How Artificial Intelligence is Transforming Modern Healthcare

Artificial intelligence is slowly, but surely, showing potential in improving modern healthcare. In the UK, researchers recently used four AI algorithms that beat doctors in predicting heart attacks. Moreover, Google’s DeepMind is fighting blindness with machine learning.

Lately, medical science is seeing potential in the ability of AI systems to find meaning in datasets that are too complicated for us to process. This potential is perfectly applicable in modern healthcare practices.

In this article, we will discuss how artificial intelligence is revolutionizing modern healthcare:

Medical Decision Making

Medical practitioners are human. Long hours can add strain and slow down medical analysis. Consider emergency room radiologists, who sometimes have to look at 200 cases per day and 3000 images per study. The sheer volume of data can be too much for radiologists, who are already a scarce resource in many countries.

Now, artificial intelligence is being considered as a tool for reducing this volume of data and to expedite medical decisions.

Data gathered and presented by AI algorithms will enable healthcare providers and doctors to see patients’ health risks and take more precise, early action to prevent, lessen the impact of or forestall disease progression. These interventions will curb healthcare costs and lead to improved patient health outcomes. – Derek Gordon. COO of Lumiata

Consider Medical Sieve, a project in progress by IBM. Once launched, this cognitive medical assistant will assist in clinical decision making in radiology and cardiology by using advanced multimodal analytics and clinical knowledge. The cognitive health assistant is designed to detect radiology images faster and more reliably.

Visual Diagnosis

AI will revolutionize the modern healthcare by accelerating the time to diagnose hard-to-diagnose conditions, such as cancers, Alzheimer’s, multiple sclerosis and other rare disorders. – Srinivas Kowta. Senior Director of the Health Analytics Vertical of Axtria

Physical examinations are a large part of medical diagnostic practices. Certain types of markings on the skin can help doctors detect conditions like measles fairly easily without the need for further tests. Conditions that can be diagnosed with behavior, like Parkinson’s (characterised by slow movement, tremors) also offer visual clues that doctors can identify.

And now some technologies are using AI capabilities, like facial recognition, eye tracking and pattern detection, to create systems that make this a little bit easier for medical providers.

The ability of AI systems to recognize patterns, learn, and make decisions based on what they figure out has always been intriguing. Google Photos has a facial recognition software which allows you to search and catalogue images of specific people amongst the thousands of photos you have on the application.

When applied to AI systems in healthcare, these catalogued images can be of thousands of patients who exhibit similar physical symptoms that are unique to specific medical conditions.

And it has already been put into practice:

RECOGNIZING FACIAL SYMPTOMS:

Technology that allows AI systems to detect faces in digital photographs is now showing the same potential in identifying physical identifiers in certain medical conditions.

To illustrate, consider Face2Gene phenotyping applications that use face recognition and machine learning to help healthcare providers in identifying rare genetic disorders. These applications pull data points from a photo and compare it to images of patients from a database, who have also been diagnosed with these disorders.

DETECTING MENTAL CONDITIONS:

A patient who has a mental condition often exhibits behavior that makes his or her condition apparent. That behavior or mental condition has to be taken into account when designing eLearning solutions for them, for example. To detect these conditions in children earlier some medical technologies are turning to AI.

Consider the eye-tracking technology RightEye LLC. The technology innovator recently launched an AI powered Autism Test which allows providers to use eye tracking technology to identify early stages of ASD (Autism Spectrum Disorder) in children aged from 12 to 40 months.

During the test, an eye tracking tool analyzes children by presenting different images on the screen. Based on this, health care providers determine which child has a healthy brain (they focus on faces on screen) and which exhibit autistic visual tendencies (focusing more on other objects on screen).

DETECTING MALIGNANT DISEASES:

Certain type of skin markings, like lesions, can be indicative of medical conditions. Identifying them can help medical practitioners detect malignant conditions like skin cancer earlier.

Some diagnostic systems are now using AI algorithms for this. DermaCompare is a prime example that uses AI algorithms to compare photos of melanoma moles with images of 50 million known moles uploaded by patients and doctors around the world.

Health Tracking

Physicians have to examine, diagnose and prescribe treatment plans to hundreds of patients. Making sure that each of them, or at least repeat offenders, keep track of their health can be counterproductive.

The bottom line is that it’s impossible. But now, innovations in AI tracking softwares are aiming to change this:

SELF MONITORING:

Self monitoring tools empower patients to keep track of their own health. Innovations in AI powered health tracking applications are helping them get there.

Consider Vi, an AI personal trainer that lives in biosensing earphones. The AI powered digital assistant monitors your heartbeat, learns from your fitness data, and uses this knowledge to personalize a training workout for you.

MEDICATION ADHERENCE:

Healthcare providers are often at a loss when patients either don’t take medication as prescribed, or suffer from conditions that make them incapable of keeping up with treatment plans.

Consider degenerative conditions like Alzheimer’s disease, in which patients exhibit symptoms like loss of memory and thinking skills. To solve this, the healthcare industry is now turning to artificial intelligence for a solution.

To illustrate, consider the AICure app that uses artificial intelligence and a smartphone’s camera to help healthcare providers confirm if their patients are taking their medication as prescribed and on time.

The facial recognition technology confirms medication ingestion and adapts to patient behavior over time.

Treatment Plans

In 20 to 30 years, we really will be living in the Jetson era. By then, big data, the internet of everything, precision medicine and AI will have converged. A body scan will verify what we can already predict based on genetic mapping. Therapy, diet and treatment will all be personalized to the individual, and treatments like chemotherapy will seem as barbaric as leeching. – Al Babbington. CEO of PrescribeWellness

For healthcare professionals, creating a treatment plan is not always a straight road. Patients might exhibit resistance to certain medication for example, or a patient may start showing new symptoms which compels providers to revise treatment plans.

Recently, AI has offered a solution to the problem:

CONSIDERING TREATMENT OPTIONS:

Medical practitioners create treatment plans based on a patient’s medical history and data on his specific conditions. To make more informed decisions, they are now turning to AI’s deductive and data mining abilities.

Consider IBM Watson for Oncology, a system that draws meaningful medical insights from structured and unstructured data in clinical reports. The data helps physicians identify key information in patient records and explore treatment options to create more informed treatment plans for patients.

PROMPT DETECTION OF MEDICAL CONDITIONS:

The problem with rare medical conditions is that there aren’t many medical records available to compare them with, which creates delays in devising treatment plans.

Medical practitioners are using deep learning, a process of interpreting information by comparing it with an extensive amount of other data. The aim is to understand these conditions better, expedite diagnosis and treatment.

To illustrate, consider Enlitic, an AI platform that can interpret radiology images in milliseconds by comparing it with other unique medical cases. According to the Economist, the platform was 50% better at identifying malignant tumors than human radiologists.

We also custom designed an algorithm with the help of Emory’s Medical Researchers & Doctors for a cloud based research web application, that mimics early stage AI bots, for Emory University.

The artificial intelligence based suggestion system we developed, analyzes given patient symptoms and suggests treatment based on them.

Wrapping up

The use of artificial intelligence has started to show remarkable results in modern healthcare. Medical systems that use AI are hoping it can fill the gaps in helping medical practitioners improve patient care. In the years to come, AI can be a game-changer for healthcare.

Want to learn more about artificial intelligence? Check out the biggest challenges artificial intelligence faces.