Aoki-sensei recognized my work just as I was starting my career as an assistant professor at Montreal Polytechnique. Together with Prof. Masaaki Hori they invited me to talk about the g-ratio at the Japanese Society for Magnetic Resonance in Medicine in 2015, and it was the first time I felt that my research may have a global impact. It was also the first time I had tried shochu, a smooth Japanese distilled spirit that sealed my bond with Aoki-sensei. Once he found out I liked shochu, he ordered four different kinds (rice, barley, potato, and black sesame) and we tasted each one in three ways: neat, iced and hot. Twelve drinks and a Tokyo earthqake later1, I had met a mentor and a friend for life.

Read more

KANTAROT / КАНТАРОТ

ChatGPT and Galactica are taking scientific papers to their logical conclusion

“... taken to its logical conclusion, every story is sad, because at the end everyone dies…

Read more

3 years ago · 3 likes · 1 comment · Nikola Stikov

In September 2024 we held NeuroLibre Day in Montreal in an attempt to answer this question. What started out as a harebrained idea about going beyond the PDF in 2018, materialized as a cutting-edge platform for hosting executable research objects. Our scope was limited to neuroscience, but over the course of five years we published 20 NeuroLibre reproducible preprints (NRPs), including this one which we dubbed the research article of the future:

KANTAROT / КАНТАРОТ

Crowdsourcing the research article of the future

NOTE: Mathieu Boudreau, Agâh Karakuzu and Pierre Bellec contributed to this post…

Read more

3 years ago · 1 like · Nikola Stikov

Exactly a year ago we found out that our grant that funded NeuroLibre has been renewed. This meant that we could take NeuroLibre to the next level, expanding our scope to include all natural sciences, and making it possible for papers to “talk” to each other. If you are unsure what “papers talking to each other” sounds like, I recommend you read Agâh Karakuzu’s tour de force review where he introduced the concept of woven literature.

Tomorrow (December 4) we will demonstrate what woven literature looks like in the age of AI. We will introduce a new article format (a wired paper) and present four examples that cover a range of disciplines. There will also be a real-time demo of the Evidentron, an LLM tool that makes it possible to remix and update papers published in Evidence. Here is sneak-peak of the new Evidence website:

https://evidencepub.io/

To kick things off, we have assembled a stellar list of speakers including John Ioannidis (Stanford University), Alan Evans (McGill University) and Cesar Hidalgo (Toulouse School of Economics). They will set the stage for Evidence, explaining the need for going beyond conventional research outputs in the age of AI.

The event will take place on the beautiful NYU Abu Dhabi campus, with partners from universities in four countries (Polytechnique/McGill from Canada, Juntendo University from Japan, NYUAD/MBZUAI from the United Arab Emirates and UKIM from Macedonia). Everybody can attend the event virtually by registering at the following link.

Evidence has been seven years in the making, and I have spent the last two years building the Evidence federation in Japan, Macedonia and the UAE. None of this would have been possible without the groundbreaking work of Agâh Karakuzu, who has been the driving force behind Evidence from the beginning. This journey has also been shaped by my ongoing fight with 10-page pdfs, a struggle I have documented in detail on this blog. The battle is not over, yet the momentum is finally shifting in the right direction. Join us as we take the next step toward making scientific publishing reproducible, reusable and ready for the age of AI!

Share

In magnetic resonance, an echo appears when we reverse the dephasing of spins caused by field inhomogeneities. A simple 180° pulse undoes that phase dispersion, and the signal comes back refocused. Small perturbations will result in imperfect echoes but will also reflect the underlying structure that caused that perturbation. Diffusion imaging is an example of this that I use on a daily basis.

In the Quantum AI experiment the principle is the same but applied to a completely different playground. Instead of nuclear spins, they use superconducting qubits and treat them as artificial atoms that can be controlled with astonishing precision. These qubits evolve under a complex Hamiltonian, spreading quantum information throughout the system. Then, by effectively reversing that evolution, the team measures how much of the original “state” returns.

They call this a quantum echo. It’s not magnetization refocusing this time, but the reversal of information scrambling. When the echo fades, it’s not because of T₂ decay or gradient dephasing, but because quantum information has become too entangled to unscramble. This phenomenon is captured by what physicists call the “out-of-time-ordered correlator,” or OTOC.

The mathematical backbone of this experiment would look surprisingly familiar to anyone who’s ever designed a spin-echo or multiple-quantum NMR sequence. The entire idea of evolving under one Hamiltonian and then inverting it is straight out of our playbook. In fact, the authors explicitly mention that solid-state NMR spin networks inspired their approach.

The impressive part of Google’s result is the speed. Their quantum processor finished a computation in about 30 seconds, whereas a classical supercomputer would take roughly 13,000 times longer. Simulating a network of interacting spins grows exponentially harder for classical machines, but quantum processors can evolve such systems directly.

The broader implication is intriguing. If quantum computers can reproduce the kinds of spin dynamics that we’ve spent decades observing experimentally, they might eventually help us simulate systems that are too complex for classical computation: dense solids, coupled spin networks, or even biological macromolecules. It’s not that quantum computers are becoming NMR machines, but that NMR, in a way, was doing quantum computing long before the rest of the world caught on.

Share

P.S. For further reading I recommend this Ars Technica article, and the original papers: “Observation of constructive interference at the edge of quantum ergodicity”, arXiv:2510.19751, and arXiv:2510.19550. Google also put up an accompanying blog post, and a nice short video on Twitter summarizing the link between NMR and quantum computing.

Just in case, here is an honest response. I decided to speak up because I no longer recognize the society I grew up with. Community efforts are being overshadowed by virtue signaling, transparency is at an all-time-low, while membership fees and conference registrations are rising. Many people asked if I was personally snubbed before speaking up, and the answer is an emphatic “no”. If anything, I am probably burning bridges that would have gotten me far if I just hummed along.

That being said, ever since I dropped my ISMRM membership I have been doing very well. I got promoted to a full professor at Montreal Polytechnique a year ago, and I spent several very productive sabbaticals in Skopje, Tokyo and Abu Dhabi. I am running a successful foundation in Macedonia, I am getting invitations to speak at international MRI meetings, and I am organizing a high-profile publishing event in the UAE on December 4.

However, I still dream of changes in the ISMRM and I hope this series of blog posts will contribute to it. For that purpose I have created a survey (link) asking the community for feedback. Feel free to use this survey to voice your concerns, but even more importantly, use it to suggest ways in which the ISMRM could improve. Your responses will inform my next blog post, thank you for taking the time to participate!

LINK TO SURVEY

Share

Read more

Subscribe now

In 2022, right before the ISMRM meeting in London, I spent a day with Derek Jones (ISMRM president-elect) in Cardiff. I told him that I was not happy with the way ISMRM has treated grassroots initiatives, and that the pandemic further degraded our sense of community. I also told him that our return to normal might require some painful reforms, but that it should make us stronger in the long run. Three years later I am even more disappointed, because the society is on auto-pilot while funds and ideas are drying up. Instead of more grassroots energy, we got a purification ritual in Toronto (2023), Sean Deoni in Singapore (2024), and a sustainability-themed meeting in Hawaii of all places (2025). I could write more on these topics, but for now I’ll stick to an issue that is closer to my heart.

* * *

Back in 2015 I was part of a new wave of MRI researchers who recognized the importance of community building. I was asked by Matt Bernstein (the MRM editor-in-chief) to create an outreach program for the journal Magnetic Resonance in Medicine, and together with Erika Raven we started MRM Highlights (150+ interviews to-date). To celebrate the publication of the first MRM Highlights magazine we threw the first Highlights party on the rooftop of the Loof bar in Singapore (2016). It was a blast. We followed it up with a crowded beach party in Hawaii (2017), and a very crowded boat party in Paris (2018). All of these parties were funded by advertisements in the Highlights magazine, and sponsors gladly advertised in exchange for drink tickets at the coolest party in town. For three years we were the new blood of the society, and the ISMRM social media channels were buzzing about us.

In 2019 things changed almost overnight. ISMRM corporatized its social media strategy, engagement dropped significantly, and we were prevented from using the magazine advertisements for the Highlights party. This decision caught us off-guard, but we survived. That is when we decided to create the MRathon (a hackathon for MRI professionals) as a way to keep the community spirit going. The MRathon also allowed us to pay for the Highlights party with the hackathon sponsors’ funds. This resulted in the largest Highlights party to-date (Montreal 2019), with 1000+ attendees spread over three floors and a queue that went around the block.

The positive energy surrounding Highlights allowed us to regroup after the pandemic. Peter Jezzard (the current editor-in-chief of MRM) was extremely supportive of our efforts, and with his encouragement Mathieu Boudreau from my lab kept Highlights going through the pandemic and beyond. We organized three more hackathons (2023, 2024, 2025) and threw a Highlights after-party in a former masonic lodge in Toronto (2023) and a jazz club in Singapore (2024). The ISMRM leadership had a great time at these parties, but every time I would tell them we need more of this energy during the conference, they would shrug and say that the liability is too big. Putting the party aside, I proposed incorporating the MRathon in the weekend program, but all three program chairs (past, present and future) said it would also be difficult.

So when the ISMRM asked me to renew my membership for 2025, I wrote the following:

---------- Forwarded message ---------

From: Nikola Stikov <stikov@gmail.com>
Date: Wed, Sep 11, 2024 at 4:23 PM
Subject: Re: Renew Your ISMRM Membership for 2024 Today - Deadline 27 September
To: Anne-Marie Kahrovic <anne-marie@ismrm.org>

Dear Anne-Marie,

As you know, the ISMRM is very close to my heart, but this year I decided not to renew my membership. The main reason is because I have been trying to bring more community and grassroots efforts to the society, but these efforts keep falling on deaf ears.

In 2015 we started MRM Highlights and we have published hundreds of interviews with MRI researchers. The MRM Highlights after-party became a huge success, and both professors and students love it. We organized three hackathons for MRI professionals in Montreal (2019), Toronto (2023) and Singapore (2024). We have not spent a single dollar of ISMRM funds for this, most of the funding comes from my own lab and from companies that recognize the potential of grassroots community building.

However, the ISMRM stopped sharing the Highlights interviews via their social media channels, prevented us from using the advertisements from the Highlights magazine to organize the Highlights party, and does not want to incorporate the hackathon as part of the weekend program.

At this point I feel that I am putting in more than I am getting out of the society. I really hope this changes in the future, and I remain open to discuss any of these issues with the leadership. I am cc-ing a portion of them on this e-mail because we have discussed these issues and they remain good friends and colleagues. I know they are doing their best within the confines of a very rigid structure.

I am writing this while on sabbatical at NYU Abu Dhabi, working on a publishing initiative that (most of the time) manages to take my mind off the ISMRM. Still, a part of me wishes I was in Hawaii with my friends and colleagues, celebrating their successes and hearing about their research. However, I decided to skip the annual meeting and write this blog post instead, hoping it will make a difference in the long run.

The Highlights party has been the most inclusive and the most sustainable party at the ISMRM. It did not take any money from the membership, it did not use public research funds, and everybody was invited. Despite this (or maybe because of it?), it was sidelined by the ISMRM. In future posts I will explain why this is symptomatic of a larger culture shift in the ISMRM, and what we can do about it.

Share

Read more

I am no stranger to moving between countries, but even by my standards 2024 was a little too hectic. The difference between traveling and moving is all about the duration of the displacement, and for me the threshold is approximately one month. Anything longer than that and I undergo a software upgrade. My body remembers my former routine, my brain forgets my preferences, and even a task as mundane as going to the toilet feels like a backward compatibility issue1.

According to the above criteria, I ended up moving nine times in 2024. Four of these moves were with the entire family, so no wonder our suitcases were rarely unpacked. Every day was an adventure, a blessing, and a surprise. I am grateful to my university for approving and financing this one-of-a-kind sabbatical, and to my family for always looking on the bright side of the move. Still, I am hoping for a more stationary 2025.

As unpredictable as my personal life was in 2024, I compensated for it by accurately predicting worldwide trends. Memories can be deceiving, so I logged each of my predictions on Polymarket and distributed $10 on ten different bets. My goal was not to optimize my earnings, but to perform a sanity check on a world going crazy. In the past I have used calibration graphs to convince myself that I can think probabilistically about questions where the answers are well-known2. This time I made ten predictions about things that haven’t happened yet. I decided to only bet on topics where I had 90% certainty I would be correct. I was hoping that I would get 9/10 predictions right, but it turns out I got all ten! Keep in mind, these predictions did not reflect my wishes. They only reflected reality.

Read more

Dear Mrs. Ballantyne,

My name is Nikola Stikov and I am an international student from Macedonia studying at Stanford. It is the first time I am writing to you, but in the past I sent two letters to your husband, Mr. Robert Ballantyne, expressing my gratitude for the scholarship your family has granted me. I just found out that Mr. Ballantyne passed away, so I decided to write to you and give you my condolences.

I am not sure if you have read my past two letters, so in the hope of establishing a regular correspondence with the Ballantyne family, I will tell you a little bit about my background and what the Elsie B. Ballantyne Undergraduate Scholarship Fund has done for me.

As I mentioned earlier, I am an international student coming from a poor and undeveloped country. Macedonia is still recovering from the Balkan wars and many young people there are left without perspective and want to leave the country. I felt the same way after finishing high school, and I really wanted to study abroad. However, if it wasn’t for the Ballantyne Fund, I could not have afforded my studies at Stanford.

Right now, I am a senior graduating in Electrical Engineering and I am using the last portion of my Ballantyne scholarship. As my undergraduate education comes to an end, I wanted to reflect on my four years here and express my deepest gratitude to your family for making my Stanford experience possible. Thanks to you, I was not only able to complete a B.S. in Electrical Engineering, but also take part in many activities that made my education complete. I became a teaching assistant for the Computer Science Department and discovered my desire to teach and pursue an academic career. I spent a quarter studying in Berlin and got a chance to see Europe (I had never been to another European country before). Stanford has opened many doors for me, and it has helped me make an informed decision about my future. I am staying in Stanford to complete my graduate degree and I am hoping to become a professor of Electrical Engineering.

However, Stanford also made me realize what my home country means to me. I would like to go back to Macedonia and use my education to improve things there. Right now, whoever gets a chance to leave Macedonia does so, and having once left, rarely comes back. I would like to be one of the few to return to Macedonia and help the country get out of the mud it is in. I hope that with a PhD from a renowned educational institution I could do a lot for the Macedonian educational system and stop the outflow of young people from the country. Only then will it be possible to bring Macedonia closer to Europe politically and economically.

As I close this letter, I would like to once again express mine and my parents’ deepest gratitude to you and your family for making my Stanford education possible. I sincerely hope that you will get back to me and that I will get a chance to get to know you better. I am deeply indebted to the Ballantyne family and it would be a pleasure and honor to hear back from you.

Thank you,

Nikola Stikov

2024 P.S. I did become a professor of electrical engineering and I am helping my country get out of the mud. Some promises are hard to break.

During my studies at Stanford University, I taught ten different courses ranging from introductory programming to graduate courses on communication and optimization. The class size varied from ten to a hundred, the classroom moved from California to Japan and back, my roles changed from teaching assistant through TA coordinator to instructor. Yet I stuck to some core beliefs and principles that have made me a better teacher and guided me to the Stanford Centennial Teaching Assistant Award in 2008.

The roots of my teaching philosophy reach back to my first TA training session, when the coordinator said that I should prepare for a lecture as if my IQ will drop by half the moment I face the classroom. This sentence made such an impact, that two years later, when I inherited his job, it was the first thing I told the new TAs. It keeps coming back to haunt me every time my vanity or my laziness get the best of me. In the years since, I have learned that the only way to make up for this apparent IQ drop is to know exactly what to say and how to present the first five minutes in front of an audience. Dry runs are only a part of it. The larger part is putting the presentation in context, and framing the beginning in a way that makes the remaining 40-odd minutes less of a chore, and more of a chorus, rephrasing and repeating the take-home point as often as possible.

Every time I get a chance to teach, I ask myself what is the single thing I want my audience to remember at the end of the day. Then I try to summarize it in one paragraph or catch phrase, so that they hold on to it while thousands of other words drift by. The more universal the summary, the more applicable it will be to everybody present. The more striking the words or images in it, the more likely it is that they will get stuck in memory an hour later. The more connections I make to the summary during the course of the lecture, the easier it is to see the importance of the concept and its place in the vast body of science.

Subscribe now

Once the goal of the lecture is clear, I try to engage the audience with questions, some with a clear answer that could be shouted from the back row (not that people in the back row ever shout answers), some so open-ended that even I do not know the answer. I spend most of my preparation for lecture coming up with those questions, and finding ways to field them as if they are a natural extension of the lecture. By varying the type of questions, I strive to appeal to different personalities in the audience, yielding my soapbox to those that have no trouble responding in public, but also teasing those who will never bother to raise their hand. While this type of lecturing is not guaranteed to be interactive, the goal is not to get everybody talking as much as to get everybody thinking. Using some of those same questions in assignments further drives the points home, and makes problem sets and exams an effective teaching and thinking tool.

I have been fortunate to learn from some amazing teachers, who incorporate many of the tactics outlined above. Without exception, the best teachers have also been the clearest about their expectations from the students. Grading is never perfectly objective, yet publishing clear grading criteria and creating unambiguous assignments is the best way to prevent headaches after the grades are in. This only proves the importance of thinking about the right question to ask, be it during lecture or during the evaluation process.1

Yet all the approaches mentioned above are worth nothing without an appropriate attitude. While I do not believe that teachers should be friends with their students, a certain level of familiarity will help the students identify with the instructor and the subject matter. There are several ways of establishing such a relationship, ranging from learning students’ names early, to asking for feedback or adjusting the office hours to accommodate the students’ schedule. I have spent numerous hours mentoring students and training teaching assistants, both as a teacher and as the Stanford Electrical Engineering graduate student advisor, and I know that the biggest life lessons do not come from getting the right answers on a problem set, but from understanding the principles of problem solving.

The first president of Stanford University, an ichthyologist named David Starr Jordan, said that every time he learned a student’s name, he forgot the name of another fish. I have learned many students’ names, and as a consequence I have forgotten many things I was taught throughout the years. Yet, some truths keep swimming around in my memory, refusing to be forgotten and resurfacing in most unexpected times and places. Whenever they brush against my consciousness, I reflect back on who put them there, and my appreciation of that person increases. I would be happiest if ten years from now I run into a former student of mine, and I instantly recall their name while they fish out a nugget of wisdom I planted in them. That alone would be worth a lifetime of teaching.

Share

1

I have changed my mind about the possibility of “clear grading criteria”. I wrote about it a the start of the semester:

KANTAROT / КАНТАРОТ

Grading is subjective. Period.

In 2017 I was asked to be on a committee choosing PhD fellowships for international students in Quebec. The criteria were strict, and the applications were long (click here to scroll through a typical one). I ended up spending an entire weekend carefully reading the 50+ page applications of 20 students whose career prospects were literally in my hands…

Read more

2 years ago · 1 like · Nikola Stikov

Четири години подоцна почина татко ми, па посакав да ги транскрибирам неговите новинарски снимки за да направам подкаст во негова чест. Татко ми, Александар Стиков, беше новинарот со најмногу изговорени минути во етерот на националното радио, а неговите снимки се дестилат на македонската спортска историја. Тие документираат стотици победи и порази пренесени на слушателите во време кога радиото беше главeн медиум на комуникација. Тие се и архива на личните радости и разочарувања на татко ми, затоа што Аце беше најекспресивен кога беше во етерот. Мало чудо е што овие снимки се зачувани на шест USB стикови (да, да, Стикови), пренесени од неговата колешка Вики низ темните лавиринти на Македонската радио телевизија и копирани повторно и повторно на различни лаптопи и надворешни хард дискови.

За жал, и во 2023 не можев да најдам добра алатка за транскрипција на македонскиот јазик. Се обидов со бесплатните отворени модели на OpenAI (Whisper) и Meta (SeamlessM4T), како и со затворениот и скап модел на Google (Chirp Speech-to-Text). Ниту еден од нив не е доволно добар за транскрипција на теренски снимки, како што се радио преносите од спортски натпревар. На поосновно ниво, овие модели исто така не успеваат во стандардни задачи за транскрипција, бидејќи мешаат прилози и придавки и случајно вметнуваат бугарски и српски зборови и граматика.

Затоа решив лично да се ангажирам и да направам нешто по ова прашање.

Се обратив до Илина Јакимовска, директорка на Институтот за етнологија и антропологија и личен херој кој ја документира македонската култура како професор, блогер и подкастер. Нејзините колеги од Факултетот за природни науки (Љупчо Ристески и Даворин Трпески) ги отворија вратите на архивата на Институтот, давајќи ни пристап до ризница од внимателно транскрибирани теренски интервјуа на различни македонски дијалекти.

Контактирав со Дејан Порјазоски, македонски докторант во Финска кој работи на транскрипција на јазици. Дејан волонтерски истренира неколку јазични модели, користејќи ги архивите на Институтот, подкастот на Илина Јакимовска (Обични Луѓе) и видеата од платформата НАУКА ЗА ДЕЦА на нашата Фондација КАНТАРОТ. Конечно, го поканив Ордан Чукалиев, директор на УКИМ Центарот за напредни интердисциплинарни истражувања (ЦеНИИс), да нè води во создавањето на Буки, отворен модел за транскрипција на македонскиот јазик.

Резултатите ме воодушевија, затоа што конечно имам начин да го документирам животот на татко ми, анотирајќи и пребарувајќи ги транскриптите од неговите снимки. Настрана личната мотивација, сега имаме и алатка за зачувување на македонскиот јазик и наследство.

Илјадници часови на снимени материјали се распаѓаат во влажните ќошиња на македонските институции, чекајќи да бидат пребарливи и индексирани. Моделот Буки е првиот чекор кон зачувувањето на ова богатство. Трениран е со само 60 додатни часа говор, но веќе е далеку подобар од конкуренцијата.

За да можеме да направиме верзија што ќе транскрибира и дијалекти и теренски снимки, би сакал да ве замолам сите да придонесете кон оваа иницијатива. Донирајте го својот глас преку платформата Mozilla Common Voice (упатства) и помогнете ни да стигнеме до 600 часа снимен говор! Ако вашиот глас го употребиме во тренингот, тогаш ќе бидете полесно препознатливи во новата верзија на Буки. Најпотребни ни се гласови на деца и повозрасни луѓе, затоа што нив најмалку ги има во тренинг податоците.

За да можете да се уверите во брзината и точноста на Буки, Фондација КАНТАРОТ ќе овозможи бесплатно тестирање во наредните две недели, користејќи ја платформата Hugging Face. Моделите ќе бидат отворени за користење денес (9 октомври) во 14:30 по македонско на следниов линк:

https://huggingface.co/spaces/Macedonian-ASR/Bookie-Whisper-Macedonian-ASR

Ќе имате пристап до моделот Буки-wav2vec2 (брз и точен), Буки-Whisper (побавен и поточен), како и до споредба помеѓу Буки-Whisper и отворените модели на OpenAI и Meta. Подолу е видео демонстрација како да го тестирате моделот1:

Се надеваме дека наскоро ќе најдеме и донатор кој ќе ни овозможи да купиме компјутер со силна графичка картичка за да продолжиме да го користиме Буки локално. Во меѓувреме донирајте глас, зашто кој донира транскрибира!

P.S. Овој напис е достапен бесплатно за сите читатели на КАНТАРОТ. За да нѐ поддржите во нашите напори...

...ДОНИРАЈТЕ!

The year since has passed very slowly. You know how they say that time speeds up when you get older? Not this year. I have been observing it in slow-motion, as every day brings a little less sorrow than the day before. I have been recounting it from three continents, ten countries, and dozens of apartments and hotel rooms. Most of these places have nothing to do with my Dad, but in each of them I kept up a virtual conversation with him, in which I was patiently listening to his radio recordings and taking notes.

My father, Aleksandar Stikov, was the journalist with the most on-air minutes on national radio, and his broadcasts are a distillate of Macedonian sports history. They document hundreds of victories and losses broadcasted to listeners back when radio was the main medium of communication. They are also an archive of my Dad’s personal joys and disappointments. He was the most expressive when he was on the air.

It is a small miracle that these recordings are preserved on six USB keys, carried by his colleague Viki through the dark corridors of the Macedonian Radio Television and copied over and over on multiple laptops and external hard drives. It is also a big shame that I have no way to transcribe, search, and annotate them, because there is no good tool for transcribing the Macedonian language. I have tried Google’s Speech-to-text, OpenAI’s Whisper, and Meta’s SeamlessM4T. None of them are good enough to transcribe field recordings, such as radio broadcasts shouted over a crowd of thousands at a sports game. On a more basic level, these models also fail at standard transcription tasks, confusing adverbs and adjectives and randomly inserting Bulgarian and Serbian words and grammar.

That is when I decided to do something about it.

Read more

Caroline is a colleague at Polytechnique who has been mentoring engineering PhDs for over a decade. She is a successful professor, enterpreneur, and a role model to generations of students at Poly. Most importanly, she has a positive outlook on academia. For each of my reasons not to enroll in a doctorate, she has a nuanced and thorough chapter addressing the issue and explaining ways to get around it. If I had to summarize her perspective in one sentence, it is that setting expectations is the most important part of doing a PhD. So let’s see how she addresses my three most common arguments against the PhD.

Right off the bat, on page 3, she shows that the cumulative earnings of a PhD overtake the earnings of technicians and engineers 20-30 years after high-school graduation. Not great, because these 20-30 years are full of opportunity costs (imagine investing in Nvidia or bitcoin instead of paying your tuition), but good enough for the pleasure delayers. She dedicates an entire chapter to the PhD after-life and while she acknowledges that very few of the students will become professors, she also shows the variety of careers that benefit from a PhD. Caroline’s main argument is that a PhD trains you to internalize original and independent thinking, and this pays off in spades in the long run. Finally, she wraps the book up with a series of advice on how to navigate the stressful environment where lines between mentorship, competition and friendship are often blurred.

Most importantly, she manages to do this without alienating cynics like myself. She says that PhDs come in all shapes and sizes and debunks the myth that you need three peer-reviewed papers to graduate (I didn’t). She realizes that work always expands to fill the time you allow it (Parkinson’s law), and knows that excellence in one topic requires mediocrity in other aspects of your work. Finally, she recognizes that one cannot do a PhD without entering a flow state, where ambitions, curiosity and inner voices do not matter. It is all about playing to your strengths and being absorbed in your work.

I found part 1 (Doctoral strategies) and part 3 (Tools of the trade) to be remarkably close to my perspectives on academia. Part 2 (Leading a research project) was the one where Caroline and I disagree the most. Her approach to a PhD is methodical, mine is chaotic. Hers requires knowledge of MBA terminology, mine avoids anything that has the words Gantt, SWOT, f2s, TRL, RDI and WBS in them. While being organized is important, I believe that part 2 is a little too structured and makes it sound like the biggest portion of a PhD is planning. She even cites Louis Pasteur who said that chance favors the prepared mind, but I doubt that Pasteur spent his time drawing Gantt charts. The biggest discoveries in the history of humanity happened because of a FAFO approach, and I don’t think Caroline gives this approach enough credit.

Subscribe now

The book has a unique visual language. Obviously inspired by the Piled Higher And Deeper PhD comic, Caroline’s cartoons have wit and wisdom. The schematics on the other hand are a little too busy, and the font choice makes it difficult to read them. I would have preferred more cartoons and less charts, but maybe that is just a consequence of my disorganized approach to research.

Finally, a word or two about things that are missing from the book. The book gives excellent advice on how to write a paper/thesis, but it does not address reviewing articles and responding to peer-review. It also does not spend enough time on the DORA principles, nor to the concepts of sharing data and code to make research reproducible (p. 89 maybe?). Finally, it only tells half the story of networking, ignoring how important name recognition is. Having a well-cited mentor correlates with your academic success, and while it may not be a causal relationship, choosing a mentor based on prestige may give you an upper hand.

All of these are minor complaints about a well-written and insightful book. I always thought I can explain a PhD to a student in a single meeting, so it came as no surprise that after that meeting most students would never want to do a PhD. I never thought that I have internalized 300+ pages that go without saying, and many of those pages actually make a good argument for sticking around. Thank you Caroline for finding my blind spot and for giving my potential PhD students a new required reading.

P.S. If you would like to join a discussion with Caroline Boudoux about her book, I will be talking to her about it today (Friday the 13th) at 2pm Eastern time. Below is a Zoom link, available to our paid subscribers:

Read more

Then we had our committee meeting, and I realized that my grades were very different from those of my colleagues. Yet, somehow, we all managed to agree on the top five candidates, who were the only ones to get the fellowship. Surprised by this easy consensus based on very noisy data, in 2018 I told my wife that I will spend exactly three minutes per application. She told me that was unfair to the students who probably spent three weeks (or maybe three months) putting that application together. I told her that they will not know this (the perks of anonymous peer review).

I opened the applications at 7am on the date of the deadline, and took three minutes to scroll through each application and assign it an approximate ranking. I submitted my scores at 8am when the evaluations site closed, and then I waited for the committee meeting…

At the meeting we reached an easy consensus once again. The best applications rose to the top, and even though our scores were noisy, our recommendations were consistent. That is when I decided to do some statistics. I entered all the reviewers’ scores in a spreadsheet and calculated the Pearson’s correlation coefficients to see how much reviewers agreed with each other. The analysis is here and the results are below:

On average, reviewers agreed with each other with a correlation coefficient around 0.5, which leaves about 75% of the variance unexplained. Not bad, not great, on par with any decision-by-committee. The correlations are significant, yet not very strong. Surprisingly, my scores did not stand out in any way. In 2017 I had a strong correlation (r = .75) with Reviewer 3, whereas all the others’ correlations were between 0.35 and 0.55. In 2018, the two other reviewers did not agree well with each other much (r = 0.3), whereas I managed to correlate with both of them slightly more (0.5 with each).

All to say that my 3-minute experiment went completely unnoticed. Nobody knew about it, nobody raised an eyebrow or questioned my judgement, and statistically my 2017 and 2018 contributions were indistinguishable. Yet, my effort was reduced by an order of magnitude. I took that as a huge win. Whether I spent a single hour or an entire weekend on a review, the scores would remain noisy, the agreement between the reviewers would stay low, and the best would rise to the top. I was particularly happy when I told this story to a group of students in a previous class, and one of them approached me to say that he was ranked first in the 2018 competition. I looked up my score, and he was my top choice, even after a 3-minute review. He was also the best student in the class.

Subscribe now

DOES THIS MAKE ME A BAD REVIEWER?

Not really sure about this one. In the years since, I have scaled down on reviewing papers and grants, and the above analysis contributed to my decision. Why waste my time reviewing when the effort is not proportional to the outcome? However, there was one aspect of reviewing that I could never escape, and that is grading my students in the classes I teach.

That is why in 2019 I started a course where the tagline is ‘3 minutes is all you get’. I told the students that they will give three 3-minute presentations over the course of the class, and that those scores will contribute to 30% of their grades. Because the class requires presentations on a broad range of biomedical technologies, I would invite experts to serve as the jury, I would give the jury a list of criteria (clarity, accuracy and production values) and we would have our own ‘Poly’s got talent’ grading sessions.

The class became popular, and I ended up with long waitlists. The course evaluations were very high in all categories, except the section titled ‘Grading criteria are clear’. That section’s scores were very low. Obviously students did not understand how we put the scores together, but I thought this lack of clarity came with the territory. The experts were three different scientists, they each took the criteria to mean slightly different things, and the same student could end up with a high score from one judge, but a low score from another. It was understandable, but also disorienting for the class.

GRADING IS RANKING

Last year I finally had a realization! It wasn’t the scores that were disorienting, it was me trying to be clear about the criteria. I asked the judges to grade based on clarity, accuracy and production values, but those three concepts are very vague. So in 2023, I told my judges to design their own criteria, and that the only thing I care about is how they rank the students. I also told the students that they will be ranked by the judges and their grades will scale down with the ranking (First-ranked gets 10 points, second-ranked gets 9.5, etc.)

This led to several students dropping out of the class. They were uncomfortable with the idea that they are competing against their classmates and thought that grades should reflect knowledge, not competition. I told them that I understand their point of view, but that the criteria are clear. I also told them that classes are one of the few places where ‘objective’ criteria are still the norm. Companies do not hire based on passing a threshold. They hire based on a ranking where the criteria are often inscrutable and opaque to the applicants. The ‘best’ candidates get the jobs, whatever ‘best’ means. A low ranking should not be taken personally, because no matter how good they are, somebody will think they are not good enough. And if they didn’t like this message, they could always take another class.

The students that stuck around gave me the highest teaching evaluations I have ever received.1 Somehow my methods clicked with this group. They got to hear from amazing speakers, including May Griffith, Danielle Levac, Julien Cohen-Adad, Stuart Buck, Adam Mastroianni, Matt Clancy, and Evelyn McLean. They asked hundreds of questions during class. They created brilliant presentations that got noticed on social media. They got internships based on their final projects. They wrote testimonies about the class (here, here), and four of them will return in class today, to give a guest lecture for my course that starts in one hour. They will only get 3 minutes, I will time them, rank them, and they will not take it personally! Hopefully that convinces this year’s class that ranking is not as bad as it sounds. We pick our jobs, our friends and our partners by ranking, so what’s wrong with assigning grades that way?

Today, following my old students’ 3-minute lectures, a group of 20 new students will show their first 3-minute presentations. I will be their first judge. They are probably uncomfortable with my (lack of) criteria, and are wondering if they made the right decision to stay in a class that explicitly tells them they will be ranked by a noisy and unpredictable algorithm. Some of them will drop out, but some of them will understand that ‘noisy and unpredictable algorithm’ is the very definition of our reality.

I hope all of them will remember the lessons of this blog post!

Share

P.S. If you like the idea for my class, you can subscribe to our new YouTube channel, or follow us on Twitter/X. Yes, social media participation is required for the students’ participation scores, and they will be ranked based on their stats.

This was one of the first things I heard during the Stanford freshman orientation in 1997. The Magnetic Wall is considered by Reddit to be one of the fun “secret places” on the Stanford campus, but it is not a secret for any Stanford alumnus. Generations of students have thrown a metal part at the wall and watched it stick, wobble and precess. In 2019, it was the first place I showed my kids when we went to celebrate the 90th birthday of Al Macovski, the founder of the Magnetic Resonance Systems Research Lab (MRSRL) and the Magnetic Wall. Below is a personal history that reaches a bittersweet conclusion today when the MRSRL scanner will be celebrated and then decommissioned.

May 1998

I finally find out what the Magnetic Wall is about! It is the home of an MRI scanner! As many discoveries, it comes on unexpectedly and from leftfield. Professor Charlie Marcus gives the prospective physics majors a most unique E&M final exam. He tells us to write about a topic that we find personally interesting, and NMR/MRI is one of the ten suggestions. I choose MRI, and even though he tells us not to go over 5 pages for God’s sake, I end up writing ten pages (students, do NOT do this!) I write about Bloch and Purcell, and use cgs units to describe superconducting phenomena. I discover that Stanford has played a critical role in this story, and finish my report with: “What is outlined here are just the most basic physical principles underlying MRI. The rest was beyond my power of understanding.”

September 2003

I wish I could say that the 1998 final exam is the moment when I became obsessed with MRI, but that is not my personality. I take my time with everything, including career choices. For five years I precess around that Magnetic Wall, first thinking I will be a physics major, then (thanks to Brad Osgood) deciding to switch to electrical engineering, then taking and teaching all kinds of EE classes, but never meeting any MRI researchers.

It isn’t until the fall of 2003 that I become the EE graduate student advisor, and John Pauly happens to sit down at my lunch table during the EE student orientation. He is a new EE professor, and there’s something about him that makes me want to be his teaching assistant. I ask Marianne Marx (the legendary EE course coordinator) to make me a TA for EE102a, the introductory signals and systems class. I get the TA job and get to watch John struggle with his first lectures. With each subsequent lecture the awkwardness fades and I realize that I am in the presence of a walking encyclopedia of human knowledge. I ask him if he would be willing to be my PhD supervisor, even though I haven’t passed the notorious EE quals. He agrees and opens the door to the Magnetic Resonance Systems Research Lab (MRSRL) to me1.

Subscribe now

2004 - 2009

Behind that door is a treasure, an MRI scanner from 1989 that somehow manages to look even more ancient than that. In the subsequent years I take Dwight Nishimura’s imaging classes, Brian Hargreaves’ EPIC course, and slowly start to understand the magic of the Magnetic Wall. I dare to scan alone, because you are not a true MR engineer unless you can operate a scanner at 3am in conditions that stretch the limits of IRB approval. I become the scanner’s safety officer2, even if that means dealing with complaints from neighboring buildings that the magnetic field is making their monitors fuzzy (likely), as well as their researcher’s heads (unlikely)3. I do weekly checks of the cold head and helium levels and wonder if this scanner burns helium faster than the sun can make it. And I perform cryogen fills (without any GE oversight), where I obsessively recite the steps inside my head (I am notoriously bad at following directions), and imagine dreadful quench scenarios. None of these scenarios materialize and I become more and more confident around machinery. Even today, whenever I get scared to drive a car (and it is my biggest fear), I remind myself that I used to drive an MRI scanner, and I did its plumbing too.

August 2019

Ten years after my PhD, I bring my family to Stanford to celebrate Al Macovski’s birthday! In preparing the MRM Highlights special issue celebrating his life (thank you Graham Wright!), I gain a deeper understanding of how crucial Al Macovski is to my research philosophy. Shiny new things do not change the world. What truly made an impact was not this ancient GE Signa 1.5T, but the culture of MRSRL and the hundreds of people who infused it with ingenuity, civility and altruism. As a new professor, I take Al’s words to heart: “Find good people and get out of their way!”

December 2022

I return to Stanford after a grueling pandemic. It is December 22, I wonder the empty campus with my family and decide to check out the scanner for which I still have the key (best $10 deposit I have ever made). I run into Dwight4, and he is changing the cryogen at MRSRL! We share a very special post-pandemic moment together and all is right with the world.

August 3, 2024

Knowing I cannot be at Stanford for the MRSRL scanner commemoration, I come to campus for a day, and sneak into the scanner for one last ride in the driving seat. It looks like nothing has changed, but then I do notice something new! A picture of all the MRSRL alumni, hanging in the corner of the console room. I feel immense pride to be on that picture!

August 30, 2024

I am teaching a class in Montreal on Friday afternoon, so I cannot make it in time for tomorrow’s MRSRL scanner celebration at Stanford. I should be preparing my lecture, but instead here I am revisiting a very special relationship with the MRSRL. Even though I left in 2009, 15 years later I still feel the magnetic pull of that wall. As many of you know, I am very good at long-distance relationships, but you need new things to keep those relationships strong. A small piece of me will quench tomorrow, but I know it will be replaced by something equally special, because it has never been about the scanner. It was always about the people who threw something at a wall, and then tried to understand why it sticks.

All that to say: I love you MRSRL!

Share

1

He also opens up a world of opportunities that I still reap 20 years later. John stood by me when I failed the quals (I was the first person below the cutoff). He stood by me when I decided not to write that first paper. He stood by me when my girlfriend (now wife) was rejected three times for a US visa. Here is what I wrote about John in my PhD thesis acknowledgement in 2009 and I still stand by every single word.

2

My personal notes the week I became the MRSRL safety officer

4

For those who know Dwight, and for those who don’t

Let’s start with a song:

This is 34 Ghosts by Nine Inch Nails, the industrial rock band that has surprisingly found their greatest recognition in the country charts1. A sample from the above song formed the basis of Old Town Road, the record-breaking country rap by Lil Nas X that topped the US charts for 19 weeks in 2018!

So how is Old Town Road compatible with Ghosts? Well, Trent Reznor, the NIN frontman decided to release their 2008 album under a creative commons license (CC-BY-NC-SA), which allowed the song to be remixed and reused for free, as long as it was not used for commercial purposes. Ten years later, producer YoungKio sampled it, Lil Nax X discovered it, and they made a ton of money from it. But Trent Reznor was not left out. While their financial arrangement remains undisclosed, he is listed as Old Town Road’s songwriter and producer, making him an unlikely winner of a country music award.

Let’s try to bring this back to MRI.

In 2010 José Marques and colleagues published a paper on MP2RAGE, a popular MRI pulse sequence that has seen wide use for imaging at 7 Tesla. The accompanying code was released under a GPL-3.0 license, which made it possible for anybody to recycle the code, including Siemens. MP2RAGE is currently a product sequence on Siemens scanners, making José one of the most cited MRI scientists worldwide.

Because the sequence is in the public domain, it is now possible to remix it and bring it to other platforms. And that is exactly what we did a month ago, developing an MP2RAGE sequence using the vendor-neutral open-source pulseq framework developed by colleagues from Freiburg University. The development was done at Juntendo University in Tokyo, with collaborators from industry and academia that made it possible to do the work in less than a month.

In keeping with the music theme of this post, we called our project “MP2RAGE against the machine: Scanning in the name of reproducibility.” To the best of our knowledge, this is the first MP2RAGE implementation that runs on GE scanners, and other vendors are in the pipeline. Our next step is to deploy MP2RAGE on the proprietary RTHawk platform, developed by the company Vista. RTHawk is an FDA approved product that is closed-source, yet Vista has made it possible for customers to deploy open-source sequences on it. That is what we did in 2022 with our award-winning VENUS article, and MP2RAGE will be the obvious next step. So we have gone from an open-source sequence, to a proprietary Siemens sequence, to a vendor-neutral open-source sequence, to a a vendor-neutral open-source sequence on the proprietary RTHawk platform. The interoperability is evident. Open science and IP can, do, and should co-exist.

So far I have steered clear of the financial ramifications of open-source technology. One need look no further than Red Hat Linux and bitcoin to realize that open-source projects can make money. In 2019 we wrote an editorial that addressed this issue and explained how the very word “patent” implies openness:

Most would agree that transparency in science is an essential goal, but there are differing opinions on how open our science really should be. The questions surrounding open science and research transparency are particularly relevant for fields like magnetic resonance, in which much of the innovation over the past 50 years has been driven by intellectual property, including patents and copyrighted software, and there is a tight link between industry and academia.
…
There is a common viewpoint that open science is necessarily incompatible with the protection of intellectual property. However, even the etymology of the word “patent” implies openness, and historically patents have offered an important alternative to companies hoarding trade secrets.

As a matter of fact, in my native (Macedonian) language the word “patent” means “zipper”. Once a patent is published, the invention is unzipped, making it possible for other people to use it. The conditions for use are not always straight-forward, but in general patented technologies can be used for research as long as they are not for-profit. And if somebody does decide to profit from a patented technology, they can always sign a licensing agreement and pay royalties to the inventor. While not always used in their intended way, by definition patents are open-source technologies.

If this sounds counterinuitive, you should read my favorite essay on IP and open-source, called The Cathedral and the Bazaar. It is written by Eric S. Raymond, and it describes the differences between the Unix operating system (a cathedral model controlled by a centralized entity) and Linux (a bazaar model where everybody can contribute and modify code). In it, the unlikely champion of the bazaar model is (drumroll) MATLAB:

There have been other software products with a two-level architecture and a two- tier user community that combined a cathedral-mode core and a bazaar-mode toolbox. One such is MATLAB, a commercial data-analysis and visualization tool. Users of MATLAB and other products with a similar structure invariably report that the action, the ferment, the innovation mostly takes place in the open part of the tool where a large and varied community can tinker with it.

So next time somebody tells you that MATLAB is outdated, remind them that there are good reasons why it is still the most used language in the MRI community.

IP and open-science have coexisted for decades, and I hope that the MRI community will realize the potential of the bazaar approach to innovation. For that goal we have developed the idea of MRI FAIR, a vendor-neutral MR bazaar (app store) where developers can share their techniques with vendors and customers. This idea won the ISMRM Shark Tank competition in 2023, and we are now working together with other vendor-neutral projects (Pulseq and gammaSTAR) to create standardized implementations of MRI pulse sequences. We have decided to start with a unified MP2RAGE sequence, because its history makes it a perfect “34 Ghost” candidate. We will continue to remix, innovate and work with unzipped IP to create the Old Town Road of MRI.

Share

P.S. A lot of the ideas for this essay fermented during yesterday’s MRathon, a community-driven effort to make MRI accessible, harmonized and vendor-neutral. Big thank you to everybody who joined us for the third hackathon for MRI professionals.

There were only a few movies that managed to genuinely surprise me, and those comprise my top 5. The rest of the batch still deserve to be seen, and I am grateful to movie studio A24 for keeping 2023 interesting. And then there are the 2023 disappointments, but those are hidden behind a paywall. Oppenheimer is there, along with the Holdovers, the Boy and the Heron, the Little Mermaid, Asteroid City, Elemental, the Killer, Anatomy of a Fall and a few others.

TOP 23 FOR 2023
(CLICK HERE FOR PREVIOUS YEARS, CLICK ON TITLES FOR THE TRAILERS)

23/23 Io Capitano - Matteo Garrone

You’ve seen magic realism, but can you imagine magic brutalism? Matteo Garrone creates a coming-of-age torture porn that doesn’t make much sense, but is nevertheless moving thanks to Seydou Sarr’s stellar performance.

22/23 Phi 1.618 - Theodore Ushev

Would you burn all the books on Earth if you knew that its contents are stored forever in invisible ink somewhere in the clouds? А Bulgarian homage to silent movies, Balkan parties and Terry Gilliam’s Brazil.

21/23 I like movies - Chandler Levack

What would Ghost World look like if its protagonists were teenage boys? 17-year-old Lawrence is working at a video store and is the least likable teenager to grace movie screens in a long time.

20/23 Three - Nayla Al Khaja

A female Emirati director creating the Muslim equivalent of the Exorcist? Despite its simplistic resolution, Three is a thought-provoking and scary experience that doesn’t shy away from uncomfortable topics.

19/23 Poor Things - Yorgos Lanthimos

Has Yorgos Lanthimos decided to follow in the footsteps of Wes Anderson? Impeccable production design and paper thin characters makes this the worst Lanthimos movie yet. However, Emma Stone's performance makes it good enough for the top 20.

18/23 Guardians of the Galaxy Vol. 3 - James Gunn

How do you get me to stay awake during an MCU movie? You don’t, but the single-take battle scene (No sleep till Brooklyn) is one of the best fight choreographies I have ever seen. If only it wasn’t for the 3-hour running length.

17/23 Talk to me - Danny Philippou/Michael Philippou

Can A24 outdo Bloomhouse? Talk to me is a grade-B horror movie that manages to incoroporate a clever allegory about substance abuse and losing your mind to drugs.

16/23 Knock at the cabin - M. Night Shyamalan

What are you willing to do to prevent the apocalypse? M. Night Shyamalan answers the question with a home invasion movie that invokes Cloverfield Lane 10.

15/23 Renfield - Chris McKay

Is Nicolas Cage the only actor to enter my top 20 more than once? Renfield is bloody, funny and perfectly Cage-y.

14/23 Jerry & Marge Go Large - David Frankel

Are you too young to enjoy a comedy about a retired couple who finds a way to game the lottery? Probably yes, but you shouldn’t miss the chance to see Bryan Cranston and Annette Bening giving the young ones a middle finger.

13/23 Hypnotic - Robert Rodriguez

Can you watch a movie, like it, forget about it, and watch it again in the same year? Maybe Robert Rodriguez hypnotized me, but I wouldn’t mind watching this breezy popcorn movie for the third time.

12/23 Beau is Afraid - Ari Aster

Can you be bored and scared at the same time? Leave it to Ari Aster to fill the movie theater with dread that is uncomfortably boring.

11/23 Cocaine Bear - Elizabeth Banks

Did you like Snakes on a Plane? Me neither, but this one is way, way, way more fun!

10/23 Across the Spiderverse - Dos Santos/Powers/Thompson

Can the sequel be better than the original? It could have been, if only it didn’t end on a cliffhanger.

9/23 Zone of Interest - Jonathan Glazer

You have heard of Auschwitz, but have you listened to the sounds it made? Glazer keeps his focus on the idyllic family life of Rudolph Höss, but makes us listen to the sounds beyond his walled garden. A truly uncomfortable experience.

8/23 Killers of the Flower Moon - Martin Scorsese

Did you know that there was a time and place when Native Americans had white servants? Scorsese takes a fascinating slice of American history and lets Leonardo Di Caprio and Robert De Niro play second fiddle to Lily Gladstone.

7/23 Barbie - Greta Gerwig

What if a Barbie doll realized that she is just a toy in a male-dominated world? Greta Gerwig gets a lot of mileage out this high-concept setup, but she loses steam when she decides to tone down the satire and play up the girl power.

6/23 Monster - Hirokazu Koreeda

Who’s the monster and who’s the victim in this story that involves pyromania, bullying and homosexuality? Koreeda’s tribute to Rashomon lets the mother, the teacher and the child give different answers, then unifies them in a transcendant finale.

5/23 Sisu - Jalmari Helander

Are you up for a tongue-in-cheek celebration of westerns, B-movies, war movies, revenge flicks and everything in between? If that’s your cup of tea, then Sisu will make your cup overflow.

4/23 American Fiction - Cord Jefferson

Can you afford yourself the luxury of not obsessing over race when everyone around you does? American Fiction exposes the religious zeal with which white people seek absolution for their racism, even if it means pigeonholing Black artists.

3/23 Past Lives - Celine Song

What if two people were meant for each other but destiny pulled them apart? Past Lives revisits them in ten year intervals and manages to perfectly capture the thrill and frustration of a long-distance relationship.

2/23 Dream Scenario - Kristoffer Borgli

Is Dream Scenario an absurdist comedy or a diatribe against cancel culture? Nicolas Cage is haunting and hilarious as the college professor who gets banished by people simply because they dreamt about him.

1/23 Saltburn - Emerald Fennell

Is there a limit to the perversions that Emerald Fennell has in store for the characters in Saltburn? This is a movie that shocks, thrills and saves the best for last.

And if you cannot get enough of Barry Keoghan and Emerald Fennell working together, then check out this photo session that you will not be able to unsee:

https://www.wmagazine.com/culture/barry-keoghan-emerald-fennell-saltburn-vampire-cover-story

And now for the disappointments. Some of them are intriguing, many of them are overlong, most of them left me cold. Keep in mind, these are not necessarily bad movies, it is just that I expected more of them.

Read more

Read more

Academic publishing is broken. Mainstream news outlets are declaring the scientific article obsolete, renowned scientists claim that research articles are just advertisements for the actual scholarship, forward-thinking enterpreneurs are realizing that pretty soon papers will be written by nobody for nobody, and literature search experts are saying that close to nothing of what makes science work exists as text on the web. Large language models are amplifying the validity of the above statements. Advances in AI are making it easier to write authoritative prose, and scientists are realizing that they need to go beyond text and conventional publishing if they want to make their science count. This is particularly true of biomedical sciences, where research is data-driven and computationally intensive.

When you strip a research article of its prose, what remains are the figures, the data, and the provenance. The first two remnants can be easily manipulated, as has been demonstrated in recent retractions by prominent scientists who generated fake data and images. And while outright fraud is rare, many scientists engage in creating misleading figures, introducing a so-called lie factor (defined as the effect size in the graphic divided by the effect size in the data). This lie factor can be virtually eliminated by publishing the figure provenance, i.e. by tightly integrating the figure generation with the underlying data. However, provenance cannot be incorporated in a static PDF. That is why we have created NeuroLibre, a living preprint server that hosts interactive research objects and replicates the runtime environment necessary to generate the figures and to trace their provenance.

Over the past three years, we have been working on our flagship NeuroLibre publication, and I can proudly say that it is finally here. 42 co-authors, 94 datasets from 18 sites, 5 interactive figures, and an award-winning dashboard, all bundled in our most ambitious NeuroLibre preprint yet:

https://preprint.neurolibre.org/10.55458/neurolibre.00023/

A joint effort between two study groups of the International Society for Magnetic Resonance in Medicine (Reproducible Research and Quantitative MR), this preprint presents the results of the T1 mapping reproducibility challenge. At the link above you can access a PDF of this preprint, interact with a Jupyter Book of the manuscript, and re-execute everything in the browser using Binder (just click the little rocket in the upper right corner here).

We also built an interactive dashboard to explore the phantom and human datasets and analyses. The dashboard (also hosted by NeuroLibre) won a recognition from the company Plot.ly for its innovative visualizations. Check it out:

rrsg2020.db.neurolibre.org

SO WHAT DID WE DO?

Challenge participants acquired T1 datasets using the ISMRM/NIST qMRI phantom (produced by CaliberMRI) and/or human brains using a gold standard T1 mapping protocol. Participants were only given the PDF article where the protocol was published and told to proceed as they see fit. This work was selected as one of the top 50 abstracts at the 2023 annual ISMRM meeting in Toronto #ismrm2023 #ismrm23

Here's a video recording of the presentation and eight take-home points taken from a recent Twitter/X thread:

1/8 CROWDSOURCING THE T1 COMMON GROUND

Now that we showed off the format, time to tell you why this work is important and why it is the only research paper I will publish as senior author in 2023.

But first, a big thank you to joint first authors Mathieu Boudreau and Agâh Karakuzu!

2/8 A PDF IS NOT ENOUGH TO IMPLEMENT A PROTOCOL

Even though everybody implemented the same T1 mapping protocol (Barral et al.), the variability was twice as big across sites than within a single site.

This shows the impact of implementation variations across data submissions!

3/8 YOU CAN CHECK IT OUT FOR YOURSELVES

Using this interactive Plot.ly figure we can compare against the phantom reference values and zoom in to identify outliers.

4/8 QUALITY CONTROL IS ESSENTIAL FOR REVIEW

For quality control, in addition to showing different phantom T1 maps submitted by participants, we gain additional insights by using a dropdown that displays the images at different inversion times.

5/8 DASHBOARDS ARE THE FRIEND YOU DIDN'T KNOW YOU NEEDED

Additionally, we acquired in vivo brain T1 maps. Using our @plotlygraphs dashboard, you can easily switch between measurements to focus on specific regions, sites or vendors.

6/8 KEEP THE LIE FACTOR TO A MINIMUM

With the click of a button, you can change the axis units (eg, % vs. absolute) or the grouping (individual sites vs. vendors) in the statistics section of the dashboard while you’re reading the manuscript.

7/8 VALIDATION IN PHANTOMS IS NOT ENOUGH

A comparison of in vivo white matter at 3T and the NIST phantom sphere that is closest to white matter T1 values at 3T. Variability in phantoms is much smaller than in vivo.

8/8 WHY I WILL NOT PUBLISH ANOTHER RESEARCH ARTICLE THIS YEAR

It took me eight years to publish my first article. This one took only three years, so I believe I deserve a break:)

P.S. I would like to thank Mathieu Boudreau, Agâh Karakuzu and Pierre Bellec for the countless hours they have put into NeuroLibre, and for their help in putting together this manuscript and this Substack post. For a more technical explanation of what we have accomplished, here is a snippet from a popular Mastodon post, courtesy of NeuroLibre co-founder Pierre Bellec:

https://neuromatch.social/@pierre_bellec/111250741252232026

This NeuroLibre preprint is probably unlike anything you've seen before. The science by Mathieu Boudreau, Agâh Karakuzu and a large team of collaborators is fantastic, but I'm talking about the tech used for the preprint itself here. First it's not just a lame pdf preprint. It's got an html version filled with interactive figures, and even a dashboard! But that's not what's unique. What really matters is that it is fully reproducible, and has been tested for it. By clicking on the small rocket, you can reproduce the figures yourself, from your browser. All the data, all the code, all the dependencies have been published alongside the preprint, and the figures have been generated by the neurolibre servers, not by the authors! Each reproducibility artefact has its DOI, and they are cleanly linked to the DOI of the preprint. It is indexed by Google Scholar, Orcid and the like. Neurolibre is based on the amazing Jupyter Book project, and authors can do 99% of the work themselves just by using Jupyter Book and the Neurolibre technical docs. The technical screening of the submission is automatized to a very large extent (it's been adapted from the awesome workflow of the Journal of Open Source Software). Check the publication process out, it's on GitHub!

This project is part of the Canadian Open Neuroscience Platform, funded by Brain Canada and several partners, including the Courtois foundation, the Montreal Heart Institute, the Quebec Bioimaging Network and Cancer Computers.

Share

Овие два збора ми ја креваа косата на глава секој пат кога требаше да прошетам со татко ми низ град. Прошетка што требаше да трае десет минути, траеше саат време затоа што луѓе го запираа татко ми на улица и го прашуваа за Вардар, за Вишњевац, за Панчев… Мислам дека му годеше вниманието. Знаеше дека кога тој зборува, луѓе слушаат.

Неговиот глас одекнуваше низ цело Скопје во недела попладне. Се слушаше од семкарите, од дворовите на љубителите на фудбалот, се слушаше дури и на стадион оти гледачите си носеа со себе транзистор за да го слушаат преносот на Аце од истиот натпревар на кои самите присуствуваа. Му веруваа на Аце повеќе отколку на сопствените очи. Зборот инфлуенсер не постоеше тогаш, но татко ми беше токму тоа, човек кој можеше со еден збор да влијае врз симпатиите на милионска публика.

Можеше, ама не го правеше тоа. Напротив, инсистираше на умереност и мрaзеше суперлативи. Се случуваше со саати да бдее над напишаниот текст во потрага по вистинскиот збор, иако знаеше дека тој збор ќе биде изговорен само еднаш на радио и ќе исчезне во етерот. Знаеше со денови да критикува нечиј пропуст, до точка до која веќе немаше да ти текне да го повториш. Кога преслушуваше прилози од негови колеги ме прашуваше што ми се допаѓа во нив, а што не. Кога ми ги читаше составите по македонски, ги чешлаше збор по збор и ме прашуваше што сум сакал да кажам со тоа. Кога ми ги слушаше презентациите постојано ми ја поправаше дикцијата. Ова инсистирање на прецизност во говорот знаеше да нервира, ама секогаш се исплатеше. Кога Аце ќе дадеше зелено светло, се знаеше дека патот е чист.

За неговите професионални достигнувања се кажа многу во претходниве денови. Неговите колеги не штедеа комплименти за неговата професионалност, како и анегдоти за неговата импулсивност. Знаеше да загрми со гласот и знаеше да удри со рака по маса, ама знаеше и да биде бескрајно трпелив со оние кои му значеа. Аце си ја сакаше фамилијата, пријателите, колегите, си ја сакаше својата работа и го сакаше спортот. Камо среќа сите да можеме да собереме толку љубов во себе и околу себе.

Кога се пензионираше се повлече комплетно од јавниот живот. Не стапна ни на еден спортски терен од каде што известуваше претходните четириесет години. Времето го поминуваше со семејството, со пријателите, со телевизорот и со компјутерот. Одеше кај Ѓутла каде ќе се напиеше една кока-кола и ќе играше карти. Ги учеше внуците Алекс (кој е крстен по него) и Софи математика и песни од Битлси. Религиозно ги архивираше сите медиумски написи за мајка ми, за мене, за Драгана и за нашата фондација, КАНТАРОТ. И не престана да следи спорт, останувајќи и до пет наутро за да го фати последниот настап на Ѓоковиќ или Јокиќ.

Се обидував да го натерам да направиме подкаст во кој ќе ја раскаже својата животна сторија, но не успеав во тоа. Најмногу што добив е една снимена прошетка низ неговите омилени места во Скопје, за која ме замоли никогаш да не ја споделувам со јавност. Жал ми е што нема да ја видите, затоа што тоа беше Аце на 75 години, луциден, духовит и со моќен глас.

Аце беше шкрт на комплименти, но неверојатно горд на сите луѓе кои имаа привилегија да им биде ментор, колега, пријател, сопруг, татко, дедо. Во последните денови се опушти со комплиментите и знаев дека не е на арно. За мајка ми рече дека била како кукла, но дека кога ќе наумела нешто тогаш се фрлала “на прва топка по земја со глава”. За Алекс дека конечно има убава фризура. За Софи дека многу убаво танцува. За Драгана дека е многу добар шофер. А мене, после години во кои инсистирав да снимиме подкаст ми го даде најубавото интервју што еден син може да го направи со својот татко.

Тоа интервју се случи една недела пред да почине, во наплив на енергија кога со саати ми раскажуваше за најважните моменти од неговиот живот. За Пољакињата од Братфорд што му ја поклонила плочата Сан Франциско од Скот Мекензи. За пошумувањето на Водно и за тоа како ја запознал мајка ми како командант на работна акција. За неговиот моден стил, кога и на преноси од втора лига одел облечен во блејзер. За Олимпијадата во Сеул како негов најголем професионален предизвик, за вокменот и патиките кои ми ги донесе оттаму (првите Рибок патики во Скопје) и за написот во Екран во кој Никола Солдатов, импресиониран од темпото со кое Аце известувал од Кореја, го нарекува новинар октопод. За двете омилени награди, Триумфот за медиуми и спорт од Меѓународниот олимписки комитет во 2004 (оти била најпрестижната) и Златната бубамара за животно дело во 2011 (оти била доделена пред најбројна публика). И на крај за мене и дека не би ме пуштил да одам за Америка доколку денес имав 16 години.

Уживав во тој муабет како што не сум уживал во ниеден претходен муабет со него. Ги слушавме заедно снимките од најважните негови преноси и за момент и јас се почувствував како новинар, обидувајќи се да извлечам од него што е можно повеќе информации пред да мора повторно да ја стави на себе омразената маска за кислород. Го пишувам ова откако повторно ги преслушав тие скапоцени снимки, убеден дека заслужуваат да бидат овековечени во еден подкаст. Тој подкаст дефинитивно ќе започне со нему омилената изрека, “Колега Стиков, се слушаме!” А до тогаш…

ТАТО, СЕ СЛУШАМЕ!

Share