- 9th August 2018
- Posted by: Manolis
- Category: Blockchain
Clinical data is a goldmine for healthcare providers. Analyzing large datasets of medical data can provide deep insights and improve outcomes across the healthcare value chain, from research to diagnosis to patient management. However, managing years of complex clinical data has been an enormous challenge for both clinicians and analysts alike. Moreover, the healthcare industryis currently reeling under a scarcity of physicians, increasing healthcare costs and explosion in health-related data. This scenario warrants rapid technological transformation with Applied Artificial Intelligence (AI) poised to become a key enabler of this change.
With the volume of data availability growing at an astounding rate along with evolving patient’s demands from their providers, the healthcare domain is ripe for a better change. In such a scenario, AI is all set to penetrate every segment of the healthcare industry. There is no question that there is a need to improve process efficiency and optimize overall cost. Furthermore, the demand for huge data processing and efficient service delivery has actually put Artificial Intelligence(AI)-enabled solutions at the forefront of the healthcare revolution.
Let’s dive deep into some of the key trends of AI:
#1. Empowering Value-Based Healthcare
The effect of data collaboration and adoption of AI among payers and providers have resulted in the latter sharing clinical data with the former. This has led to tighter coordination between care management, case management, and utilization management. A range of AI applications used in a collaborative way is benefitting the business outcomes of the payers and providers alike.
Payers are mostly using AI-based solutions to cut down unnecessary ER visits for patients and detect fraudulent insurance claims. For instance, AccuHealth, a Chile-based startup has come up with a patient-based healthcare model, which focuses on home-based remote care rather than reactionary based care. This has resulted in 42% decline in emergency room visits leading to 30% savings for participating insurance companies.
Hospitals and healthcare clinics use AI for various applications with a maximum focus on disease diagnostics and customer-facing apps. Medalogix for instance, helped Alternate Solutions HomeCare to reduce the frequency of readmission of patients after completing the treatment.
AI-powered health IT tools are expected to reduce the time and complexity involved in data processing. We can also expect future AI solutions to leverage social media information – such as health communities, and information about patients vitals form wearable device to achieve deeper health treatment.
#2. Migration from Doctor-Centric Diagnostics to Data-Driven Diagnostics
The high cost involved in diagnostic processes, over-testing of low-risk patients, and limited capability for early detection of chronic diseases are driving the adoption of AI in diagnostics. The good news is, AI is bridging the gap between doctor-centric diagnosis and AI-based diagnosis through pattern recognition and machine learning technologies. Oncology, for instance, is the most prominent field for developing AI-based diagnostic solutions. In fact, the maximum number of trial runs has been carried out in this field itself. When it comes to diagnosing Retinopathy (causes childhood blindness) through analyzing images of infant’s eyes, the doctor’s diagnostic accuracy can be 82% but the use of AI can raise the level of accuracy to 91%. According to Nathan Buckbinder, Co-Founder & VP Operations, Proscia Inc, “There are many hidden patterns in a patient’s cancer data that can impact the entire course of diagnosis and treatment and which can benefit from the use of AI and Machine Learning.”
#3. Transformation of Drug Discovery in Pharma Industry
Increasing drug development time and cost is driving pharma companies to embrace digital transformation. According to Deloitte, in 2017, 12 largest biopharma companies received just 3.2% returns from their drug-research units, as against 10% in 2010. The use of AI in drug discovery is expected to expedite the overall process. Applied intelligence can improve drug discovery success rates by 8-10%, resulting in savings worth billions of dollars for the industry. Finding drug discovery compounds and precision medicine are the major trending applications over other pharma based AI applications.
In fact, pilot projects on AI-based drug discovery tools have demonstrated a significant reduction in development time. For instance, IBM Watson’s Clinical Trial Matching (CTM) eliminates the need to manually compare clinical trial enrollment criteria. CTM uses AI to read through patient medical data and match the right patient to the right clinical study. This has reduced the prescreening wait time by 78% during the trial period (16 weeks) and automatically eliminated 94% of patients who failed to meet clinical trial requirements.
Increased availability of patient and medical data sources is paving the way for healthcare providers to design custom medicines, identify early signs of critical diseases and most importantly, ensure individual care to patients. On the other hand, there are challenges that are cropping up with regard to data source integration, talent availability and limited implementation of AI systems. Going forward by 2025, it is expected the bright spot- AI in this domain will be involved from population health management to digital avatars capable of answering specific patient queries.