How AI is Revolutionizing Patient Engagement and Care

Doctors have been trying to personalize patient care for decades, but the technology and regulations are now available to make this a reality. Initially this was simple telehealth but today, with AI-driven insights from data collected by providers and patients alike, we can identify at-risk individuals earlier than ever before and treat them with the best possible solution: an individualized plan of action to take into account their unique needs and lifestyle.

In 2012 I was presenting at a conference and showed Cue – a unique cube diagnostic system in which one could insert samples of pinprick of blood, urine, saliva etc and it would diagnose a range of medical issues. It then connected to your physician, your pharmacy and your schedule to make the process of diagnosis and treatment seem-less. I gather it ran into some FDA hurdles along the way but the FDA has now approved it for at home diagnostics.

It is a very cool system although I suspect companies like Apple with both existing sensors in their devices, are probably already working on something like this. To integrate this kind of functionality into a phone makes a lot of sense. And the tech is possible, as Cue proved.

There are numerous other approaches in diagnosis of conditions. A friend of mine who is an architect is working with a tech company that have mirrors that have diagnostic ability so when someone looks at themselves in their bathroom mirror, the mirror is assessing any micro-changes that could be predictive of certain medical conditions and flags anything suspicious. She also has toilets that analyse their contents and flag anything suspicious. My architect friend is already using these in homes that she is designing for clients who are increasingly health and wellness focused.

But what about complex conditions like oncology?

Treatment in oncology of any type, especially when there are multiple cancers involved, is fairly unique to every individual patient, so personalization of medical treatment is crucial. In this world, AI can be utilized to tailor treatment to specific needs (and is also used in predicting drug responses), improving cancer diagnostics and treatment, and assisting general patient care. This is because AI can be used to analyze and combine masses of data to come up with the optimal recommendations.

Diagnosis

We have seen variations of simple at home diagnosis in a few conditions recently with the Google Dermatology AI diagnostic tool.

I recently came across an AI start-up who are leveraging AI to improve cancer diagnosis and treatment delivery by building an individualized software framework that keeps track of all your data from many sources – from DNA sequencing information, lab work results and a lot more for both yourself as well as other patients to benchmark your data against. They use state-of-the art deep learning with data science expertise as well as expertise from research scientists, oncologists and statisticians in order to tap into decades of medical knowledge while also training AI algorithms how best to identify possible cancers.

There already are many AI-powered tools that aid diagnosis for oncology including for breast cancer, thyroid cancer, melanoma and more. On top of the sophisticated tools, even simple tools powered with AI are becoming more widespread in oncology.

Treatment

Wearables are being used to assist fight chemotherapy- induced nausea with fitness plans, while smart phones are being used for a variety of diagnosis and detection. A start up called ‘Butterfly network’ are using smart phones for ultrasound for tumour detection.

We already have companies (e.g. Medtronic) releasing digestible pill cams (minuscule cameras in pill) using tech to provide images inside the body from a pill. These are used in GI today. It is not such a stretch to see them being used to take biopsy of cancers. Nanobots (tiny robots the width of a hair) also are likely to be used for biopsies as well as providing targeted treatments.

At home care

Monitoring and interventions for symptoms are the pillar of cancer care by oncology nurses. Using digital health technology allows nurses to assess and monitor more effectively. There are quite a few of these on the market including the Symptom Care at Home system. This system tracks eleven common symptoms in patients undergoing chemotherapy and gives instant feedback for self-management directly while also notifying healthcare professionals when symptoms exceed thresholds. A study found that this system gave 43% reduction in symptom severity over usual symptom care. There are other tools similar to this one.

Another interesting approach being used is the Advanced Symptom Monitoring System tool that is a mobile phone–based tool that can monitor toxicities from chemotherapy during the treatment phase of care.

There are also tools that have been developed for palliative care to monitor and track symptoms, distress, classify pain severity and to identify important clinical discussions.