The digital health space refers to the integration of technology and health care services to improve the overall quality of health care delivery. It encompasses a wide range of innovative and emerging technologies such as wearables, telehealth, artificial intelligence, mobile health, and electronic health records (EHRs). The digital health space offers numerous benefits such as improved patient outcomes, increased access to health care, reduced costs, and improved communication and collaboration between patients and health care providers. For example, patients can now monitor their vital signs such as blood pressure and glucose levels from home using wearable devices and share the data with their doctors in real-time. Telehealth technology allows patients to consult with their health care providers remotely without having to travel to the hospital, making health care more accessible, particularly in remote or rural areas. Artificial intelligence can be used to analyze vast amounts of patient data to identify patterns, predict outcomes, and provide personalized treatment recommendations. Overall, the digital health space is rapidly evolving, and the integration of technology in health
Headlines featuring phrases like “AI beats doctors” or “AI beats radiologists” are capturing the public imagination, boasting the impressive abilities of AI algorithms in potentially life-saving tasks such as detecting breast cancer or predicting heart disease.
While such headlines reveal how much exciting progress has been made in AI, in a way they are also misleading – because they suggest a false dichotomy.
It’s true that an AI algorithm must perform on par or even better than a human expert on a specific task in order to be of value in healthcare. But it’s not actually a matter of one “beating” the other.
A lot of the decisions that clinicians make on a daily basis are incredibly complex, and, as I will argue, require more than an AI- or data-driven approach alone. Both clinicians and AI have their unique strengths and limitations. It’s their joint intelligence that counts if we want to make a meaningful difference to patient care.
Here’s why.
The relationship between machine learning (A.I.) is and will be symbiotic. A.I. can identify objects, language, and images that humans may miss. Humans will cross-check results from Machine learning as well as identify errors of machine learning.
Two pairs of eyes see more than one – the need for human oversight We know that even the most experienced clinicians may overlook things. Missed or erroneous diagnoses still occur. In fact, one study by the National Academy of Sciences in 2015 showed that most people will experience at least one diagnostic error in their lifetime [6]. The truth is, however, that AI algorithms will never be infallible either. They may err, too. Even if it’s only in 1 out of 100 cases. Clearly, that is a risk we need to mitigate when an algorithm is used to assess whether a patient is likely to have a malignant tumor or not.
That’s why human oversight is essential, even when an algorithm is highly accurate. Applying two pairs of eyes to the same image – one pair human, one pair of AI – can create a safety net, with the clinician and AI compensating for what the other may overlook.
AI can help track the volume of brain tumors over time by extracting information from MRI images
It’s a compelling example of the power of human-AI partnership. And there are other collaborative models that also rest on the same four-eyes principle. AI can act either as a first reader, a second reader or as a triage tool that helps prioritize worklists based on the suspiciousness of findings [8]. In other words, A.I. can tag images that deserve human interpretation. It would improve efficiency discarding 'normal' and forwarding suspicious images to the physician. This has been done for many years with electrocardiograms.
Deep learning algorithms allow us to see more than we ever could with our own eyes. But unlike human experts, they do not develop a true understanding of the world.
Camouflage graffiti and art stickers cause a neural network to misclassify stop signs as speed limit 45 signs or yield signs.
A cautionary example from the AI literature is a study into visual recognition of traffic signs. An algorithm was trained to recognize STOP signs in a lab setting and did so admirably well – until it encountered STOP signs that were partly covered with graffiti or stickers. Suddenly, the otherwise accurate algorithm could be tricked to misclassify the image – mistaking it for a Speed Limit 45 sign in 67% of the cases when it was covered with graffiti, and even in 100% of the cases when it was covered with stickers [9].
Some have called it the Age of Artificial Intelligence. That may be so, however, integration is the other player in this decade.
Thousands of applications and software solutions have been marketed in the Health care space.
Small vendors have their own niche, electronic health records, remote monitoring, telehealth, natural language processing, analytics, practice and revenue cycle management, patient portals for scheduling appointments, secure instant messaging, and prescription refills
The core application is the one the user, provider, patient signs on to first. Which one is difficult to predict? The evolution of HIT is so rapid prediction may be moot. Most vendors look to the future to plan their next pivot.
A.I. is enhancing integration among different applications in radiology. The artificial intelligence applications will not operate in an isolated role. Provider input will overly and parallel results from A.I. and machine learning. Humans have the ability to error check and identify metrics that A.I. cannot. Some images may have an artifact which will confuse A.I. applications. Providers can use A.I. to confirm their findings and/or refresh their own knowledge base.
Natural language processing, AI, and machine learning travel in the same space. The breakthrough is a result of the collaboration of several technology giants, Google, Amazon, and commercialized by Nuance, is a long-standing player in dictation and transcription. Nuance is a player in the consumer market place, as well in the health care market. Their offerings have expanded considerably and Nuance is in full marketing mode.
Products often appear to be useful when introduced, but the real test is in clinical use. The small things often make a product a winner or a lackluster performance when caring for patients.
Dr. Tarik Alkasab demonstrates PowerScribe 360 Clinical Guidance at Mass General Hospital
Workflow and data entry are key components and individual physicians accomplish their own tasks in different ways from each other. When electronic health records came into existence workflow and computer data entry did not get along. This markedly reduced clinician efficiency less time with face-to-face interaction with patients, and longer work hours. It led to substantial physician burnout and early retirement for many physicians.
NLP has advanced with time and usage. Algorithms require huge amounts of data, text, charts, or images to 'learn'. There is really nothing intelligent about machine learning. The algorithm must be fed, like the robot in the movie,Short Circuit, where a 1986 American comic science fiction film was directed by John Badham and written by S. S. Wilson and Brent Maddock. The film's plot centers upon an experimental military robot that is struck by lightning and gains human-like intelligence, which it embarks to explore.
"Johnny" or Number Five from "Short Circuit"
The robot earns the name 'Number Five', apparently, his serial number given to him by its builders. "Number Five is alive" becomes a trademark announcement by the misguided robot shocked into life by an errant lightning bolt. The lightning bolt reprogrammed Number Five with an algorithm again to machine learning.
Number Five had a consummate thirst for knowledge. When queried his response would often be 'need more data'. He had the unique ability to recognize when he could not compute the answer.
NLP and machine learning also needs to be given data and learns. NLP is not a static program, it will learn, the more it hears the more it can translate into written language. The process is autonomous and should not require manual reprogramming.
Nuance is marketing an expanded set of NLP and artificial intelligence applications:
A unified worklist platform that optimizes workload management and prioritization, consolidating access to imaging and patient information across multiple sources.
An industry-leading, end-to-end solution that simplifies creating and managing lung cancer screening programs, tracking high-risk patients from eligibility through follow-up and reporting.
Automated identification and proactive management of incidental findings recommendations and follow-up compliance, closing the loop on patient follow-up for improved outcomes.
NUANCE and Microsoft Teams are Integrating DAX to be used in telehealth and clinical encounters. Nuance promises the integration of their NLP into the EHR and is embarked on that goal with Epic EHR systems. It is in limited use at several health facilities. This is a work in progress and will be expensive to perfect.
The Exam of the Future using DAX
This first version of DAX is merely a cut and paste version and does not form fill in the correct fields of the EHR. In essence, it is a marketing tool to develop an interest in a later version.
Nuance Dragon Ambient eXperience for Microsoft Teams
In much the same manner DAX Telehealth records in real time as the examination proceeds with little to no action on the part of the examiner.
All these improvements are early renditions, and much more will come swiftly. It is a step by step process. Good things are coming.
Everyone remembers the IBM Watson project. M.D. Anderson was a partner in that adventure into 'artificial intelligence" and machine learning. Here was the first large scale attempt at integrating M.L. into a clinical setting.
The oncology department of M.D. Anderson performs many clinical trials for new drugs. There is a constant flow of drugs in the developmental pipeline. The combinations and permutations are huge and the use of M.L. in the process looked promising. The partnership between IBM and M.D. Anderson lasted briefly. IBM did it's best, however, the clinicians terminated the project due to critical incompatibility between Watson (software) and Anderson. That project seems to be dormant, or at least neither partner has much to say about it publicly.
Physicians have been and are still skeptical of Artificial Intelligence. The inexperience and lack of education in what A.I. can and cannot accomplish are part of the problem. Knowledgable people know that A.I. can only do what it is taught to do using algorithms and a large database of information, such as images, and/or raw data, charts, and natural language. Computers continue to be as stupid as they ever were. It remains doubtful if a computer will ever duplicate the human brain. Humans are creative, computers do not have that ability. Regardless of how much they 'think' has anyone seen a. computer create a new idea? Can a computer have an 'aha!" Do they have an epiphany? Can a computer dream? The answer may lie in the movie iRobot whereSonny discloses human emotionspontaneously when he becomes angry. He did not recognize that he had a human emotion.
Will computers become sentient? In this case, Sonny was not aware of his transition from an automated machine by acquiring human emotion. Was this developed by his creator in programming or did it occur organically?
Despite my reservations, large companies such as IBM, Google, Microsoft, and Apple pave the way for machine learning in smart home devices, smartphones. Software developed by Amazon is being licensed to all varieties of devices manufactured by Sony, Bose, Samsung, and others. This feature has been around for at least 20 years used by Nuance in the original "Dragon Speaking" dictation software for entering text into word processors. Nuance now offers a new iteration, " Dragon Ambient eXperience
IBM has lagged for ten years in its effort to enter the health IT space in a meaningful way (Paddington).
Several rumblings are occurring in IBM's financial structure according to the Wall Street Journal. IBM's lackluster performance in cloud computing against rivals such as AWS, Google Cloud, Apple cloud, and even Dell Cloud have IBM considering a financial split of their IT to develop revenue flow to support such a venture.
Such splits have been disappointing. Boyar Research, in a study last year of 246 spinoffs, found that in most cases the outcome was disappointing. Such companies generally did well in their first year, the New York investment research firm said, but after five years, just over a third were performing better than the S&P 500 index. Parent companies mostly did poorly in the five years after spinoffs, the study found, and on average shares in the parent trailed the stock index. IBM has experience with shedding businesses to reinvent itself. The company was a pioneer in personal computers but exited under pressure from more-nimble rivals such as Dell Technologies Inc. IBM sold its PC business to China’s Lenovo Group Ltd. in 2005 for $1.25 billion, a deal widely regarded as shrewd.
IBM unloaded its semiconductor manufacturing business to the chipmaker Globalfoundries in 2015, an acknowledgment, some analysts said, that it didn’t have the scale or focus to compete in that arena.
Now IBM is exiting a business in a market that has uncertain fortunes. IT services, which include infrastructure services, will decline by 6.8% this year, although the sector is set to rebound next year, according to the research firm Gartner. Mr. Krishna said the business IBM is spinning off will be the largest player in the field and will be vying for a market worth $500 billion, with operations in more than 100 countries.
Measuring success looking at stock prices obviously disregards its value in the workspace. Many outstanding products languish in the stock exchange while delivering real value to the users.
Health IT is often a risky investment. It is a highly specialized area and market pressures change rapidly. Mergers and acquisitions occur almost at the speed of light.
ref: This article is a synthesis of numerous sources:
