Thursday, July 2, 2020

Future Of AI in Healthcare

The post Future Of AI in Healthcare appeared first on healthcare nt sickcare™.

Future Of AI in Healthcare

Future Of AI in Healthcare – The AI’s capabilities in solving some of the major healthcare challenges are beyond doubt. There are encouraging stats and reports of AI delivering reliable performance and in many cases surpassing the capabilities of expert physicians as well. While the importance of the human medical experts remains paramount, it is the AI’s superb multitasking speed and computational accuracy (processing, analyzing, and prediction functions) along with the adaptive learning model that makes it a revolutionary tool for reforming healthcare sector.

The major concern, however, is how to seamlessly integrate it with the present healthcare at a wider level? Along with technical and social restrictions, there are tricky ethical and regulatory issues as well. In this blog we will mention the major obstacles that need an immediate response before we can implement AI on a wider scale across the globe:

Future Of Artificial Intelligence in Healthcare

The real power of AI lies in its sophisticated analytical, logical and reasoning capabilities that allow it to determine the most relevant solution out of several options and take independent decisions with zero or minimal help. 

Technological Boundaries

While AI is registering a remarkable constant growth, presently it is gaining strength in specific ranges only. The present AI systems can accurately perform specific enormous tasks that are simply impossible for human beings (even with the aid of traditional computing). However, we need to develop different algorithms for different tasks. The Deep Blue of IBM can defeat the world champions in chess but it cannot prepare a tea nor do banking activities- something that most of the average chess players can do with ease. 

 The present AI systems can study data sets and learn to identify the patterns and logically obtaining the problem-solving skills that can be applied to specific domains. However, it is not yet capable of transferring the acquired cognitive capacity to other domains. 

In simple words, unlike humans, the AI cannot manipulate the acquired knowledge and utilize it across different fields. It necessitates the development of different algorithms for different purposes- which is both effort-intensive and expensive. 

What we need is the domain-agnostic AI that can intuitively apply the obtained knowledge across different domains with uniform dexterity. 

Performance Potential Determined by Data Type

Also the AI feeds on the supplied data and thus its knowledge skills depend upon the volume, integrity, and wholesomeness of data. The risk of bias- whether intentional or unintentional- is also there. 

We can take the instance of radiology here. Numerous images are feed into AI to make it learn about the imaging patterns. One of the key primary steps is framework conceptualization for the algorithms which determine/influence the learning, functioning, and evolution cycle of AI. There are 2 limitations here. 

  • This foundational part of the process is exposed to the subjective supposition of the production team that automatically narrows its potential. 
  • Secondly, the developed algorithms’ forecasting functioning is based on past cases and so that might not be able to accurately tackle the future instance like resistance to or side effects of the treatment. 

Tricky Ethical Issues

Apart from other technical limitations the ethical issues are also there, which need to be immediately addressed before implementing and utilizing AI in healthcare on a wider scale. Unlike other devices that follow the instructions of the doctors the AI-based systems can evolve on their own and take independent decisions as well. 

Moreover, it is not always possible to conclude the logical process/reasoning behind a certain decision made by AI. 

  • The question is who would be held responsible if an AI algorithm takes the wrong decision, or provides an inaccurate diagnosis. The doctor would plan the treatment based on the same. In such a scenario who is then legally answerable for the safety of patients- the doctor, the AI-developer, or the (least possibility) AI itself? 
  • One possibility is to build protective boundaries around the system. But it brings to another concern- building security boundaries will defeat the very objective of using AI by limiting its independent evolution and can thus prevent us from realizing the full potential of AI.

Limited Availability of Digitized Data or Past Records

The wholesomeness of data, as mentioned earlier, is vitally important. However, it needs the data in a digitized format. 

The reality, however, is that majority of the healthcare facilities and medical practitioners in the resource-poor locations generally manually write the diagnosis and treatment notes. Such data is not preserved and it can cause critical gaps in data availability. Such information will thus be omitted while preparing AI systems for analyzing specific situations or forming preventive, diagnostic, or treatment suggestions. 

It can create an unintentional but critical bias towards a large number of populations and the irony is that this is the specific community (residents of under-resourced areas) where AI-systems can have a major positive impact

Healthcare Gains May Cause Economical Losses To Many

While the above paragraph is true to a large extent, the use of AI may certainly reduce the utility value of human workers. The established surgeons, doctors, and experts don’t have to worry at least for now but the real concern is regarding the workers like lab technicians. The concern here is that these issues are to be adequately tackled before implementing AI on a wider scale. We need to make sure that the profits gained on the healthcare front shouldn’t cause the loss of income source for the healthcare workers. 

The Haze and Disappointment Caused by Exaggerated Reality 

While the AI is certainly loaded with lots of innovative capabilities, the popular media (not the authentic sources) is creating hype around the subject because of their vested interests. It not only creates the misconception and subsequent possible disappointments in the future but may also misguide digital citizens

There already is a big confusion regarding the definition, limitations, capabilities, and the process of AI. Moreover, the hype over the topic makes things hazier. There are some media reports of AI ruling over the people while others talk about the doctors being redundant once the AI is adopted in the healthcare sector on a wider scale. It will possibly prevent many physicians to lend their active support in developing AI systems. 

The possibility of the same multiplies when we talk in the context of developing or poor nations/localities where many physicians and primary healthcare professionals are not technically proficient or updated with the IT capabilities. If they Google the term and come around such misleading information like AI snatching away the doctors’ jobs- they can be discouraged  to share their data or prescriptions that are critically important for the wholesome progress of AI-based healthcare ecosystem.


I would like to introduce myself as Jitendra Bhojwani, A tech writer and blogger currently exploring new opportunities. Published articles: How Can Blockchain Technology Improve Cloud Storage?

You May Also Interest To Read; Preventing Prediabetes To Becoming Diabetes

All material copyright healthcare nt sickcare. 2017 – 2020. Terms and conditions & Privacy Policy of use. The contents herein are for informational purposes only. Always seek the advice of your physician or other qualified health providers with any questions you may have regarding a medical condition. Source: This article inspired from various online articles and our own offline experiences. The content meant for public awareness and regular post to the clientele of healthcare nt sickcare.


The post Future Of AI in Healthcare appeared first on healthcare nt sickcare™.



Post a Comment

Please give your feedback for this post