The potential for artificial intelligence in healthcare

Artificial intelligence has come a long way since it was first established as a field in 1956. Over 60 years ago at Dartmouth College, a group of scholars organized by computer scientist John McCarthy coined the term. The Dartmouth group wanted to explore the possibilities of having machines solve problems that humans typically solved using their natural intelligence. Today, AI has evolved past that early research and development stage.

In recent years, there has been an amplified focus on the use of artificial intelligence (AI) in business and society to resolve complex issues. Likewise, the adoption of artificial intelligence (AI) in healthcare is growing while radically changing the face of healthcare delivery. AI is being employed in a myriad of settings including hospitals, clinical laboratories, and research facilities. AI approaches employing machines to sense and comprehend data like humans has opened up previously unavailable or unrecognised opportunities for clinical practitioners and health service organisations. Some examples include utilising AI approaches to analyse unstructured data such as photos, videos, physician notes to enable clinical decision making; use of intelligence interfaces to enhance patient engagement and compliance with treatment; and predictive modelling to manage patient flow and hospital capacity/resource allocation. These technologies have the potential to transform many aspects of patient care, as well as administrative processes within provider, payer and pharmaceutical organizations.

The complexity and rise of data in healthcare means that artificial intelligence (AI) will increasingly be applied within the field. Several types of AI are already being employed by payers and providers of care, and life sciences companies. The key categories of applications involve diagnosis and treatment recommendations, patient engagement and adherence, and administrative activities.

There are already a number of research studies suggesting that AI can perform as well as or better than humans at key healthcare tasks, such as diagnosing disease. Today, algorithms are already outperforming radiologists at spotting malignant tumours, and guiding researchers in how to construct cohorts for costly clinical trials. However, for a variety of reasons, we believe that it will be many years before AI replaces humans for broad medical process domains.

Rather than robotics, AI in health care mainly refers to doctors and hospitals accessing vast data sets of potentially life-saving information. This includes treatment methods and their outcomes, survival rates, and speed of care gathered across millions of patients, geographical locations, and innumerable and sometimes interconnected health conditions. New computing power can detect and analyze large and small trends from the data and even make predictions through machine learning that’s designed to identify potential health outcomes.

Machine learning uses statistical techniques to give computer systems the ability to “learn” with incoming data and to identify patterns and make decisions with minimal human direction. Armed with such targeted analytics, doctors may be better able to assess risk, make correct diagnoses, and offer patients more effective treatments.

Healthcare delivery has over years become complex and challenging. A large part of the complexity in delivering healthcare is because of the voluminous data that is generated in the process of healthcare, which has to be interpreted in an intelligent fashion. AI systems with their problem solving approach can address this need. Their intelligent architecture, which incorporates learning and reasoning and ability to act autonomously without requiring constant human attention, is alluring. Thus the medical domain has provided a fertile ground for AI researchers to test their techniques and in many instances; AI applications have successfully solved problems with outcomes comparable to that of human clinicians. As healthcare delivery becomes more expensive, stakeholders will increasingly look to solutions that can replace the expensive elements in patient care and AI solutions will be sought after in these situations. However, cold technology cannot totally replace the human elements in patient care and a model that incorporates both technological innovations and human care has to be investigated.

Clinical Applications of AI Today and in the Future

There are numerous applications of AI on the market today or awaiting approval that can improve patient care and potentially save lives.

Those applications involve pattern recognition, robotics and natural language processing, which includes speech recognition and translation. Machine learning, a “technique that trains software algorithms to learn from and act upon new data to continuously improve performance,” is also increasingly used today,

Here are a few examples of the latest tools that leverage AI and its subsets to augment various areas of medicine and healthcare, such as:

  • Virtual assistants: This AI-driven technology can help people with Alzheimer’s disease with their daily activities. For example, 59-year-old Brian Leblanc, who was diagnosed with early onset Alzheimer’s disease in 2014, started using Alexa on his Amazon Echo Dot for reminders to eat, bathe and take medication. What it enables him to do is to have more control over his life.
  • MelaFind: This technology uses infrared light to evaluate pigmented lesions. Using algorithms, dermatologists can analyze irregular moles and diagnose serious skin cancers such as melanoma. Although this technology should not replace a biopsy, it helps with giving an early identification.
  • Robotic assisted therapy: Bionik Laboratories in Toronto and Watertown, Mass., use robotics and AI to assist patients in their stroke recovery. A robotic arm and hand use digital algorithms to detect motions that patients can’t execute during therapy and guides them through it. It can help patients perform more recorded movements per hour than they would have if working with a physical therapist alone.
  • Caption Guidance: The Food and Drug Administration just approved this AI-powered software, which can help medical professionals capture, without any specialized training, echocardiographic images of a patient’s heart that are of acceptable diagnostic quality. Machine learning trains the software to spot high-quality 2D ultrasound images of the heart and even record video clips of it, changing the way heart disease is diagnosed.

Is your solution truly patient-centric?

We live in a world of consumerism and companies are ever-increasingly focused on doing more to satisfy their customers. However, healthcare is different as a customer is not a typical value or experience seeker, but a patient in need of care. Healthcare in most countries is a fundamental right of an individual, and one of the core tenets of evolution and progress. And hence, every year all the stakeholders including the government, academic and research institutes, pharma and medical device companies, hospitals and the public pour crores into research and identification of solutions for better care.

As all these stakeholders focus on innovation and the next-best thing in healthcare, one must ask if the problem of improving patient care and health outcomes is truly being solved. Every year India sees innovative solutions either driven indigenously or imported from other countries and contextualized for the local healthcare system. However, most of these solutions providers given the siloed nature of the industry take a very microscopic view of the disease, trying to solve for one pain point through their product or service. We have identified a couple of simple yet commonly observed case-studies to highlight the problem and share a patient-perspective as they interact with these innovative solutions, developed to help improve their condition and quality of life.

Case-study I:  An elderly patient on regular dialysis

Let’s take a case of an elderly patient having diabetes and related kidney failure and requiring dialysis on a regular basis. The doctor will do everything he or she can by prescribing the right care, regular procedures and follow-ups. However, there are other facets to the disease management and things that a patient can do himself or herself as well such learning more about the disease, precautions, side-effects, adhering to a regular medicine schedule, regularly measure hemoglobin levels as they tend to deteriorate over time, work with a dietitian on an adequate meal plan and so on.

As we look at these challenges, there are multiple solutions that are presented to the patient including reminder apps for improving compliance, lifestyle management application for diabetes management, diary to maintain a regular log of hemoglobin control, or a home-based hemoglobin measurement device (like the home-based glucometer). As we look at these many great solutions from a patient perspective, a few natural questions that emerge are: (a) how many applications and devices can a single patient manage? (b) how does all the data come together from these multiple solutions, and (c) what if the patient has other disease which hasn’t been covered as part of these disease management solutions?

Case-study II: A child suffering from Juvenile Idiopathic Arthritis

Juvenile or childhood arthritis is a condition where a child undergoes bouts of painful swelling inflammation of joints. While the course of treatment is relatively well-established including physical therapy, pain-killers, corticosteroids, DMARD’s and surgery in some rare cases, given the young age of the patient there are multiple open questions that remain to be answered: (a) how do you motivate the child to get regular physical therapy, (b) how do you alleviate the pain during treatment and, (c) also how do you support the education of the patient during this journey?

With today’s technology evolution such as introduction of gamification in healthcare, digital patient education platforms, and virtual reality to help combat pain, the future of this treatment looks promising. Stakeholders will have to however work on building a holistic ecosystem and not siloed solutions using the foresaid technologies, ensure that the experience for the child remains personalized and more importantly humanized, and continuously ensure that all this eventually leads to improvement of clinical outcomes.

During past few years of my experience in studying the patient journey for various diseases, evolution and deployment of various innovative solutions and the unmet needs, I have realized that there are four key elements that define any patient-centric solution:

  1. Information / Engagement: Quality of information shared, and the degree of engagement achieved with the patient
  2. Channels: Select the most appropriate and effective channel for each activity
  3. Patient Journey: Ability to engage at each stage of patient journey and bring the right value proposition
  4. Personalization: Personalization or customization of the services to the extent possible and in the right context

Below map indicates a reference framework for identifying different pain-points and areas of considerations as we design a solution for any disease area.

Stakeholders need to consider the following factors and answer related questions such as:

  1. Information / Engagement: Are we enabling the patient with all the information and engagement aspects across the spectrum? A pharma company might be more inclined to share prescription guidance and a dietitian might be more inclined to share only a meal plan, but then such an approach would mean that the patient must shop around the ecosystem to for information
  2. Channels: What are the right channels for disseminating the information amongst the available options? Each provider might not have all the channels, but there are ample opportunities to collaborate and ensure the right channel partner is involved
  3. Patient Journey: What stage of the disease journey is the patient going through, and based on that are we addressing the right needs of the patient and the care-giver?
  4. Personalization: Does the solution require personalization, humanization and contextualization based on the various outlined parameters?

During introduction of the recent healthcare reforms in US, one of the most emphasized roles was that of a ‘care coordinator’, someone who can stitch a holistic experience for the patient. We seeing some interesting solutions being introduced in India as well for 360-degree management of diabetes and a few other chronic diseases, which include delivering the medication to the patient on time, timely reminders, education on using the insulin pen correctly, diet advice, answering any other queries the patient has, lifestyle guidance and counseling. While most stakeholders do not have all the capabilities internally, the most encouraging sign is that they are collaborating with each-other with one common goal of helping the patient. Are you working towards the same?

Opportunities & Challenges in Healthcare in India

Very recently I had this great opportunity to give a talk regarding “Opportunities & Challenges in Healthcare Sector in India” during one of the Start-up Saturday sessions organized by Headstart Network, at Ahmedabad Management Association. Let me say this upfront that I am not expert on the healthcare sector in India, still trying to find my feet on the ground in past 15 months working on my own preventive health assessment start-up Remedy Social. However, here are a few experiences and observations based on my efforts in the past few months that I had shared during the presentation & wanted to share with a broader audience so that more people can benefit from the same.

t is not secret that healthcare is one of the largest spend drivers in developed countries, responsible for anywhere between 12% to 18% of GDP for various Western countries. However, despite all the major developments and economic progress, India still remains one of the countries with a lower spend on healthcare with the number being somewhere around 5-6% of the overall GDP being spent on healthcare. Pure macro-economic interpretation of these numbers suggest that Indian economy will be growing at 6-10% per annum and on top of it our healthcare budgets will catch-up with Western nations taking the industry CAGR to somewhere around 18-20%, which is highly impressive.

Also, there are a plethora of opportunities in this field as one has an opportunity to play across multiple sub-sectors as covered in various verticals in the below diagram, and an opportunity to develop a niche across various horizontals based on expertise in specific skill-sets that you bring to the table. While I show this framework, I am of course cognizant of the fact that describing the entire sector is less of an article and more of a book writing exercise, and also acknowledge the fact that one can never be 100% comprehensive in covering the breadth and depth of the sector.

Not going through the pain-staking task of describing all the boxes above, I would like to mention a few that excite me the most as we look forward to the progress of this sector:

  1. Marketplaces: One of the key challenges in healthcare has been access to right information. More than anything else, marketplaces start addressing information asymmetry allowing people to access the best information, products and services without having to pay an additional premium for the same
  2. Digitization: We have been a few years behind the other sectors when it comes to IT sophistication, but digitization creates avenues for a lot of user-friendliness, data creation and analytics, improved efficiency, service quality, etc.
  3. Home Care: Most importantly, home care brings healthcare to the patients home, providing care within an environment which is more preferable for the patient and less expensive as well. It generates convenience, improve care, reduced cost and higher satisfaction. This sector had always existed but in a fragmented and unorganized manner. Now it is being organized and structured, creating new avenues in improving care and patient experience.

Moreover, there is a high level of perceived or actual unmet need amongst both consumers and the care providers that I have experienced. Consumers or patients are looking for more: (a) awareness & transparency in the system, (b) improved accessibility to services, (c) cheaper solutions that make care more affordable, (d) improvement in quality of service, and (e) above all an improve experience while getting care.

At the same time, healthcare providers are struggling with: (a) people issues as workforce availability, recruitment and retention continues to be one of the biggest challenges, (b) process issues in terms of ensuring operational efficiency and managing EBITDA’s, (c) performance pressure for ensuring quality and consistency, both in terms of clinical quality and customer service, and (d) publicity as they need to constantly identify appropriate avenues to improve awareness and attract more patients

Despite of all these gaps and growth opportunities in the industry, if only it were that easy to establish and grow a healthcare enterprise … what more could one wish for. This a very nuanced industry, requires a lot of context around the science, the process and the inherent age-old culture that drives this industry. Here are the top six challenges that one should be aware of and build a mitigation strategy around as you venture into this sector:

  1. Business of Trust: Health is always a sensitive area and people don’t trust you very easily with serving them, it takes time to gain trust of doctors & patients in this industry
  2. Set Stakeholder Interactions: Healthcare ecosystem stakeholders have strong connections, set incentive and revenue sharing models, which you need to understand and carefully position your service within that construct
  3. Very Nuanced: Healthcare is not straight-forward like any other industry where one can go ahead and build aggregators, marketplaces, etc. and leave it open to vendors for listing and servicing individuals. Regulations, workflows, service quality levels, etc. are more complex and nuanced.
  4. Talent is Expensive: As a result of the nuanced nature of industry, talent is expensive in this field. It ain’t easy to get people who are knowledgeable, talented and experienced in this industry
  5. People Don’t Spend: Despite of high cost involved in developing capabilities and servicing individuals, people still don’t spend that easily on healthcare. Health is not always a top priority for many individuals unless it becomes an emergency
  6. Competition from Large Incumbents: Above all these is aggressive competition from large incumbents with deep pockets, that are heavily expanding their service portfolio to provide a 360 degree coverage. Whatever service or product innovation you might develop, usually it is very easy for these large players to emulate and provide it across a wide geography

Now all this being said, this list wasn’t meant to scare anyone or discourage from entering this market. However, the idea is to help others like myself be better prepared as you venture into this promising space. It always been my core philosophy that there is no point pondering … the only way to do a start-up is to “get started” … & find a way to just survive long enough, till you succeed …