John Mazziotta is Professor of neurology, CEO of UCLA Health, and vice chancellor of UCLA health sciences. He was also a founding member of the OHBM. He co-authored the first book on whole-body, cross-sectional anatomy using CT. He’s been involved in the first PET studies in normal subjects and with patients with epilepsy and Huntington’s disease. He was the principal investigator of the ICBM brain atlas, a key tool for brain normalisation. We interviewed him as part of our OHBM Oral History series, to find out about the early days of PET, (f)MRI and the inception of OHBM.
Nils Muhlert (NM): Thank you very much, Professor Mazziotta, for joining us today. I'd like to start by asking you about your background: Why and how did you become interested in neuroimaging?
John Mazziotta (JM): Well, I wanted to be an architect. That didn't work out because I spent a lot of time in Manhattan with architects when I was an undergraduate, and they didn't seem very happy. I like science and went into a lab where I was doing early molecular biology and that was interesting but very isolating. I thought, “Well, I'll go to medical school.” I hated medical school, memorizing bones and things of this sort. Ultimately, I met a neuroscientist in the medical school. The school also had a very active biophysics department and were building the first CT scanner that could image outside of the head. This is now mid-1970s. I got involved in that project and we physically built that machine, soldering wires. We had a functional scanner that worked anywhere in the body.
I decided I would become a neurologist, moved to Los Angeles and UCLA and immediately met the group that had moved from Washington University in St. Louis. They had been involved in the development of PET and all worlds connected, so I got involved in research with PET and then MRI.
NM: Should I ask who the neuroscientist was, that you met during your med school years?
JM: The neuroscientist in my medical school years was a woman named Betty Hamilton. And ironically, Peter Fox and I were in the same medical school class, and there were other neuroimagers in medical school with us. It was an interesting coincidence. The seeds were set there.
NM: And what was it about neuroscience and neuroimaging that really attracted you?
JM: Clinically, it was the approach to the patient, the gathering of the data, the deductive logic of coming to a diagnosis and then having confirmatory tools like imaging that could show you structural and/or functional abnormalities, confirming whether that diagnosis was accurate or not.
NM: And what do you see happening with neuroimaging in the US nowadays?
JM: Obviously there's been a big shift to MR-based strategies rather than PET. But now we're starting to see a resurgence of PET to study patients with neurodegenerative and psychiatric diseases. In the mid-1980s I was confident that psychiatric disease would be completely understood by PET because they were chemical disorders. You could give somebody chemicals and they would become psychotic or delusional or hallucinate. And you could take a patient who had mental health issues and give them medications and they would get better: it was a chemical issue, no structural changes. But after 40 years of scientists, very diligently, pounding away and making ligands and imaging patients, we still have not provided, to my mind, actionable insights in mental illness through imaging.
NM: And do you think there will be breakthroughs over the next 20 to 30 years? Is there anything you'd peg your hopes on?
JM: I’m just as confident now as I was then [laughs]. It's a tough problem and a very expensive problem. When developing a new ligand it might be useful if, as a community, we encouraged key places to be the factories of ligand development. It's so expensive, and like CERN or other high energy physics projects, talented scientists could go there, do the work and then return to their home institutions. Once proven, the recipes for the ligands could be distributed widely.
NM: That's an excellent idea, to really pool the resources and expertise. Going back to your own work, what research or other contributions are you most proud of?
JM: Always a humbling question. My research with PET began with studying normal individuals. We studied the visual system. Our first papers were on visual responses to different types of stimuli, sensory deprivation, auditory stimulation, and a variety of various states in normal subjects. That was very exciting. Every time we did a scan, it was a new day. You never knew what you were going to see. It was an exciting time.
On the clinical side, I was involved in the first PET studies of patients with epilepsy, the first studies of patients during seizures, Huntington's disease and depression. Combining genetics with imaging, in the Huntington study, was a good example of developing probabilistic approaches to individuals who are at risk, then testing them genetically and seeing the outcome and looking at the scan. Those were also exciting times, in our collaborations with what was then the Hammersmith group in London; ultimately with Karl Friston, Richard Frackowiac and the others, Terry Jones in that group, and our colleagues in UCLA with Mike Phelps and Henry Huang, myself and some of the fellows that I had, Roger Woods and Scott Grafton. We carried out a lot of the early work on blood flow measurements with that combined group.
NM: And many neuroimagers will know of the ICBM brain atlas. You had a pivotal role in the development of that. How did it come about?
JM: That was a painful part of history. We were all struggling with how to normalize data, in our own labs, among individuals within a modality and then across modalities, and then ultimately, to pool data from multiple different laboratories. And it was clear that this was an enormous problem, and it was unlikely that one laboratory on its own would solve it.
I invited to Los Angeles, where I was then working with Arthur Toga, a group of the individuals around the country and around the world that were doing this kind of work and were frustrated by the fact that the problem was a difficult one. We all worked together for a couple of days: Peter Fox, Alan Evans, the people I mentioned, some individuals from Europe and Asia. We were all natural born enemies. We're all vying for the same funding dollars, all doing a lot of similar work. There was a lot of posturing and opening remarks. In the end, we emerged from those two days saying the only way to really solve this problem is to do it together. And if we did it together, we'll actually get it done and would emerge with something that makes sense.
Later, we teamed up with other groups, particularly the group from Jülich, who did all the amazing work on the histology. It was a big program, Arthur Toga perfected the sectioning of human heads; that data went to Germany, to Karl Zilles and Katrin Amuints and their talented teams and went on from there. It's still going on today with the BigBrain project and the collaborations between McGill and Jülich. So that was another satisfying addition to the contributions by the group. That group continues to meet, always in Hawaii, always in the first week of November, this year will be the first exception in something like 26 years.
NM: It's a great example of fruitful collaboration, not just between national institutions, but across the globe. And you were also involved in the creation of OHBM. What was that like? And what did you imagine OHBM would be like?
JM: It was less about what it could be, but what it would eliminate for us. We were at a meeting in San Antonio that Peter Fox hosted. I was there, Arthur Toga, Alan Evans, Bernard Mazoyer. We were bemoaning the fact that we were all going to all these meetings every year, the cerebral blood flow and metabolism meeting, this meeting, that meeting, meetings about MRI, meetings about PET, meetings about everything. It didn't make any sense because most of those meetings had nothing to do with what we were interested in, which was trying to map the human brain.
Sitting around in a little conference room, the idea emerged,: "Well, why don't we just have a meeting about what we're interested in and not have to go to all these other meetings?" So everybody said, "Yeah, that'd be great. But it's going to be such a pain to do and who's going to front the money." But people were compelled to do it just because they were so frustrated with the current situation. And when we thought about how much money was wasted sending fellows and students and everybody else to all these other meetings that were low yield, we decided we'd roll the dice. Bernard decided he would really roll the dice and put a deposit on a center in Paris. And the rest is history.
NM: And how did you think it would evolve over the years?
JM: Well, we didn't know if it would work at all! So initially, it was a matter of trying to stabilize the finances to the point where we could at least be confident that if we advertisd the next year, there would be enough funding to get it to happen. The more grandiose envisioned something more like what the reality is now, that there would be books and journals and a subdiscipline of neuroscience that was basically doing this. We also felt that it would be important to have educational components: that a graduate student in psychology might not necessarily be exposed to the physics of the machines that was generating the raw data, or that a mathematician who was doing modeling wouldn't necessarily understand neurophysiology and neuroanatomy. And so once it was clear that we could sustain the meeting, then the next part was to make it something that had value not only as an information exchange but also as an opportunity to provide training to the field.
NM: And thinking of your own involvement with OHBM. What have you found most rewarding about that?
JM: Walking into the rooms with the posters or the lectures and seeing all those people, and the energy of the students and the fellows and people who are seeing their careers and their professional motivation, joined by colleagues who were like minded in the same place. That's very satisfying to see. From five people sitting in a room to thousands of people who are all thinking about these problems and trying to make progress.
NM: Definitely! Are there any experiences you've had attending OHBM that really stand out?
JM: One of the things that we did for a while was Richard Frackowiak and I would summarize the whole meeting at the last session in 30 minutes. So we had to somehow boil down 1000 presentations into 30 minutes. That was always challenging and fun. And I have to say it does focus your attention on the content of the meeting, rather than just cruising around and talking to people and reading a few posters. For the first 10 years, we had an “L & L party” with people from Los Angeles & London labs. We would jointly fund that party and it was one night a week and we'd bring the entire lab to these parties. We had them on boats, nightclubs, bars - they were all over the place. That's evolved into the social events that are on the different nights of the meeting,
NM: So that's been there since the very start?
JM: The first one.
NM: Brilliant. And have you seen any changes in how the meetings run or different angles coming into it that perhaps weren't there at the start?
JM: Well, when things are small, they're easy, everybody's in one room. And as things grow, they become more specialized and subcategorized. Then you have to pick and choose which things you want to attend or not. That's a natural evolution of any process like this. But I appreciate the fact that the named lectures and the other components have been maintained as unifying parts of the meeting. Attendees really look forward to those sessions and everybody's there.
NM: And a final question then. Not an easy one to end with! But, what do you see as the future for neuroimaging?
JM: [long pause] Notice the long pause [laughs].
The future is bright. But the pressure will be there to deliver on the clinical side, truly valid biomarkers from imaging. We don't have any of those yet. If you really look hard, and ask the question: if we have this imaging result, then this is the diagnosis. Those are few and far between, from functional imaging. That will be the question that will be put out there. Can you do that and if not, then why should we fund it? So that's one.
The second issue was the one I mentioned earlier: insights into purely functional and chemical illnesses like mental illness. And that's a big lift, and an important one. And similar one will be neurodegenerative diseases.
Then the most profound and the most interesting question is: how does the brain work? I would envision in the future, that through techniques we don't know about today or in some parts, extrapolations of some of the physics of MRI, that we'll get to the point of actually being able to image neural conduction, synaptic activity. It wouldn't be microscopic, but in large ensembles. With improved temporal resolution, we'll be able to understand the choreography of signaling in the brain. Once that level is achieved and massive data can be managed in four dimensions, then the insights will come more rapidly.
NM: So there's bridging of scales.
JM: Yes both spatially and temporally.
NM: Fantastic. Professor Mazziotta, thank you so much for taking the time to speak to us. It's been a pleasure hearing about your history with OHBM. Thank you very much.
JM: Thank you.