The Medical Story No One Tells

Radiology: built in a war

Consider what happened with X-rays.

The rays were described in 1895, by Wilhelm Röntgen, with little implementation, then when the stakes were at their highest, Marie Curie developed the field of radiology, at scale.

It was the First World War, and the casualties from wounds and trauma were piling up. Marie Curie knew it was time to bring X-rays into clinical practice, or the situation would turn to the worst. So she developed the mobile X-ray machine, the petite Curie. She wrote the safety procedures. She created the workflow, and she trained the staff to read what the images showed. Around 1.2 million wounded were examined.

She did what needs to be done to make adoption possible at scale, no one could have done it better.

Anaesthesia: proof in the open

On 16 October 1846, in the surgical amphitheatre at Massachusetts General Hospital, a dentist named William Morton administered ether while John Collins Warren, founder of the hospital and first dean of Harvard Medical School, removed a tumour from the neck of Edward Gilbert Abbott. Abbott felt nothing. Warren turned to the sceptical audience of physicians and students and said: "Gentlemen, this is no humbug."

That day is remarkable, because the medical community saw proof, and when there is proof, there is adoption.

The mRNA vaccine: a decade of work in a year

The most recent proof is the one we all lived through.

For years, mRNA sat in the category of promising and not yet delivered. The platform existed, the science was maturing, but no mRNA vaccine had reached the world at scale. Then the need arrived, all at once, and everything changed.

The genome of the new virus was published on 10 January 2020. Moderna and the NIH had settled the vaccine's design the next day, 11 January, one day later. The first clinical batch was completed and shipped to the NIH by 24 February, forty two days from sequence to vial. The first participant was dosed on 16 March, sixty three days from sequence to a human arm. Emergency authorisation came on 18 December 2020, with an efficacy of 94.1%. A process that normally takes a decade or more was done, safely and through full trials, in about eleven months.

That mRNA platform did not appear from nowhere. It stands on decades of quiet, foundational science, and part of that foundation is the discovery of split genes and RNA splicing, the understanding of how the genetic message is edited before it becomes a protein. That discovery earned Richard Roberts the Nobel Prize in 1993.

Decades later, when the world needed a vaccine in months rather than years, the science was already there, waiting. That is the quiet miracle underneath every fast adoption. Necessity moves us, but it can only move as far as the foundational work already laid.

A personal note. Richard Roberts has been my mentor since I was fourteen years old. From high school, through university, and in the years since, we have never lost touch. He shaped my professional path, and he taught me to pursue an idea with courage and never to take no for an answer. That the foundations of his life's work would one day help the world through a pandemic is, to me, the whole argument in a single life: do the science well, and you never know whose life it will one day save.

Compare that to almost any other field

Construction builds today in ways a bricklayer from 1970 would recognise. Its innovation intensity has been persistently low for decades, and it has captured very little from the technological breakthroughs happening around it. In one survey of ten sectors, workers ranked construction dead last for technological competence. AI use in the sector sits around 1.4%.

Healthcare workers, in the same survey, ranked second.

Other fields adopt aggressively too, such as finance, where adoption is fast and often runs ahead of medicine on machine learning and automation. Aviation is deeply digitised, and innovates with deliberate restraint because its safety record was bought that way.

Bodies vs airplanes

But look at what the validation is worth. Aviation trains in simulators, and so do we, but there is a caveat. Both fields face external variance that no one can standardise, the events that arrive from outside the machine and cannot be engineered away. However, aviation solved internal variance. It standardised the machine itself and engineered that machine to withstand the external world. Every Boeing 737 is, to a remarkable degree, every other Boeing 737.

Medicine cannot do this, because the machine is a human body, and there is no standard body. Every patient varies biologically, biochemically, physiologically and anatomically, and again by ethnicity, by sex, by age. The internal variance that aviation engineered out of existence is, in medicine, the permanent condition. And the body is not a more complicated aircraft. It is complex on a different order entirely, still only partly understood by the people who spend their lives studying it. You cannot standardise what you have not finished mapping.

Due to these variance constraints in regard to the human body, the stakes of innovation are higher than in aviation, and yet, as the history shows, the profession innovates at the pace of aviation, and often ahead of it.

Speed is not recklessness. In medicine, speed is treatment.

There is a finding that should be quoted more often. Hospitals that adopt new technology fastest have survival rates about 3.3 percentage points higher than the slowest adopters. That gap accounts for roughly one third of the total improvement in heart attack survival over two decades. Researchers found that a one standard deviation increase in the rate of technology diffusion has the same effect on survival as doubling every other input in the system.

Read that again. In medicine, adopting faster is itself a life saving intervention. Hesitation has a body count. The profession has always understood this at some level, which is exactly why it moves the way it does.

Physicians adopt new medicines within months. Specialists move first, family physicians follow within roughly four to six months, and the curve plateaus inside two years. The pioneering instinct in medicine has always lived closest to the patient, in the individual who sees the need first and reaches for the tool. That is where the courage to try something new has always come from, and it is where it comes from now.

AI is the next instrument, and it is ours to command

Look at what is happening right now. Clinicians are taking up AI at remarkable speed, exactly as they took up ether, X-rays and the vaccines that ended a pandemic. The world watches this and reaches for the wrong story, that doctors are being replaced by the technology.

That is not what is happening. That has never been what happens.

Radiology was not the end of the physician, it was the physician gaining a new sense. Every instrument in the history of medicine was picked up by a clinician, argued over by clinicians, tested by clinicians, and then commanded by clinicians in service of a patient. The instrument never led. The clinical mind led, and the instrument extended its reach.

The clinical mind is more powerful than the world gives it credit for, and, on the harder days, more powerful than clinicians give themselves credit for. What a doctor or a nurse does in an encounter, holding a story, a body, a history, a context and an intuition built across thousands of patients, and then deciding under uncertainty, is not something a model performs. It is the thing the model serves.

So the task now is not to fear AI, and it is not to submit to it either. It is to lead it. To demand that it shows its sources. To demand that it surfaces uncertainty instead of hiding it. To keep the right to overrule it, always, and to insist it be built expecting exactly that. To pick it up the way this profession has always picked things up: with rigour, with argument, with scrutiny, and then with total commitment once it earns its place.

Where Keikku stands

We build Keikku as an instrument for the clinical mind, never as a substitute for it.

Everything it does is designed to extend what a clinician can hold and reach in the moment of care, evidence that comes to you cited and grounded so you can judge it, documentation that carries your reasoning rather than replacing it, and diagnostic signal from the body itself, brought to the clinician who decides its medical role. The clinician stays always in command.

Medicine has never been conservative. It has been careful, and it has been bold, at the same time, because the person in front of us deserves both. AI is simply the newest thing we get to argue about, test, and then master.

We have done this before. We are very good at it. And we will do this again.

Introducing the clinical augmentation you deserve: Keikku across tablet, laptop, phone, and iPad.

Sources