AI for better health

With the rapid evolution of Artificial Intelligence, many fields are witnessing the emergence of new business values. Likewise, Artificial Intelligence is increasing its outreach to improve our health. According to Accenture research, when combined, key clinical health AI applications can potentially create $150 billion in annual savings for the US healthcare economy by 2026 [1]. AI has upgraded many aspects of healthcare: Health Monitoring, Disease Diagnosis, Precision Medicine, Drug Development, and Patient Flow Management.

Health Monitoring

Personal health assistants are devices capable of monitoring vital health indicators. These devices, phones or watches, are equipped with medical sensors that monitor health indicators and notify users of any abnormal behavior. For instance, a personal health assistant can inform its user of his increasing heart rate during sleep, preventing imminent heart disease. Likewise, these assistants can provide long term reports to doctors, easing diagnosis [2]. 

Disease Diagnosis

Last year, we witnessed the genesis of a health disruptor. An AI system developed by Google Deep-Mind and medical institutions in the UK to diagnose breast cancer outperformed doctors in their diagnosis [3]. In the USA, this intelligent system demonstrated an absolute reduction of 5.7% in false positives, where patients can become frustrated while being healthy, and 9.4% in false negatives, where cancer goes undetected. Computer vision helped detect small tumors in cancer screening. The following figure shows an example of cancer prediction through Deep Learning:

(Example of cancer diagnosis through Deep Learning [4])

Today, AI can use any vital health indicator to diagnose other diseases than cancer.

Precision Medicine

Since data can be employed in personalization [5], treatment can be further personalized depending on health indicators, genes, and symptoms of each patient. Precision Medicine represents the field of medicine where treatment depends on each patient's conditions [6]. This field when combined with AI can enhance the treatment's personalization. Depending on patients’ health conditions, machine learning and deep learning models can be used to recommend the most suitable treatment.

Drug development

Drug development is a set of feedback-based experiences used to identify a certain compound that can be minimized through AI. Since drug development is based on analysis and deduction, AI models can detect patterns in trials’ results aiding biologists to evaluate compounds. A drug to treat obsessive-compulsive disorder required 12 months of AI development to get to human trials while casual drug development usually takes five years [7]. With AI, drugs can be industrialized in no time. 

Patient Flow Management

Patient flow encompasses the movements of patients between hospital facilities. If mismanaged, it can cause dangerous health issues and even increase the mortality rate inside the hospital. However, AI can automate the re-affectation of patients to other wards based on hospital data. Clinical administrators can also control their intakes better. Accordingly, patients’ engagement will increase.

Artificial Intelligence is powerful because of its applications. It is redesigning health care with innovations in diagnosis, drug development, and administration. However, AI hasn’t matured yet to be generalized in healthcare. As AI moves from R&D to practice, it raises key questions as to

whether trust intelligent models with human lives?

References

[1] “Artificial Intelligence: Healthcare's New Nervous System” —Accenture

[2] “Use Fitbit's New Wellness Report To Have A More Informed Conversation With Your Doctor” —Fitbit blog

[3] “International evaluation of an AI system for breast cancer screening” —Nature

[4] “Assisting Pathologists in Detecting Cancer with Deep Learning” —Google AI Blog

[5] “Why Data is the New Superpower?” —Oussama KIASSI

[6] “Powering Precision Medicine with Artificial Intelligence” —Intel

[7] “Artificial intelligence-created medicine to be used on humans for first time” —BBC News

Further Readings

“Artificial Intelligence for Enterprise” —Oussama KIASSI