AI & Robotics Transforming the Healthcare Industry

By adoption of artificial intelligence and robotics, healthcare industry can be transformed drastically. Numerous companies have already started investing in the same industry. Tata Elxsi is amongst the world’s leading providers of design and technology services across industries including Automotive, Broadcast, Communications, Healthcare and Transportation. Tata Elxsi is helping customers reimagine their products and services through design thinking and application of digital technologies such as IoT (Internet of Things), Cloud, Mobility, Virtual Reality and Artificial Intelligence. In an email interaction with Nitisha from BISinfotech; Biswajit Biswas, Chief Data Scientist, Tata Elxsi explains the transformation of healthcare in India through the usages of AI and robotics.

Covid-19 pandemic gave a boost to telemedicine and telehealth as people were indoors. What are your takes on it?

You are right. Covid-19 took the entire world by storm. There was hardly any time to prepare for what was to come. Countries around the world reacted with stringent actions and guidelines to tackle the spread. While some countries contained the proliferation of covid-19 to a large extent, most of them bore a high social and economic burden. However, a common theme between all the countries was the unprecedented stress that overwhelmed healthcare systems. The covid-19 outbreak has exposed how inadequate our global healthcare systems are to deal with the rising healthcare demand. We saw elective surgeries getting deferred to a later date and hospitalizations only for life-threatening conditions.

Telehealth became a necessity for people and healthcare providers to continue regular consultations and care for chronic conditions. In the US, telehealth services have been used 78 times higher for outpatient care and consultation compared with February 2020, according to public reports. Apart from the willingness of patients and providers to use telehealth services, government backing also contributed immensely to the upsurge in telehealth adoption. The government realized the need and waived off regulatory requirements. Likewise, they also brought it under the purview of reimbursable services.

The demand for telehealth has since stabilized following the first increase last year. It is, nevertheless, considerably higher than before the pandemic. We have noted improvement in the attitude of public and healthcare businesses towards telehealth. The government has also made certain telehealth services permanently reimbursable. The other marker that shows increased activity in this space is investments in digital health. For example, 2020 saw 3-times more venture-capital led investments compared to 2017. In conclusion, telehealth is here to stay for good, and many medical specialties will immensely benefit from it.

Tata Elxsi has been working on building telehealth solutions for more than 10-years. We have established ourselves as a preferred design, engineering, and regulatory compliance service provider to healthcare businesses. Our end-to-end capabilities have allowed us to successfully commercialize solutions ranging from connected medical devices, mHealth applications, Software as a Medical Device (SaMD) to a complete telehealth platform for video consultations. Last year, we launched our HIPAA and GDPR compliant customizable digital health platform that virtualizes major touchpoints across the patient journey leading to improved patient experience. Since its launch, we have witnessed high interest not only from the healthcare providers but also from the medical device and pharmaceutical drug manufacturers. It indicates that healthcare businesses are more open to transitioning to digital platforms for healthcare services and patient engagement. In conclusion, telehealth is here to stay for good, and many medical specialties will immensely benefit from it.

What are your points on the increasing popularity of Remote diagnosis and efficient health monitoring measures using AI?

Remote patient monitoring (RPM) is establishing itself as an effective tool for chronic condition management and longitudinal care. As providers aim to replace a considerable number of in-patient monitoring with RPM in the coming years, we expect further growth in this field. Moreover, traditional technology businesses like Apple, Google, and others have started investing in healthcare and wellness devices and solutions, expanding their product and services ecosystem.

The volume of patient health data has expanded significantly with the massive influx of wearable devices. While RPM devices generate valuable patient health data but collecting the data is not the end goal. To integrate these RPM-derived data streams into clinical workflows, data analytics and clinical AI are essential to produce accurate predictions for preventive measures. For instance, the combination of AI and RPM can help providers generate a holistic view of a patient or population for actionable insights and timely intervention leading to better clinical and financial outcomes.

AI is now considered a vital piece of all remote monitoring and diagnosis programs. On the upstream, AI can help providers recruit patients likely to respond to RPM interventions by analyzing their clinical, socioeconomic, and experiential data. Further downstream, it enables proactive care such as early diagnosis of chronic conditions, anticipating adverse events, and mitigating disease progression. Moreover, the combination of remote monitoring and AI also promotes self-management of chronic conditions. We are developing digital health solutions that ingest patient health data and leverage AI to help patients with chronic conditions to gain more insights into their health and improve their self-management practices with personalized insights.

AI can also influence the design of RPM devices. Although RPM has significant benefits such as reducing hospitalization days, chances of re-admissions, and improving the quality of diagnosis and treatment, the challenge is the reluctance of patients to accept the technology that restricts their mobility and various social and cultural issues. While the technology has improved significantly over the past few years, for example, the smaller form factor of the wearable device, longer battery life, there is a lot of scope for further improvement using AI.

Using advanced AI algorithms, manufacturers can reduce the number of sensors in the wearable device and still maintain the device’s efficacy. AI algorithms can identify complex conditions using minimal data points. We are seeing many RPM products are already using state-of-the-art AI-based diagnostics and doing a lot of heavy lifting in the cloud, making the device side thin and non-intrusive to the patient. AI-driven RPM will have a relatively lesser learning curve which will further propel its adoption.

The government of India initiated digital health IDs for all Indian citizens. This initiative will enable data aggregation at the individual and macro levels to leverage digital services using AI. So all these indicate the steps in the right direction for AI and data-driven services.

How technology is changing the role of doctors as well as the role of patients?

Technologies are disrupting the care delivery landscape. We are seeing the shift in the locus of control from physician to patient. The patients are becoming more aware, involved, and empowered with this shift. For example, digitization has moved a lot of traditional medical knowledge, which was exclusive to medical doctors, to the public or digital sphere. Additionally, technologies such as IoT, sensors, telecommunications, AI, etc., are enabling “care anywhere” for patients.

Doctors, on the other hand, are the new-age system architects and managers. AI is bringing one more layer to augment the decision process and reducing the workload on doctors and caregivers.

Also, when healthcare is shifting to outcome-based reimbursements, doctors and clinicians have more incentive to upskill themselves for technologies. As it is evident from the current scenario, technology will eventually become widespread in the clinical environment. Medical practitioners must design care pathways that harness technology capabilities to deliver optimal clinical outcomes and financial outcomes.

Has AI reduced the healthcare costs? Explain.

Perhaps the most common incentive to leverage AI in healthcare is to reduce the cost of service when the burden for outcomes is on the providers. Providers tend to incur high-cost due to wasted medical expenditure. One of the common reasons for high medical spending is treatment variability or misallocated treatments with little to no therapeutic value. AI can tap into patient data such as EHR and RPM data to identify optimal care pathways personalized to the clinical need of the patients. Moreover, providers can enable preventive services for patients with chronic conditions to lower the likelihood of hospital re-admission resulting in optimal utilization of healthcare resources, including hospital staff.

Artificial intelligence or machine learning is also anticipated to play a key role in cost modelling by revealing prospects for cost reduction in specific diseases and treatments. Researchers have shown that 75% of all healthcare expenditure can be attributed to high-cost patients who make up less than 20% of all patients. Addressing high-cost patients could result in an immediate impact on the overall healthcare expenditure. Such machine learning-based cost prediction models can also be applied to specific disease treatment and cost estimation.

For example, adopting AI in radiology diagnostics has already started reducing the cost of diagnosis, early detection is helping to reduce the cost of recovery and prognostics. AI-enabled automation has been reducing the overall time for the overall process and thereby reducing the cost of treatment.

As more and more AI applications such as virtual nurse assistants, administrative workflow assistants, fraud detection, connected machines, image analysis, cybersecurity, etc. are getting adopted, it improves the prospects of maximizing their savings.

What are the challenges of AI in healthcare?

While its role is valuable in healthcare, AI has key obstacles to deploy and use in the actual world. The key challenge is to transition doctors from traditional decision-making to AI-supported decision-making. Healthcare software vendors need to collaborate with medical practitioners throughout the AI-based solution lifecycle to ensure the clinical validity of the solutions. Healthcare solution providers can also establish a feedback loop with the users to optimize their AI models for better efficiency and accuracy. Apart from improved accuracy of the model, it will also ensure easy integration in the clinical workflow. Additionally, medical practitioners will feel involved and empowered to use AI-based solutions.