The healthcare industry has undergone significant upheaval. Healthcare has been made better thanks to the wealth of data created nowadays, especially through Electronic Health Records (EHRs), pharmaceutical research, genomic sequencing, medical imaging, etc. Despite the abundance of Big Data that has been collected, the healthcare sector does not fully exploit it. Every year, more than 1.2 billion clinical documents are produced on average. However, there is a shortcoming in using it to enhance care.
With data insights for a healthier world, Apixio is here to improve the use of big data in healthcare. Apixio’s machine learning technology improves the decision-making process for greater healthcare quality. Additionally, the business incorporates artificial intelligence into its platform to convert diverse data into useful knowledge.
According to Darren Schulte, the CEO of Apixio, “Our technology examines unstructured and organised datasets to model patient care.” To obtain the best results, “we help providers and payers better undertake important operations including risk adjustment and quality reporting.” The Apixio data processing pipeline runs in a parallel computing environment for increased speed and scale. He continues, “In mere hours, we can process hundreds of millions of clinical notes.”
“Data Insights for a Healthier World”
Accessing healthcare data can be challenging without the proper tools. Your go-to tool is the Apixio AI platform. Our AI technology pulls and encrypts documents, photos, billing claims, and other data types from our customers’ source systems via a simple and secure extraction method, according to Darren. The data is subsequently transmitted to the Data Loader and ultimately to Apixio’s cloud-based platform for additional processing and analysis.
On the Apixio platform, the data is then sorted using a proprietary data specification. To make sure we have the essential components for customer projects, he adds, “We run hundreds of different data validation checks on imported datasets. The Apixio Data Coordinator then distributes files to various processing procedures based on their attributes. For instance, the optical character reader (OCR) pipeline is used to process medical photographs.
Once processed, the files are monitored on the Customer Data Inventory system from data import until insight creation. The Apixio Patient Object Model is then used to store the imported data (APOM). The APOM is similar to a phenotype in that each APOM contains details about a person’s medical history, illnesses, treatments received, biometrics, etc. The APOMs can be examined using ensembles, prediction models, and classifiers. All of these circumstances are combined when a person must make a choice in order to justify that choice.
Healthcare data put to good use!
All of it boils down to machine learning with AI integration. According to Darren, “Our platform uses a range of machine learning algorithms to extract a variety of signals from our data.” We employ ensembles to combine signals to develop insights, which are then packaged into customizable application workflows to enhance user decision-making.
Additionally, the user feedback obtained is once more saved in the APOMs and used in the future to retrain and refine the algorithms and machine learning. Our models are regularly updated by user annotation and data labelling (using automatic and manual methods),” says the CEO.
For further training of the models, supervised and unsupervised techniques are used. According to him, “our unique science infrastructure has methods built to deal with noisy annotations and labelling errors.” This is done specifically to set up workflow apps to reduce errors and increase accuracy. Apixio promises to continue to enforce this since they think it’s crucial for creating and sustaining high-performing algorithms.
Discover the Pioneer
Darren Schulte, CEO
Darren has over 14 years of expertise in the healthcare analytics and technology fields. Prior to being made CEO in 2014, Darren was Apixio’s Chief Medical Officer and President. Darren had executive leadership positions at Alere Health, Anvita Health, and Resolution Health prior to joining Apixio. Darren co-created 25 clinical indicators that were approved by the National Quality Forum to assess the quality of ambulatory treatment using electronic data. He is a well-known national speaker on healthcare analytics and quality enhancement.
Darren studied at the University of California, Berkeley for his B.S., Harvard for his M.P.P. and Stanford for his M.D. He completed his internal medicine residency at the University of California, San Francisco. At the Chinese American International School in San Francisco, Darren is a board trustee. Co-author on two US patents, he.
