Artificial Intelligence and Data Management: AI Meets Healthcare Innovation

By on

artificial intelligence“Remote patient monitoring is essentially going to put a rocket launcher on telemedicine,” says Waqaas Al-Siddiq, due to years of experimentation in healthcare technology with Artificial Intelligence (AI) and the lowering of costs.

He credits the lowered cost for chips and connectivity, along with the proliferation of the Internet of Things (IoT), with creating a cross-pollination process – where technological advances in one area drive progress in another by clearing barriers to accessible, intelligent medicine. Al-Siddiq is the CEO of Biotricity, a remote patient monitoring company, and as such, is uniquely qualified to weigh in on emerging trends in healthcare and AI technology.

Biotricity was created to fill a growing need for remote, proactive health monitoring. The company’s real-time, high-precision mobile telemetry solution enhances a physician’s ability to monitor a patient’s condition while enabling the patient to take an active role in managing their condition.

Al-Siddiq explains how Biotricity was conceived:

“We saw an interesting gap in the marketplace, in which there was a lot of technology and diagnostic equipment stuck in the hospital. It wasn’t cost-effective enough to bring to individuals within their homes so that they could manage their condition on an ongoing basis. Biotricity can now monitor the patient while they go about their daily life, because our intelligent medical-grade device is constantly monitoring and looking for any anomalies, it is basically interpreting and analyzing your ECG signal in real-time.”

When there is a possible issue, an alert is sent to a call center. If it’s an emergency, Biotricity and other appropriate personnel can respond accordingly. Having started with the diagnostic cardiology market, the company is now planning to apply its proprietary remote monitoring platform across a spectrum of connected health applications, including fetal, sleep apnea, and COPD monitoring.

Al-Siddiq says that early experimentation with integration of technology into medical workflows has built a foundation that organizations can learn from and build upon:

“We’ve gotten to a point where some of those experiments have ultimately failed miserably, and some of them have been very successful. The successful ones are now going to move on to the next stage,” he said.

In recent years, experiments using AI with data sets to explore possibilities have informed new developments. “I think we’re going to see sample implementations, prototyping, experimenting –  things where you’re using AI to come up with a suggestion. It’s not really diagnosing, but an experimental process, where the AI is providing feedback to a physician and the physician is giving feedback to the AI,” Al- Siddiq explains

More Experimentation: Watson Goes to Med School

Sloan-Kettering is using IBM’s Watson for cancer screening, but the process is less like a true screening and more like a series of suggestions. After feeding historical data into Watson, doctors work with the question answering computer system to understand the diagnostic process.

Al-Siddiq illustrates what happens when a radiologist shows Watson a scan and asks for a diagnosis:

“Watson will say, ‘Oh, I think it’s cancer,’ and circle five areas, for example, and the radiologist will say, ‘Well, no, you’re wrong. These three are cancer; these two are not, and here’s why,’ and I believe there’s going to be a lot of this type of experimenting going on next year.”

Although he sees significant expansions in the use of Artificial Intelligence within the biggest healthcare organizations, Al-Siddiq doesn’t expect this trend to be widely accessible to the general public in 2018 because clinical decisions can be “a very complicated area.” According to Al-Siddiq, true use of AI is “going to be the last piece of the puzzle in healthcare transformation because of the regulations.” He also insists that while patients may be comfortable with data collection and having input from AI, they want to be reassured that a human doctor is still involved.

Drivers for Innovation

Al-Siddiq believes that there are a few interesting trends driving innovation. Bigger hospital systems like Providence or Kaiser “are absorbing private practices and specialty clinics, and they’re basically becoming a multi-care organization.” He sees this as an advantage, because for-profit multi-care organizations are:

“Free to experiment a lot more. They are not driven by reimbursement – they’re driven by efficiencies. They’re competing. Kaiser is competing with Sutter and wants a better patient experience, better outcomes.”

Providing innovative care attracts more patients. “An implementation of some experimental technology will drive an improvement, so that they have more patients, and ultimately have more profits,” he remarks “They will play with it, but for a mass kind of adoption, you get back into regulatory and payment problems.”

Excellence Distributed via Rule-Based Triage Workflows: The First Real Segue into AI

Al-Siddiq details how organizations such as Sutter, Sloan-Kettering, and the Mayo Clinic are “sitting on huge data sets of clinical outcomes.” He continues:

“This year, they’re going to be taking their historical data and come up with rule-based workflows, so it’s a kind of triage system. It’s collecting information from the patient to determine what the most likely scenario is and where to send the patient for the best care.”

Al-Siddiq sees even further expansion into areas of healthcare where top specialists are in short supply. A heart patient in Texas will go to Texas Heart, “because they have the best heart doctors there.” In 2018, he said, “These organizations [will] try to replicate these types of tools across their organization because you only have one top heart surgeon, or [at most] three top heart surgeons, [whereas] you [only] have five facilities.”

Cancer patients and those with other high-risk conditions will also benefit, as well as ER patients. Someone complaining of shortness of breath might be asked a series of questions about activity level, lifestyle choices, and symptoms, and that information will be fed into an app, which “will collect all of this information and essentially triage you,” says Al-Siddiq. “That is going to be the first real segue into AI that we will see, and I think that’s going to show up this year.”

Partnerships Multiply Top Docs

Al-Siddiq insists that an increase in partnerships among hospitals, specialists, and technology providers will expand the use of AI for diagnosis and treatment in 2018.

He explains:

“AI has consumed the clinical database on outcomes, and now, [it’s] providing suggestions. It’s saying, ‘Oh, I think it’s this,’ and ultimately, it’s the doctor making the call, but in that process, he’s training it, which can essentially replicate the knowledge of that top doctor across the organization. If you’ve got a world-renowned electro-physiologist at Mount Sinai and he’s sitting with an AI device, and that AI’s providing him with a suggestion and he’s telling it when it’s right or wrong, he’s essentially taking it to medical school.”

A new electro-physiologist hired just out of school can then be paired with the AI, expanding the influence of the top doctor that did the training. “In 2018 we’re going to see a lot more of these mash-ups,” says Al-Siddiq.

Previous Efforts Make Way for Improvements

This year, some organizations will also be able to build on past developments and move beyond experimentation into practical implementation.

Al-Siddiq points out:

“Chips hips have become a lot cheaper, they’re a lot easier to implement, and there’s a lot more module providers on the cellular side. You’ve got cellular chip manufacturers who are now creating modules for devices because of the IoT. As a result, it will be easier and cheaper for medical wearables to integrate connectivity.”

He goes on to explain how a simple snap-on device can now:

“A small connected device attached to an inhaler could be used to track how many times someone is using their inhaler. You can cross-reference that with weather patterns to determine the next time someone’s going to have an asthma attack.”

Bluetooth-enabled stethoscopes and glucometers that can record and store blood glucose levels and heart rates in the Cloud are now available and networks like AT&T and Verizon are getting more devices connected.

He references earlier telemedicine experiments that simply involved a face on a screen traveling around the hospital: “A lot of these experiments have failed, because telemedicine has a very limited scope. There’s only so much you can diagnose through a Skype call.”

Previous efforts at remote diagnosis have been limited to issues that rely on visual data, such as burns, but now, “If you’ve got an asthmatic patient, you can collect the asthma data, and if you have a Bluetooth stethoscope you can collect their lung sounds,” he says. “Suddenly, you can [completely] diagnose this issue remotely.”

Solving the Reimbursement Problem

Reimbursement is another area where innovation can run into complications, but the Centers for Medicare and Medicaid Services (CMS) have already approved telemedicine billing, so Al-Siddiq expects this won’t be an issue for long. “In a lot of states, from a physician perspective, it makes no difference to them whether they’re seeing the patient remotely or physically – they make the same amount of money,” he argues.

Cross-Pollination and the Tipping Point

Al-Siddiq sees 2018 as the tipping point at which healthcare organizations will see previous barriers to technological advancements removed, allowing for areas of cross-pollination. For Bluetooth and cellular connectivity, there’s a few things that need to happen. According to Al-Siddiq, one is that “chip manufacturers have to say, ‘OK, we’re going to bring the price down.’ Telecom had to come in and say, ‘Let’s try and bring in connectivity.’”

Al-Siddiq insists that once connectivity is no longer a barrier, remote patient monitoring and telemedicine become viable options:

“This cross-pollination has essentially now come to a point where in about a year and a half – probably in 2019, we will have a snowball effect, where everything’s going to get accelerated. The more data we have, the better our screening, and the better our ability to take the knowledge base of renowned surgeons and clinicians and make it available across organizations. Then, the more connectivity we get, the more telemedicine we get.”

He predicts a major evolution in healthcare delivery after years of transformation with underlying technologies and cross-pollination, and he sees 2018 as the beginning of this evolution: “These foundational layers are actually years deep,” and “it’s going to be very exciting to see how fast healthcare’s going to transform.”


Photo Credit: Panchenko Vladimir/

Leave a Reply