Microsoft Healthcare NExT

Initiative of Microsoft Healthcare NExT (New Experiences in Technology)

7 trends in health care

7 trends in health care of 2017

• "Effect of a tsunami" in terms of load of a health care system (expenses on maintenance of human health at the age of 65+ in 4-5 above, than on people are younger than 65 years) ^[1]
• Changes in age structure of the population, growth of chronic diseases create new long-term needs for rendering medical services
• It requires other profile and models of rendering medical services (telemedicine; permanent, long-term contacts with patients, especially elderly and suffering chronic diseases; Internet of Things, and so forth)
• IT solutions should be integrated with medical services within all system of interaction: houses, at work, in society, in the system of medical and social security.

2017

Announcement of Microsoft Healthcare NExT

On February 16, 2017 Microsoft announced several initiatives and solutions in health care. The software giant intends to use the cloud services and artificial intelligence technologies (AI) for the benefit of medical institutions and partners.

This program of Microsoft received the name Healthcare NExT. It will integrate research and development of medical products and also new model of strategic cooperation of corporation in the health sector. Microsoft will provide to partners resources for creation and implementation of artificial intelligence.

Microsoft will provide cloud services and artificial intelligence for the health care benefit

Within Healthcare NExT the Microsoft company announced several new projects:

HealthVault Insights

This research project will allow Microsoft partners to understand more deeply health of patients and to develop precise plans of treatment, using the last scientific achievements in the field of machine learning. It will be possible to connect fitness bracelets, scales with wireless communication to the HealthVault Insights platform and other devices, having provided remote condition monitoring of the patient.

Microsoft Genomics

The project develops the genomic analysis using the Azure platform and an ecosystem of partners in this area. Objectives:

• It is fast interpretation and the analysis of DNA in community with partners on the basis of high-performance cloud computing in Microsoft Azure
• The secondary, tertiary analysis and integration with electronic medical records
• Providing access to the cloud storage of data through specialized API
• Use of analytical opportunities of Microsoft Azure

Virtual interlocutor

Using artificial intelligence. This technology will allow Microsoft partners to create tools for patient care with smart voice actuation.

Project InnerEye

Software development using AI for planning of radiation therapy. The purpose of this project — to provide to meter men and beam therapists tools for creation of 3D models of an organism of the patient on the basis of medical images in several minutes instead of several hours required now.

• The MS Research project on software development using AI for planning of radiotheraphy
• Creation on the basis of standard pictures of KT of a 3D model of the sections affected with a cancer tumor within minutes (instead of several hours) for the subsequent treatment
• Provides faster and accurate classification of a disease and the best end result of radiotheraphy
• Does not require use of work of doctors of super-high qualification (economy of means)
• Cloud computing and algorithms of machine learning

Use of the algorithms approved in Microsoft Kinect

The same algorithms [are used [machine learning

Gallery of a demo templates

Gallery of a demo templates of solutions for health care using machine learning technologies.

• Prediction of duration of stay of the patient in hospital^[2]
• Binary classification of data for identification of breast cancer
• Prediction of breast cancer using Azure Machine Learning for data analysis
• Prediction of heart troubles using a data set from a repository the University of California in Irvine UCI Machine Learning repository

Architecture of on-premise of the solution on prediction of stay of the patient in hospital using SQL Server R Services

The Skype for Business templates for use in telemedicine

• Open Source templates for Skype for Business which allow partners, to clinics to create solutions for remote interaction of doctors among themselves and with patients in different situations and scenarios
• Different options of deployment are local servers, cloud scenarios or a hybrid, for compliance to requirements of scaling, security, reliability and a cost structure
• Consultations "doctor-doctor"
• Consultations
• Services of "second opinion"
• Consultations between the doctor and the patient

• All necessary instruments of communications 'near at hand' – a chat, a voice, video, a presence status and the location
• The spontaneous and planned conferences
• Support of video transmission in H.264 format, the joint work mode with materials and a possibility of data transmission from other systems in real time
• The web application of Skype for Business gives a connectivity from any place with a full range of functions of meetings available in the version of the client for the PC (Internet Explorer, Firefox, Chrome, Safari)
• SDK and API for developers (embedding in functionality of MIS)

Microsoft HoloLens

The wearable holographic computer - technology of augmented reality which allows to introduce the digital world to the real world.

Tools for creation of robots (bots)

For the benefit of health care 'Microsoft’s AI health chatbot technology

Different scenarios of use

• Primary sorting of patients on the basis of standard medical protocols or completely adapted
• Possibility of providing answers on privileges and the prices
• Possibility of complete setup and embedding in MISY and portals of interaction with patients

Intellectual opportunities of setup of dialog

• "Boxed" ready functionality
• Support dynamic interactive dialog settings
• Knowledge deployment from the previous contacts with the patient
• Use of information from HealthVault with data on the patient, including the used medicines given from wearable devices
• Use of industrial knowledge bases for generation of more relevant answers

I want that our research opportunities and hyper scalable cloud computing helped our partners to achieve great success in the world of health care — the head of the Healthcare NExT program Peter Lee said.[3]

Scenarios of use of a cloud platform of Microsoft in health care

Amounts of data and requirements to computing powers in health care

In a health care system 30% of world data in a digital format^[4] are generated^[5]

• Amount of data in a digital format in a health care system increases by 48% annually^[6]
• The enormous amounts of data connected with the analysis of human genome (genomics) for the purposes of "exact medicine"
• The person is "Big Data": billions of characters are coded in each of us
• The genetic code of one person is about 100 GB of information
• Today the analysis of DNA of the person costs about $1000 - it is about 2% of the cost of this procedure in 2010. It is possible thanks to high-performance cloud computing and data storage in a cloud. The technology becomes a mainstream thanks to falling of the price and availability
• Experts estimate that by 2025 genomes of 100 million people will be analyzed.
• Cloud computing – the ideal tool for such analysis
• Complete interpretation of this code is about 24 hours of calculations on rather powerful virtual machine (Azure D14 instance: 16 cores, 112 GB of RAM, 800 GB of local SSD memory and so forth)
• The amount of data which are generated in the field of genomics doubles each 7 months and will be there is in area a zettabyte soon (10 in 21 degrees of bytes) (~ 40 exabytes of data a year by 2025). ^[7]

Notes