed to radio waves of a particular frequency, such that we human beings sitting in the mri, correct me here, become a living, breathing radio transmitter. Is that how the MRI works?

[A]

Some of this stuff, even to me, after more than a quarter of a century as an imager, some of this stuff still sounds like magic. But yes, we are living, breathing radio transmitters. So, so here's how it works. Inside the nuclei, inside the atoms, inside the water, in you and me, there are these little sort of magnets, like compass needles. They sort of align themselves in a magnetic field. And normally we don't notice this at all, but if you take these little nuclear magnets inside us and put them in a strong magnetic field, they align a little bit, and then if you bump them with, with radio waves, they start zipping around in the field and creating radio signal. So, you know, very tiny, very faint, not, not anything you'd hear on a classic radio. But if you, if you create the right sorts of detectors, you can get a signal. But now you still need to tell here from there. And this is where, you know, the Nobel Prize winning, you know, innovations came in. Basically, some brilliant people named Paul Lauter and Peter Mansfield figured out that if you change the magnetic field so that it's strong on one side and weak on the other side, so say it's strong, stronger on the left of your body and weaker on the right, then those little nuclear magnets are going to zip around faster, where the magnet is stronger and slower, where it's weaker and you can take the signal that comes out and basically assign all the fast stuff to one side and all the slow stuff to another, and boom, all of a sudden you've got a projection, you know, what's where, and from there on, you know, it works just like ct. You get lots of different projections and create your image.

[B]

And that. One of those individuals you spoke about, Mansfield, I think, was a mathematician and was his contribution the back projection component of this.

[A]

So actually Louderber was a chemist or a physical chemist, and Mansfield was a physicist. And they actually came at it from different perspectives, but they ended up with basically the same technique, the mathematics, interestingly enough, the mathematics for reconstructing images from projections that had been discovered by a guy named Radon back early in the 20th century. He couldn't get a Nobel Prize because you have to be alive for a prize. And he had, he had passed away long before. So the mathematics was actually discovered and rediscovered and rediscovered several times.

[B]

Yeah, fascinating. I want to briefly. You actually allude to this briefly in your book, but this idea of spin warp imaging, it really seems to be a strong interest or a passion of yours. Could you briefly talk about that?

[A]

Yes, yes, yes. Well. And don't you love the space age name, like spin warp imaging? You know, it's like warp speed engines or thing, but, but basically all that is is really a way of sweeping through different projections. So, you know, the details are technical. But remember, when you're generating a tomographic image, what you're doing is you're gathering projections from lots of different angles. Spin warp is a way of kind of rastering through your projections in a, a regular order, almost like you're doing a search grid and rapidly collecting all the information you need, which then can be converted into an image. You know, as, as a, as a magnetic resonance physicist by training. This is the sort of stuff I love, you know, but the details, you know, ultimately don't take away from, from the remarkable magic of creating an image with magnets.

[B]

No, absolutely. And just to bear out the value of that magic, you cite a 2001 survey, I believe, where a number of medical professionals were surveyed on the most important medical innovation in the last 30 years. And CT and MRI came out on top, meeting coronary bypass surgery and angioplasty. So it seems like, you know, that revolution that we talked about at the beginning with the X ray has just continued to transform the industry and really have a profound impact on medical healthcare.

[A]

It really has. And I guess one of the key principles is if you can see it, then you can understand it and ideally you can cure it. And so before we had all of these imaging techniques that could see inside the body, surgeons and other doctors were more or less operating blind. And so not just for the everyday person looking at an X ray image of their hand, but for physicians, medical imaging was a revelation. All of a sudden, instead of just going and, you know, cutting things open and hunting, they could slice through any way they liked without cutting, discover what they needed to discover, and in many cases, not even need to do surgery. There's this whole concept of exploratory surgery, which used to be a lot more common, but if you think about it, not sure anyone really wants a surgeon exploring inside you and, you know, discovering things new. So imaging really had a great impact in the way physicians could steer care and understand disease.

[B]

Yeah, absolutely. Before we move on, I probably should have asked you this. I'm sure it's a top of mind question for anyone on mri, but what is that loud clunk we hear?

[A]

Absolutely no. Everyone wonders, why does the machine have to bang so much? Basically, what's happening? Remember I was saying that in order to get a tomographic image, you need lots of different projections? Well, it turns out every time you change the, the projection you're gathering, say the angle you're gathering it from, you need to change some magnetic fields, Right? I was saying that the magnetic field is strong on one side and weak on the other. So you have to change that field to be pointing in the right direction. Every time you change a field, remember you're sitting inside this very strong magnet. You have to run some currents through wires and that creates a force on those wires. So that force creates kind of clunk. So believe it or not that when you hear that sound in the mri, you know, that kind of thing, that's really a whole bunch of individual clunks from gathering each projection one after the other. So that's what you hear. And so many people have tried to kind of adjust the way we gather projections to make things quieter and so on, but it's sort of a fundamental feature. One last fun story I'll tell you is in my field of mri, we have a conference yearly by the International Society for Magnetic Resonance in Medicine, and for a number of years running, we had this cool sort of extra session called Sounds and Visions in MRI where people actually programmed the MRI machines to play music to give instructions to the patient. So, you know, think of them as million dollar subwoofers. People are playing Beethoven and Bach on the MRI machines themselves.

[B]

Oh, that is awesome. We talked about getting the best quality image here, and you've talked about some key metrics that you want to optimize for. One metric you discuss is what's called the signal to noise ratio as a key measure of sensitivity. Why is that so important?

[A]

Yeah, well, really, it's important in any discipline of science or even beyond science. You can think about signal to noise with an audio analogy. Right. If you're in a loud room with lots of people talking, there's lots of noise there. But let's say you want to listen to a particular speaker. That's the signal. Basically. You need the level of the signal to be high enough above the noise to distinguish that voice from everything else that is not of interest. The same is true in an image. You want to see the signal you care about, let's say, the distribution of water in the body, like you get with mri, over and above all, the sort of random signal you get from other sources. And so signal to noise ratio is just a measure of, if you will, how clear the image is, how much that signal rises above the background of uninteresting randomness.

[B]

Yeah, that was very fascinating how you think about that. So far we've talked about medical imaging, but a related field, obviously is astronomy. And you have a very nice section, and you weave in a lot of the development of astronomy through imaging in your book very well. Can you discuss how radio astronomy specifically led to a better understanding of the Crab Nebula?

[A]

Absolutely, absolutely. And first of all, I think one of the purposes of the book was to show people how connected all these different modalities of imaging are. Our own natural vision, an MRI machine, a radio telescope, a microscope, they're really all kind of kindred techniques for mapping out the world. But when it comes to astronomy in particular, it's a real remarkable story of how people have used essentially any probe that's available to them to create images. So the whole electromagnetic spectrum, from radio waves all the way up to gamma rays, has been used by different types of astronomy. What's really cool is by looking at these different frequencies of light, you can tell different things about a structure like, say, the Crab Nebula, which is actually this vast cloud of gas around a pretty amazing spinning rock, a pulsar, which is, I think, as I recall, around the size of Manhattan that's spinning many times a second. Basically, if you look at the lower energy parts of the electromagnetic spectrum, like radio waves, you can tell about low energy processes, things like the gas around the nebula expanding. But as you go up higher in energy, you can tell these, you know, you can tell about the, the higher energy, things that happen closer to that huge, you know, fast spinning pulsar. So basically you get not just one view, again, same theme. You can get lots of different views of the same thing that tell you about how it behaves in different regimes of space and time.

[B]

You mentioned that there is a conceptual relationship between astronomy and medical imaging in terms of how we see the world. But you actually had a certain, I guess, eureka moment when you were collaborating with a colleague, astrophysicist, Professor Venkata. I think you're both preparing for a conference and I'll let you describe it, but literally you both had the same mathematical equations on your presentations. And I mean, that was amazing. Could you talk about how sort of, I guess, mind blowing that is, that the same formulas for, you know, the macro and the universe and the galaxies are the same ones you're using to look inside or your colleagues are using to, to peer inside the human body?

[A]

Absolutely. No. It turns out inner space and outer space are way more connected than we might, we might realize. And I discovered this, you know, really kind of succinctly and strikingly when I was sitting down with a colleague. Basically there was this conference called From Cells to Galaxies that wanted to bring together imagers from, you know, the very, very small to the very, very large. And interestingly enough, imaging is advanced enough now that those different communities don't always talk. The people who use the electron microscopes and the people who use the radio telescopes are sort of in different worlds. So this was to try to bring them together. I was supposed to talk about medical imaging for astronomers, and Urvashi was supposed to talk about astronomy for medical imagers. And so we wanted to coordinate our two talks and we sat down and went through the slides and a few slides in both of our jaws dropped because basically we were describing the fundamentals of how an MRI machine works or how a radio telescope worked. And as you said, with some small changes in notation, we had essentially the exact same equations on the page and we would say, wait, but what, what does that, you know, mean in yours? Oh, that means the same thing in mine. Oh, what does that mean? It's the same thing in mine. It was like discovering a long lost sibling. And we basically then in the talks were making the point to people that a radio telescope is basically the same as an MRI machine from the outside in.

[B]

Yeah, that was an amazing anecdote you talked about. We earlier talked about the revolution, the X ray. You later on in the book discussed the revolution in photonics. And specifically, and please, I think you're probably going to correct me here. But at the microscopic level, there is what's called the diffraction limit. But several advances were made to bypass this by manipulating light. Now, can you correct me and expand on what exactly?

[A]

No, that did. You're exactly right. Basically, that chapter was concerned with how enterprising scientists just weren't satisfied with the limits that physics appeared to be giving them. And this is an ongoing theme in science as well. Like, as soon as someone specifies a fundamental limit, everybody immediately starts trying to break it. Just like with, you know, a speed limit, you know, a race, anything like that. And so, yes, there are some limits that people have started to discover about optical imaging with microscopes or telescopes. And basically it arose from the fact that light is a wave. It's an electromagnetic wave. And if you send a wave through a narrow opening, you've probably seen this in breakwaters or, you know, around docks, other things like that. The wave starts spreading out. Well, when that happens to light waves, then you lose your spatial resolution because the light starts spreading. And people thought, oh, this is fundamental. We can't get around it. And then over the course of the last century, one after another after another scientist figured out a clever way around it from things like moving your instrument close enough to the object that light doesn't even really behave as a traditional traveling wave anymore. If you move closer than the wavelength in, you can get these types of approaches, you know, to go beyond the diffraction limit, which means you can get down to see, you know, even individual molecules. And one of the ways, interestingly enough, that people figured out, or a number of the ways people figured out, ended up being remarkably similar to tomography. In other words, take a bunch of different views and then weave them together. So a remarkable history of creativity which has led us now to be able to see, you know, for example, the COVID virus. When it came, we were very quickly able to image it and create that, you know, very well known and feared kind of spiky sphere that we see all the time. That was the result of all of these innovations in seeing smaller and smaller and smaller things.

[B]

Got it. And shifting the other. And aside, back to astronomy, was the development of interferometry. Interferometry response to this need to get around the diffraction limit.

[A]

Absolutely. So it turns out that radio astronomy has a particular problem because it uses radio waves which have a very large wavelength, and the diffraction limit is related to the wavelength, the longer, the larger the wavelength. The longer the wavelength, the more your resolution is limited. But basically what astronomers figured out was a way to coordinate the signals from different telescopes, different dishes, in order more or less to do tomography. Again, this is part of the reason that radio astronomy is so similar to mri. And so what this also meant is that unlike with microscopes, for example, telescopes over time got bigger and bigger and bigger. They had bigger and bigger collecting dishes to gather more radio waves, and they started spreading radio dishes across whole deserts or ultimately around the whole world to create an effective radio telescope with the size of the planet itself.

[B]

Yeah, no, that's fascinating how you go through the physics there. And I know you later suggest we might be able to do something with each of our individual mobile phones to create the largest. Yeah, that might be getting into the sci fi area of the book, but definitely thought provoking. We've talked about a number of developments in mri and two big ones seem to be parallel imaging and compression sensing. And you actually played a role in this development. Could you, you talk a little bit about that and maybe the smash versus sense different approaches?

[A]

Sure, sure, sure. Well, you know, I include a number of profiles in, you know, of, of imagers in the book just to try to give people a sense of, of what the enterprise of imaging is really like for everyday people on the ground. And I figured it was only fair if I give a little bit of, of my own background on how I came into imaging, which as it happens, was completely by accident. So I was doing a rotation with a physician, Warren Manning, who did cardiac imaging. He imaged the heart. And, you know, I was just there for a month. I was supposed to do a, you know, a write up of what I found. And I was reading through the literature and it started striking me, well, why can't we image faster? Because, you know, it's considered bad form to stop the heart while you're imaging it. And the heart just keeps on beating, you know, about, you know, once a second. And so it's hard to kind of capture everything fast enough. And what struck me as I was reading through the literature and thinking about it is, well, we're slow because we're gathering one projection at a time, right? Because remember, with tomography you need to gather all of these different views. What I reasoned was, well, why couldn't we gather multiple different views simultaneously? And after a week of scribbling and furiously trying to convince myself, you know, that there was something there, I realized that actually you could, you could gather multiple different Views at the same time, which means that you could gather the same image that much faster. And that was what became parallel mri. Parallel, because we're gathering projections all at once rather than one after the other. And that makes sense. Meant that now we could image two times as fast, three times as fast as we could before.

[B]

Excellent. And then compression sensing.

[A]

Well, so, yeah, so parallel imaging kind of created this speed bump. Actually, not speed bump, speed boost. It was sort of like a turbo factor that then got incorporated onto many scanners. But then about 10 years after that, some very bright people that I talk about in the book realized that there might be another way to speed things up. Because if you think about it, when we gather an image, right, let's say we want to send an image by email, first thing we do is we say, okay, you know, you click send and you get this option, small, medium, or large, right? And if you choose small, it compresses the image with, you know, say, JPEG compression. And people started thinking, well, wait a second, if I'm just going to compress my image, if I'm going to throw away all of this information, why am I spending so much time gathering it in the first place if it in fact ends up being compressible? And it turns out there were these really cool mathematical principles that could be used to pre compress an image. If you gather an image in a certain way with a certain amount of randomness, then it turns out you could effectively reconstruct just the compressed image with a smaller amount of data, and then, you know, basically reconstruct it like we normally do into a full image. So it was a clever way of being very efficient about the projections we gather.

[B]

Yeah, fascinating that. That brings me to a larger question. In the book, you note that the information that the images can convey is limited only by the laws of physics and the boundaries of human ingenuity. And throughout the book, you talk about the importance of bigger is better in some approaches, less is more, as others to drive advancements in the field. But you actually state the key going forward will be to image differently. What do you mean by that?

[A]

Yeah, well, I think I told you how Astronomy, for example, and the same thing happened in MRI had this bigger is better trend in order to get more signal, finer spatial resolution, all of that. But people in the last decade or so have started realizing that if we actually want to make imaging accessible to everybody, if we want to get MRI machines out to people who need them around the globe, then probably less is more is better because these big machines are Extremely expensive and hard to maintain and so on. But you lose image quality as you shrink systems like an MRI machine or a radio telescope, for example. So I argue in the second part of the book, which is about the future, that we really need to start thinking differently. For example, how can we use AI to connect different scan sessions together? So all we need to determine is what's different between one scan session and another. Maybe we don't need as much data, maybe we don't need as fancy a machine. And so I think we need, rather than just saying, I want my machine to be big or small, I think we need to start thinking, how do we coordinate information differently? How do we apply different AI approaches to get us what we want from imaging without always just focusing on getting a perfect image Right. Right now?

[B]

Yeah. You touched on a theme that's driving transformation throughout the entire technology economy and industry. And I found your chapter on AI very nuanced and thought provoking. And you talk about specifically the notion of self supervised learning. Why do you think that's so important?

[A]

Yes, well, for a number of reasons. Basically, what self supervised learning is is learning without sort of labels of ground truth. A classic way to train a neural network is to give it a bunch of examples of something, images of a cat or of cats and images of dogs. And you train it to distinguish between cats and dogs. And to do that, you need to tell it which ones are cats and which ones are dogs so it can correct itself and learn. Self supervised learning is learning without those labels to rely on. And what it does is it maybe blocks out a piece of an image and then trains the network to fill in what's missing. And by if you do that enough, it kind of learns on its own what the relationship of different parts is and learns to distinguish cats and dogs, even though it doesn't know what cats and dogs are. The reason that's so important is there's a lot more unlabeled data out there than there is labeled. It's very laborious to go and label every image with exactly what's in it. The second reason I think it's important and interesting is it's actually the majority of how we learn as children. So, you know, when you're a child, sure your parent is telling you, oh, this is a fork, this is a knife. But most of the time you're just taking in all of this stream of sensory information and your brain is kind of sorting out how it behaves. You're learning object permanence, you're learning laws of physics. You're learning that if you drop something, it falls. You're basically creating in your brain what my colleague Yann Lecun calls a world model. That's what self supervised learning can do. It's essentially getting familiar enough with the way things work that you can discover the, you know, the way the world works.

[B]

What I really liked about your chapter on AI is rather than just saying, hey, it'll be another tool to help radiologists, you conclude that, hey, AI can help us funnel this deluge of data that we're all swamped in down to a manageable stream where we can sense and detect change. Why is that so important? Detecting change?

[A]

Yeah, yeah, yeah. Well, if you think about it, in medicine, what we care most about is change, right? If you've been healthy, you're going about your daily life, that's your normal. What we really want to see is, are you yourself today or not? Is there something worrisome that's brewing inside you? Is there a disease that's developing? But the way we normally image is we just look at what we can see today. And there's this issue. If you think about using imaging for screening, you know, in advance to give early warning for cancer, for example, there's this issue that, well, if you see something that's a little bit suspicious, oh boy, you better report it. And then you can create these false positive results and then somebody has to go for all these tests, which might be unnecessary, but a really simple expedient is to look back at a previous image and see, hey, is this finding the same as it was last time? If so, it's your normal. Why are we worried? And so I think one of the things AI can do in a pretty remarkable way is connect different imaging sessions, connect different information we have about you to determine what's normal, not just for the population, but for you, and then only raise a flag when you have changed from your normal baseline.

[B]

Fascinating. Before we move on, I want to note, you started the AI chapter talking about cats and you end it talking about bats, which incredible. And my humanities chapter heading would be AI From Cats to Bats. But is it, I mean, can you briefly just talk about, you know, bats, they hear better than us, they see better than us, and they're doing it, quote all on the fly, your term. And that's very nice. Are bats just evolutionary, from an imaging perspective, better than us?

[A]

I think, you know, better is always dangerous, you know, because the question is better, better at what? Better at what tasks? But bats are incredibly well evolved for what they need to do. And if you think about what a bat needs to do, and by the way, I've been studying them recently just because, you know, as an imager, I stand in awe of them. Bats need to navigate through a complex world, often at night. They need to discover their prey and track it down quickly. They need to avoid dashing their brains out on a tree trunk or a rock. And they're doing this all predominantly. They do have some, you know, optical visual sense, but they're doing this predominantly with echolocation, basically with sound. And what amazes me is that a bat in a, in our noisy everyday world with all the, the, you know, the sound we've got coming at us, they can send out these effectively sonar pulses and adjust the frequency and adjust the focus and home in on a moth that's flying around in underbrush and manage somehow to swoop in, pick it up and swoop away without getting caught in the underbrush. They do it at speed. They do it literally on the fly because they're flying. And I, I think this highlights a feature of natural vision which we would be really, you know, well served to emulate in artificial vision, which is imaging continuously rather than taking these sort of still frames, as we do in medical imaging, as we do often in astronomy or microscopy, instead going from snapshots to streaming. How do we create imaging devices which can work on the move while a body is doing what it's doing and still give us all the information that we need, like batscape.

[B]

Excellent. And I think you reference, are there technologies like Mr. Multitasking and Mr. Fingerprinting that are looking at, trying to do this sort of capturing real time, I guess, dynamic imaging.

[A]

Absolutely. And listen, you know, video, you know, movie creation has been part of imaging for a long time, but still usually you, you know, try to set up the movie camera and, and keep it fixed in place, maybe on a tripod and you try to control the environment as best you can. I think particularly for tomography, it's a real issue because you have to gather all of these different projections, which means your shutter speed is kind of long. Well, recently in medical imaging there's been this interesting paradigm that's been developing which is more or less just gather your information, a continuous stream and sort it out later. And if there's some motion, okay, we figure it out and we correct it. And so these techniques, a technique I call grasp, you know, Mr. Fingerprinting, Mr. Multitasking, they're basically gathering lots of diverse information. Again, that theme lock diverse information is Better than singular information. And they're gathering in a continuous stream. That has proven to be very useful when you're imaging moving organs, when you're trying to collect a lot of different information about somebody sort of at the same time, and minimize the amount of time they spend inside our big imaging tubes.

[B]

Yeah, that really was another just very, very interesting, very fascinating discussion on that same theme. You address some of the issues that might be holding the industry back, and you label the tyranny of image quality as sort of the number one. What is that?

[A]

Yeah, yeah, yeah. So at some point when you have optimized the technology enough, there's a risk that you become, as a developer, your own worst enemy. And this has been well documented in industry. For example, something called the innovator's dilemma is well known. Sometimes businesses that create the most innovative products then lose out to competitors later on because they have trouble breaking out of the habits they've developed and thinking of whole new markets, whole new ways of doing things. Because they know too much, they've kind of gotten locked into their perspective. Well, our perspective in imaging for a long time has been create the best possible image, take the time you need, be careful, gather your orderly ranks of projections and create a really good, pretty looking snapshot, and then move on to the next snapshot. But for a lot of the reasons we've just been discussing, that isn't necessarily going to get us where we need to go. And so if we keep focusing on getting the very best quality in each pixel of our image, we're going to basically end up with exactly the same type of imaging devices we've been optimizing for years. I think instead, what we need to think about is not so much image quality as information quality. What information do we need about a galaxy, about a person? What is the most efficient, most effective way of getting it? How can we use all the other information we know about that galaxy or that body in order to clarify the picture? So that's how I think we escape the tyranny of image quality.

[B]

Hmm. Yeah. Can you talk a little bit about the advances in imaging you think we'll see over the next five to 10 years and how that's gonna transform the healthcare system?

[A]

Absolutely. I think imaging, as we've discussed, has already been transformative for healthcare, but I think it's gonna be even more transformative in the future by for a couple of reasons. First of all, though, imaging nowadays is used to diagnose disease once you already have symptoms in the future, I predict that it's going to be used to catch disease early while it's still curable, before you even have symptoms. It's all. We're also going to be able to bring imaging to people where they live and work. Right now, if you want a medical image, you need to come into an advanced facility and, you know, sit in this big tube. It turns out about 2/3 of the people in the world have zero access to mri. It's not at all equitable, but I think in the modern era of AI and miniaturized devices and so on, we can get imaging out to people where they live and work. In order to do this without too many false positives, as we were discussing, we're going to need to use AI to incorporate what we know about you and your baseline state of health so that all we need to detect is any worrisome changes. And in fact, we have brand new data coming out of my lab showing that this works. Basically, the more we see you, the faster and the better we can image you and the better we can predict the trajectory of your health.

[B]

That is fascinating, Dan. And one of the concepts you introduced to sort of catch disease early is this idea of everywhere scanners, a type 1, 2 and 3 network. I found that really fascinating, and both from the technological perspective, but also the cost perspective. Can you talk about that?

[A]

Absolutely. So I. I've had this project that has been basically making me lose sleep now for a couple years, and I've become singularly focused like, like Ahab with his whale on the. The everywhere scanner project, hopefully with a better outcome than that. But the idea is, let's say you come in and you get a baseline MRI scan and a baseline set of blood tests that establishes kind of your current state of health. Well, the next time you come in, why do you need to go into this state of the art scanner if all we need to do is detect change? Why can't we build an MRI machine into this chair that I'm sitting in or that you're sitting in? Why can't we build it into a bed at home or, you know, a bench in a cvs? Maybe we can get what we need just from that more rudimentary imaging device. Because we already know your baseline and we have AI to distill it and fill in what's missing. In fact, maybe we don't even need a full imaging device at all. Lots of people are talking about wearable sensors, right? Wearable devices. People have aura rings and Apple watches and they have, on their own, very limited information. They can't see very far into you. They just get a small set of signals. But maybe if we already have a lot of information about you from previous scans, maybe just the information from the sensors is enough to tell us if there is a change from your baseline. So again, the more we see you, the better we know what's normal for you. And the cheaper and cheaper the technology we need in order to detect change. That's the idea of the everywhere scanner.

[B]

Yeah. I hope we see it soon. Dan, that sounds like such a great idea. Hopefully we can implement it.

[A]

We're working on it as you and I speak.

[B]

Fantastic. I look forward to that. In the last chapter, Dan, you present basically a utopia or dystopia vision. We're on the cusp of that. On the one hand, artificial eyes, telepathy, X ray vision. On the other, hacked eyes, hacked brains and sensor dust. Where do you come down on this?

[A]

Yes, yes, yes. Well, I think every powerful tool can be used for good or for ill, right? Nuclear power, nuclear weapons. Imaging, I contend, is an extremely powerful tool. Every time our species has expanded its vision, we've also expanded our minds. We've also been able to do remarkable new things. And I think if you look at the future of AI driven imaging, it can lead to remarkable changes in healthcare. We catch all disease early, while it's curable. We, you know, share. We image our brains as we're, we're interacting and we share our thoughts. But we already know we live an increasingly surveilled life. Right? I mean, how many places can you walk in a modern city without being viewed by one video camera or someone's cell phone? And so I really think the uses of imaging are kind of up to us as individuals and as a society. You asked where I come down. In part because I'm inherently an optimist and an idealist, and in part because I'm a medical imager. I come down on the side of light. I think that these advances are things that are going to open up new capabilities, are going to improve our health, are going to allow us to see each other better. I think medical imaging, for all of the costs and other things, has been a net good for humankind. But I think we need to be very vigilant and not just give away our privacy without thought, not just train imaging devices in places that, you know, should be private, for example. So we need to, we need to be aware of what's coming our way. Which again is part of the reason that I wanted to share this story with people.

[B]

Nice. You end the book by saying the next imaging revolution may look something like a new stage of evolution. You mean by that?

[A]

I think I mean that artificial imaging, which was inspired by our bodies, but separate from it, may just come home to our bodies again. And what I mean by that is if we have all of these remarkable imaging devices and we have. And scientists are starting to develop some very interesting ones, we have brain machine interfaces, whether that be, you know, augmented reality glasses or direct neural implants, then we might just be able to expand our natural sensorium. You know, imagine that you could now take signals from imaging devices and pipe them into normal brain pathways so that you could see what these devices see. Then all of a sudden, we are basically creating artificial senses. And you might say, okay, you know, science fiction, like, how real is that? But I can tell you I have colleagues who are working on modulating thought patterns, on modulating, you know, neuron firing using ultrasound from outside. There are people who are taking signals that are being passed down your wrist to control your hand and repurposing them to control machines. You could imagine sending other signals back the other way. So I'm not saying this is tomorrow or even in five years time, but I think down the line, we might have a whole new set of senses.

[B]

Yeah, that is absolutely amazing. And you know, what your lab is working on today really sounds like magic on one hand, but at some point, Dan, we've had a great discussion about great technologies, great themes, and we've talked about at one point in the book, you introduced a number of key figures in the field that you worked with, you think highly of. So I gotta ask, why did you leave your brother Aaron out of the profiles of the. Of the key figures?

[A]

Well, first of all, I did say that I could brag to you about my brother Aaron, and I'm going to take the opportunity to do that now. So Aaron Soddickson is a remarkable scientist and human. He's a physicist and a physician. He's a leader in emergency radiology. You know, those heroic doctors who are kind of behind the scenes in shows like the Pit. They're the ones taking the images that help the ER docs know what to do and how to save their patients. But he's also been a leading innovator in CAT scanning. I guess he is, of course, in my innermost circle of imagers. We've published together, we talk about the future of imaging a lot. But I wanted to expose people to kind of the larger circle of imagers as well. And so for the profiles in the book, I wanted to be careful to provide a cross section of stories that illustrate the vast scope of imaging. And, you know, therefore, I thought I would tip my hat at my remarkable brother, but give you a range of stories to choose from.

[B]

No? And you do that quite well. That concludes our interview. I will note there's a very nice acknowledgement section at the end of the book in which you thank your parents, family, friends, the taxpayers, and your high school Latin teacher on your life. It's obviously, you know, if we're the sum of everyone we've met, you are very aware of that and appreciative of that. So that was a very nice acknowledgment. The book is the Future of Seeing How Imaging is Changing Our World by Daniel Sodakson. Dan, thank you so much for your time and writing such a thought provoking book.

[A]

Thank you so much. Greg Sa.