人們一生中要做很多選擇。有些選擇微不足道，例如早晨穿什么衣服；另一些則可能改變整個人生，例如選擇從事的職業。但有一項非常重要的決策不由任何人定，即來到這個世界成為人類一員，以及天生的疾病和缺陷。

這便是醫療行業明顯有別于其他行業之處。醫藥行業經常可以決定生死。而在世界上絕大多數地方，醫藥行業都很混亂、低效且昂貴，急需大力改革。

先看些讓人揪心的數字。2015年美國近四分之一非老年成人出現醫藥費逾期。比起2012年這還是進步，當時逾期比例接近30%（比例降低的原因之一可能是平價醫療法案下醫保覆蓋面增加）。2015年的密西西比州，37%人口欠醫療機構錢。醫療支出是個人破產最主要原因。去年聯邦政府預計每人醫療支出將超過1萬美元，為史上第一次。

公共和私人醫療系統如此昂貴，實際運作如何呢？現實是想面對面看醫生得等好幾個星期，超過6500的地方醫療專業人士沒法滿足病患需要，而且美國醫院的治療結果比起很多發達國家都很一般（甚至比不上一些欠發達國家）。

現狀不妙，說明是時候顛覆了。要來一場真正的改革需要各方參與，從政府、行業到醫療消費者自身都要調整，目前來看，拖著美國醫療行業不情不愿走進21世紀的倒是一群數字醫療公司。

數字醫療的好處在于擯棄了實體局限，例如以前非得建起醫院，現在則是充分發揮移動技術的力量，通過潛心研究人類自身尋找復雜疾病的治療之道。

為了一窺科技引領的未來有哪些可能，我們先來看看21家創新公司在忙什么，這些公司分為五類，每一類都在挑戰各自領域的傳統治療方式。

歡迎參觀數字醫療革命。

There are many choices we make over the course of our lives. Some are fairly insignificant, like the clothes we put on in the morning; others, such as the vocations we settle on, have life-changing consequences. But there’s one critical decision we don’t get to make: the choice of being born into a human body—and all the arbitrary ailments and inevitable biological breakdowns that follow.

This is what sets health care apart from other industries. The ?business of medicine is quite literally one of life and death. And throughout much of the world, it remains a messy, inefficient, ?expensive sector in need of radical reform.

Just consider some of the heart-wrenching numbers. Nearly one in four non-elderly American adults had past-due medical debt in 2015. That’s actually an improvement from 2012, when the figure was closer to 30% (insurance coverage gains under the Affordable Care Act are one likely reason for the decline). In Mississippi, more than 37% of the population owed money to care providers in 2015. Medical expenses are the top driver of personal bankruptcies. And last year the federal government projected that the nation’s health care bill would top $10,000 per person for the first time in history.

So what do we get for these extravagant private and public costs? A system where it takes weeks to see a doctor face-to-face, where more than 6,500 locales are officially deemed to have too few medical professionals to meet patients’ needs, and where U.S. health outcomes are consistently mediocre compared with those of many of our developed-nation peers (and even some of the less developed ones).

This status quo is ripe for disruption. And while true reform will require all the relevant parties—government, industry, and health care consumers themselves—to make major adjustments, an insurgent group of digital health companies is doing its best to drag American medicine into the 21st century kicking and screaming.

That means superseding physical constraints like having an actual hospital by harnessing the power of mobile technology, making the act of taking your medicine less of a hassle, and peering into our very biological building blocks to wage war on the most intractable maladies.

To offer a preview of what this tech-optimized future might look like, we identified 21 innovative companies in five categories—each of which is challenging the conventional approach to medicine.

Welcome to the digital health revolution.

Telemedicine

How health care is transcending physical walls.

At the turn of the 20th century, getting a checkup in America frequently meant your doctor came to you. Armed with a modest black bag of tools and old-fashioned medical know-how, physicians of yore would often take care of you right at your bedside.

As quaint as that image may seem today, it’s in some ways a vision of the future. New technologies are bringing back the house call—or a digital version of it, anyway.

The Companies

A slew of telehealth companies, from the Boston-based ?American Well to San Francisco startup ?Doctor On Demand, are again bringing health care to your doorstep—or more commonly, to your workplace. The National Business Group on Health last year surveyed 133 large companies employing 15 million Americans about their benefit practices: An astounding 90% said they expect to make at least some telemedicine services available to their workers this year. By 2019, nearly all of them will.

So what does that brave new world look like? Say you’ve been working after hours and suddenly feel feverish and woozy. You would walk down the hall to your company’s on-site digital health station—that’s a fancy word for “kiosk” (which, in turn, is a fancy word for “booth”)—where you would consult with a physician immediately by phone or video. That same doctor can take your vital signs—temperature, pulse, blood pressure—and if needed, send a prescription to the nearest pharmacy. Or maybe just tell you to go home and get some rest.

In December, American Well partnered with Concentra, which has more than 300 medical centers in 40 states, to provide some of those on-call physicians.

One obvious advantage of telemedicine is that the doctor can be anywhere—which is not the case for our traditional health care system. At hospitals across the country, in fact, there is an entrenched doctor shortage, which has grown only more acute as millions of Americans have gained health coverage in recent years.

New digital platforms are helping to solve that problem too. One New York startup, Nomad Health, pairs doctors with hospitals in need of physicians in three specialties—internal medicine, emergency medicine, and psychiatry. Nomad cofounder and CEO Dr. Alexi Nazem, in an interview with Fortune, likens the system to an Airbnb model for medical staffing. The professional matchmaking process lets health systems find doctors with specific credentials (say, an internal medicine specialist with five years’ experience who is free to work at a New York–area hospital in July) and vice-versa. The platform then automatically takes care of tasks like providing malpractice insurance if the hospital and doctor strike a contract.

This isn’t just convenient from a medical resource perspective; it also plays into a nascent “gig economy” in health care that’s become increasingly popular among younger, tech-savvy physicians. “With very little effort we’ve been able to tap into a huge flow of doctors who want to do short-term, freelance, flexible work,” says Nazem. “And I think that trend will continue [among millennials] not only because of the attitudinal shift [on freelance medical work] but also because the existence of technology actually makes it possible to do this kind of work.” Last summer the company raised $4 million in an early funding round led by First Round Capital and RRE Ventures.

And it’s not just medical startups driving this mobile health surge. Ride-sharing giants Uber and Lyft have started new programs to provide non-emergency medical transport for patients who struggle to get to a hospital. One platform that launched last fall, called Circulation, integrates medical records into Uber’s API so that nurses, caregivers, and hospital transportation coordinators can more easily schedule rides for patients and accommodate their needs (such as if they have a wheelchair or trouble seeing).

The Future

These technologies may one day fundamentally shift the way that health care is delivered and consumed. But before that transformation takes hold, some other changes will have to happen—including new reimbursement rules from insurance companies and policy shifts that make it easier for physicians to practice across state lines without gaining extra licenses or accreditation. There’s also the matter of whether mobile medical care will ultimately reduce national health spending: At least one recent report suggests that the technology may well cause people to pursue care they don’t need precisely because it makes it so convenient to get.

Still, don’t bet against the trend. For consumers, after all, convenience has always been king—and convenient and cheap? Well, there’s no beating that.

算法醫藥

大數據和人工智能推動學習。

近來“大數據”談得太多，人們已經很難想起大數據的“大”到底怎么回事。據IBM統計，每天新增數據達2.5百億億字節。醫療行業里，每小時新增的研究報告、臨床試驗、科學研究和病人醫療信息不計其數。醫生和醫療研究人員得顧及每個細節。

機器學習和人工智能要參與。計算機沒有人類的弱點，例如要吃要睡，所以可以幫助人類迅速遍覽學術文獻，還可以查看CT掃描，電子病歷，以及海量臨床試驗和基因研究數據。人工智能可以告訴醫藥廠商哪些病人療效最好，還能改變醫院的管理方式。

相關公司

藍色巨人IBM的蛻變仿佛一夜之間，向來西裝筆挺示人的百年老店從呆板的大型機生產商和咨詢行業巨頭變成了數字醫療行業的領頭羊——而且全靠一款人工智能產品。IBM應該好好感謝超級計算機“沃森”（Watson），這款贏下美國智力游戲Jeopardy!的認知機器明星。如今IBM旗下沃森醫療部門業務十分繁忙，而且似乎每天都在拓展人工智能的應用領域。

沃森醫療成立兩年來，迅速與不少知名學術機構達成合作，例如史隆凱特琳癌癥研究中心、注明生物醫藥公司輝瑞、美敦力和強生公司等。每起合作中沃森的角色基本一致：處理大量數據尋找暗含的規律，最好能得出新觀點。沃森不僅能解讀電子病歷，還能處理所謂的非結構化數據（例如X光片或腦補掃描片提供的數據）。

跟遠程醫療一樣，沃森的處理結果也可實現移動化。例如今年早些時候，沃森負責腫瘤的部門首次與佛羅里達朱庇特一家327床位的社區醫院合作，利用超級計算機為癌癥病人尋找最可能奇效的療法（也要多虧史隆凱特琳癌癥研究中心精心審查過的臨床數據支持）。

要是人工智能和深度學習能幫醫生分析一下病人的臉，不用掃描或測試就可以診斷該多好？波士頓創業公司FDNA就在努力通過Face2Gene平臺實現。該公司收集了罹患2000種罕見基因病的病人照片建立數據庫。醫生可以拍下病人的照片，上傳到FDNA應用里，程序可以通過面部特征對比分析出可能存在的病癥（這項技術并非診斷工具，但能縮小可能的基因病范圍。）FDNA的目標是大為縮短罕見病的“診斷歷程”，目前罕見病病人平均要就診七次才能確診。

這類技術的主要目標是降低成本，盡可能及早開始治療，也可以用來解決一些醫院的管理流程痼疾，例如等待時間過長。去年10月，通用醫療(GE Healthcare)和約翰·霍普金斯醫院(Johns Hopkins Hospital)聯合成立了全數字中心，以提高日常運營效率。該中心名叫朱迪賴茨指揮中心（Judy Reitz Capacity Command Center），每分鐘會從十多個霍普金斯醫院信息系統獲取約500條信息，通過預測分析將大量數據轉化為行動建議，避免流程不暢，幫助病人要么盡快入院治療或要么迅速出院。

從約翰霍普金斯醫院的案例來看，初步效果非常不錯。該醫院表示，指揮中心指派救護車前往別處的時間減少了超過一小時，急診室病人安排床位的時間減少了30%。

未來

來看看嚴峻的現實吧：就算計算機能力再強，如果數據不共享也無法根據數據作出判斷。可能聽起來不可思議，但白宮抗癌登月計劃特別小組前執行主任格雷格·西蒙（Greg Simon）表示，“在醫療世界里信息還十分匱乏。”

近年來，聯邦政府和私營機構一直鼓勵數據共享，舉辦了類似基因組數據共享之類的“統一數據儲存庫”活動，鼓勵公開研究結果以加快癌癥治療。但若想取得實質進展只靠一些公共私營領域的活動還不夠，要改變的是整個思路，要變得愿意積極共享，而且切實做到才行。

Algorithmic Medicine

Big data and AI push learning.

The term “big data” gets tossed around so casually that it’s easy to forget just how big “big” really is. Consider that 2.5 quintillion bytes of data are created every day, according to IBM (ibm, -1.79%). In health care, this amounts to an hourly avalanche of new research papers, clinical trials, scientific studies, and patient health information. And it’s impossible for doctors and medical researchers to keep up with even a tiny fraction of it.

Enter machine learning and artificial intelligence. Shorn of human weaknesses like the need to eat or sleep, computers are now speed-?reading through not only the vast academic literature but also CT scans, electronic medical records, and mountains of data from clinical trials and genomic studies. AI is also giving drugmakers critical insights into who benefits most from their treatments and changing the way hospitals manage their administrative operations.

The Companies

With IBM, the metamorphosis seems to have happened overnight: The starched-shirt centenarian went from a stolid mainframe and consulting giant to an upstart digital health pioneer—all in the blink of an AI. Big Blue can thank Watson—its Jeopardy!-winning, cognitive computing superstar—for that. And the company’s ravenous dealmaking business unit, IBM Watson Health, seems to be redefining daily what AI applications can be used for.

In the two years since Watson Health launched, it has struck partnerships with an impressive array of academic institutions, such as the Memorial Sloan Kettering Cancer Center and prominent biopharma companies like Pfizer (pfe, -1.14%), Medtronic (mdt, -1.52%), and Johnson & Johnson (jnj, -0.86%). In each case, Watson is given roughly the same task: Feast on reams of data and find hidden patterns—and hopefully new knowledge—within them. Watson can do this with electronic medical records as well as with so-called unstructured data (like that found in the image of an X-ray or brain scan).

And as with telemedicine, its insights can be made mobile. Earlier this year, for instance, Watson’s oncology-focused unit struck its first deal with a community hospital, a 327-bed outfit in Jupiter, Fla., that can now harness the supercomputer’s power to match cancer patients with the treatments most likely to help them (thanks to a clinical data set expertly reviewed by Memorial Sloan Kettering).

But what if AI and deep learning could help doctors figure out a patient’s disease simply by analyzing a face—no scans or testing required? That’s precisely what Boston-based startup FDNA is trying achieve with its Face2Gene platform. The firm has put together a photo database of people with more than 2,000 rare genetic diseases. Doctors can snap pictures of their patients and upload them to FDNA’s mobile app, which then spits out a list of disorders they might have by analyzing telltale facial features associated with those conditions (the tech is not a diagnostic tool, but rather a way to narrow down the list of possible genetic suspects). And the company hopes the system can drastically improve the “diagnostic journeys” that those with rare diseases typically face: Such patients, on average, are seen by seven doctors before the correct diagnosis is made.

Cutting costs and catching on to illnesses as early as possible are major goals for this type of tech. But it can also be used to combat administrative headaches like long hospital wait times. Last October, GE Healthcare and the Johns Hopkins Hospital launched a fully digital hub to better manage everyday operations. The Judy Reitz Capacity Command Center gets a constant influx of data about important events at the hospital; it receives about 500 messages every minute from more than a dozen different Hopkins IT systems and with the help of predictive analytics turns this swamp of data into suggestions for action that prevent bottlenecks and get patients both into and out of the hospital faster.

And, according to Johns Hopkins at least, it’s showing impressive early results. The hospital says the command center has shaved more than an hour off the time it takes to dispatch an ambulance to another facility and that emergency room patients are assigned a bed 30% faster than before.

The Future

Now for a sobering fact: All the whiz-bang computing power in the world can’t find knowledge in the data if the data isn’t shared to begin with. And surprising as it may seem, says Greg Simon, former executive director of the White House Cancer Moonshot, we “still live in an information-scarce medical world.”

The federal government and private organizations have tried to encourage sharing in recent years through initiatives like the Genomic Data Commons, a “unified data repository” intended to hasten cancer cures by making research public. But real progress on this front will take more than a handful of public-private initiatives: It will require a change in mind-set—a radical new willingness to share and share alike.

Next-Generation Capsules

The evolution of drug delivery.

The hypodermic needle made a splash on the medical stage in the 1850s by combining two key inventions: the conventional syringe (eventually converted from metal to glass so that users could better measure doses) and sharp, hollow needles. While the innovation was initially used for ?purposes such as injecting pain sufferers with powerful opioids, it became a true game changer once insulin came on the scene in 1921. Unlike painkillers, insulin can’t be ingested—it has to be administered via an injection or a pump in order for the body to absorb it and control blood sugar.

The way that we take our drugs has continued to evolve in the decades since then—syringes became disposable, for example—and the progress is far from over.

The Companies

In the past few years companies like Braeburn Pharmaceuticals, Intarcia Therapeutics, and Proteus Digital Health have set out to create better medical mousetraps through devices that make existing drugs more effective. That means peace of mind for diabetes patients, who no longer have to constantly prick their fingers to measure blood sugar and manually adjust insulin doses, and hope for recovering opioid addicts, who might face a relapse if they don’t adhere to a strict treatment protocol.

Last May, Braeburn and partner Titan Pharmaceuticals became the first companies to win Food and Drug Administration (FDA) approval for an implant to treat opioid addiction. Their product, Probuphine, is made up of matchstick-size implants placed into patients’ upper arms in a simple outpatient procedure. These devices dispense a drug called buprenorphine—itself an opioid, but one that doesn’t produce the sort of euphoric and addictive high that more powerful painkillers such as OxyContin and morphine do.

But rather than have patients take buprenorphine manually, Probuphine dispenses small amounts of the drug continually into the bloodstream, ensuring that people actually stick to their prescribed regimen. The device can be used for up to six months.

This type of automated, long-term delivery system could make it easier to treat everything from brain diseases to diabetes. Indeed, diabetes is one affliction that Intarcia has in its crosshairs. The Boston company is seeking FDA approval for an under-the-skin pump system that it says can dispense a steady stream of diabetes drugs for a period of six months or longer. But ?notably—even before it has won marketing approval—some think the technology could be an effective weapon against another deadly scourge: HIV. The Bill & Melinda Gates Foundation said it will invest up to $140 million in Intarcia, in the hopes that its device can deliver prophylactic medicine—and help shield patients at high risk of infection from HIV.

The common theme of these innovations is that medicine itself is necessary but not sufficient: Patients actually have to take it as directed for it to work. Nonadherence to prescribed regimens is also costly, resulting in nearly $300 billion a year in wasted spending, according to some studies.

Automating the process is one way to tackle this problem; Proteus Digital Health, however, is taking a different approach. The Silicon Valley startup’s “smart pill” platform, Discover, helps doctors track patients’ biometrics—and whether or not they’re sticking to a drug protocol—with the help of ingestible (and on-the-body) sensors that communicate with a smartphone app. This way, patients with chronic diseases like hypertension and diabetes (and their physicians) can figure out the most effective dosing regimens.

The Future

Improving the way we take drugs could save those in the developed world many billions of dollars in annual health care expenses. It could also transform how the world’s poorest nations prevent and treat disease—whether through patches that administer vaccines or through long-acting implants that dispense HIV/AIDS drugs. Even in medicine, it seems, the delivery business is in the throes of revolution.

Genomic Revolution

Precision-editing the code of life.

Adenine. Thymine. Guanine. Cytosine. These four tiny compounds provide the basis of life. They are the “letters” that make up DNA’s code, and their various permutations can determine everything from our physical appearances to our risk of being born with a devastating disease.

So it’s not surprising that being able to manipulate these chemical building blocks is one of the most exciting prospects in medicine. And thanks to a gene-editing technology known as Crispr-Cas9, it’s become a whole lot easier to do just that.

The Companies

Crispr has been widely celebrated as one of the most (if not the most) groundbreaking biotech discoveries of the 21st century. That’s not to say gene editing is new (it isn’t), but Crispr simplifies the process by using molecular scissors that can be precisely targeted to snip out aberrant regions of genetic code, which can then be replaced with the correct sequences.

The technology’s possibilities are staggering—in theory, allowing medical scientists to do everything from cure genetic disorders like sickle cell disease to identify gene targets for combating HIV. Silicon Valley billionaire and cancer immunotherapy patron Sean Parker is funding the first U.S. Crispr trials in humans, which are expected to begin this year at the University of Pennsylvania and allied institutions; in March, pharma giant Allergan (agn, -1.80%) struck a $90 million deal with Crispr specialist Editas Medicine (edit, -3.06%) for access to the biotech’s experimental therapies to treat rare and serious eye diseases.

And Crispr-Cas9 isn’t even the only type of Crispr out there: On April 12, researchers at the University of Texas Southwestern Medical Center announced they had successfully paired the gene-editing tool with a different kind of enzyme, called Cpf1, to correct mutations associated with the devastating muscle-wasting disorder Duchenne muscular dystrophy. Crispr-Cpf1 could potentially prove even more promising than the Cas9 variety because the Cpf1 enzyme is smaller and can target certain genomic regions that Cas9 can’t reach.

What makes Crispr so exciting is that, thanks to its precision, the tool has opened up a world of innovation to research facilities that previously wouldn’t have been able to handle the expenses or complexities of genome editing. The possibility that different Crispr-associated enzymes may be more effective than others is fueling the scientific competition. So is the fact that Chinese scientists at Sichuan University in Chengdu launched the first-ever human Crispr trial, in a lung cancer patient last October—a milestone that American scientist Dr. Carl June predicted would “trigger Sputnik 2.0” and a “biomedical duel” in the field between China and the U.S.

The genomic revolution isn’t just a scientific one, though. It’s also regulatory. Gene-related tech has a way of putting the fear of God—or the fear of man playing God—in people. And regulators have been cautious when it comes to the field.

That’s partly what makes 23andMe’s landmark FDA victory in early April so notable. The Alphabet-backed Silicon Valley startup, valued at $1.1 billion, became the first company allowed to sell genetic tests (and accompanying health-risk reports) for 10 different diseases directly to consumers without a prescription. That includes conditions like Parkinson’s, Alzheimer’s, and celiac disease.

Victory wasn’t always assured. In November 2013, the FDA sent 23andMe cofounder and CEO Anne Wojcicki a stern warning saying that the company’s tests and health reports, which it was already selling straight to customers, were unapproved medical devices that hadn’t been cleared by the agency. The firm had to shelve many of its services as it worked to convince regulators that its genetic tests were accurate and the accompanying medical risk reports clear enough that they wouldn’t confuse or harm customers.

Now 23andMe has pulled off the kind of comeback that’s rare to see in biopharma. “The FDA has embraced innovation and has empowered people by authorizing direct access to this information,” said Wojcicki in a statement following the clearance.

The Future

Genomic technology has evolved from the stuff of science fiction to a tangible reality, with massive medical and financial implications. Just how high are the stakes? Enough that the brilliant minds behind Crispr-Cas9—University of California at Berkeley’s Jennifer Doudna, her academic partner Emmanuelle Charpentier of the Max Planck Institute for Infection Biology in Germany, and rival Broad Institute of MIT and Harvard scientist Feng Zhang—and the various biotechs affiliated with them are embroiled in an ugly, global patent spat over the rights to the tech. (Zhang and the Broad won a key intellectual-property ruling in the U.S. earlier this year, but the matter is far from settled in markets like Europe and Asia.)

And ethical concerns will continue to dog this space. The technology isn’t quite advanced enough to birth a world of “designer babies.” But even in the case of 23andMe’s home DNA kits, some question the morality of telling a customer he is at high risk for Alzheimer’s when there’s little the person can currently do about it.

No one said revolution was easy.

Pharma's New Frontier

The radical new models for drug discovery.

The lines between Big Pharma and small biotech are eroding. Traditional drug giants have caught on to the not-so-dirty little secret that outsourcing drug research (and in-licensing) may be a more effective strategy than trying to create groundbreaking new molecules within the confines of a single lab.

“No one company can corner in or create a monopoly on the best thinking that’s out there. And so if we’ve built the lab and we invested a billion dollars in discovering medicines, we can only do what we could in the four walls of our village,” as Allergan CEO Brent Saunders, a licensing-and-acquisition maven, put it during an interview with Fortune earlier this year. “But if we walk outside of that, we’re fishing in an ocean vs. a pond for innovation.”

But shifting collaborative models aren’t the only forces changing the face of drug development. Some companies are thinking not just outside the box but outside the planet when it comes to improving medicine.

The Companies

Elon Musk’s groundbreaking private space outfit, SpaceX, flew its 10th mission under a NASA commercial resupply contract in February. The Dragon spacecraft delivered payloads to the International Space Station (ISS), including some high-profile biopharma cargo from Merck (mrk, -1.24%) and others.

Merck has been collaborating with the Center for the Advancement of Science in Space (CASIS), which has been tasked by NASA to oversee the ISS’s U.S. National Lab since 2012. CASIS’s mission is to encourage “use of this unparalleled platform for innovation”—and Merck is taking this to heart by experimenting with drug development in the realm of microgravity.

Such an environment, it turns out, presents all sorts of opportunities that aren’t available down here on Earth, as Merck structural chemistry scientist Paul Reichert explained to Fortune in an interview.

For instance, you don’t get the kind of gravity-driven diffusion that makes molecules scatter to a specific destination based on their density (think of excess sugar falling to the bottom of an oversaturated glass). One of Merck’s main interests involves growing protein crystals, which form as larger and more organized structures in microgravity.

On the ISS, Merck has been doing experiments on its next-gen cancer medicine Keytruda. The mission, says Reichert, is “to understand the impact of the microgravity environment on the structure, delivery method, and purification” of these types of compounds. The lack of gravity-induced molecular changes on the U.S. National Lab could help Merck researchers improve drug delivery and manufacturing methods on Earth.

The Future

From tech incubators to world-class research facilities, from tweaks to biological machinery to computers that can unlock its secrets, and from research conducted on our pale blue dot to the cosmos themselves, the health care revolution is within our grasp. But ultimately realizing it will require collective changes in policy and scientific culture—and recognizing that technology, like humans, has its own limits.