As per a report released by Centers for Diseases Control and Prevention, the suicide rates in all states but one has sharply risen. One of the main causes for people to take such an extreme step is depression. Suicide has become the 10th leading cause of death in US and the second-leading cause of death for people in the age-group of 15-34. Major depression is the biggest reason for mental disorder in US.
People from all age-groups can be impacted due to depression. Older adults, children, teens, professionals at work and women are all susceptible to depression.
Image Source: https://cdn.pixabay.com/photo/2014/10/27/22/03/sad-505857_960_720.jpg
As per a recent article in techemergence, depression is affecting 16 million Americans on an average. As per WHO (World Health Organization), the yearly global economic bearing of depression stands at $1 trillion. It will be a leading disability by 2020. Medical practitioners and data scientists are working closely to enable AI and AI-enabled technologies to deliver value for mental diseases.
To that effect, Takeda, the largest pharmaceutical company in Japan, in a joint effort with ConvergeHEALTH by Deloitte, a research and development data science institute are putting in efforts towards better understanding effective therapies for difficult diseases like depression.
They are using insurance claims records to draw information like diagnoses, prescriptions and medical practices used to treat depression. There were several datasets run like linear and non-linear with the intention to identify data factors that had the highest outcome in forecasting patient results. As a result, they realized that if the right question were combined with the right responses, then the deep learning model and predictability improved.
In case of a complex disease like depression, the patient must be given several medications before finding out the one that works. The idea is to study the patient trajectory, their treatment journey and investing time to understand the patients that might be resistant to any treatment. The predictive model, after analyzing the patients’ historical records, helps to identify the benefit of switching a medication or perhaps using it as a first-line treatment. It will enable a shorter time to serve the right medicine to the patient. The models use different methodologies to enhance the predictive power to help patients suffering from depression to find a viable, fast and effective treatment.
For doctors, it is an impossible task to study the various facets involved in medical records. Deep learning method, on the other hand, can predict medication switches and identify patterns. The combination of AI, deep learning and machine learning adds more accuracy to a future prediction.
Amazon Web Services or AWS, which includes GPU on demand servers, was used to build and train these models. Deloitte’s Deep Miner tools and Amazon’s SageMaker tools were used to create pipelines and execute machine learning models. Earlier running machine learning analysis on large data sets was a costly and time intensive task. However, analytical tools, ease of scaling up infrastructure and data availability has made data science experiments economical and it’s possible to execute them within a short time.
If you wish to leverage the power of AWS in your enterprise, our cloud experts will be happy to understand your business requirement and discuss how AWS best practices, resources and strategy can boost your business growth.