Regional |HM Hospitals and Microsoft apply artificial intelligence in cancer research

After the firm last March of the collaboration agreement between HM Hospitals and Microsoft to apply data science in the field of health and thus advance in the so -called 'Precision Medicine', the Alliance has taken important steps in the incorporation of technologiesof artificial intelligence (AI) to promote the personalization of health care, especially in the field of oncology research.

On the one hand, AI is being applied in the analysis of diagnostic images and in the automation of data obtaining from the medical records with a double objective: guide clinical activity and support research, accelerating the development of new proceduresand medications.On the other, the use of conversational interfaces in the training and training of patient care professionals to optimize this service and raise the quality of interactions is being evaluated.

A new comprehensive medical check -up service at HM Puerta del Sur, a key prevention tool

Una de las herramientas disponibles para fomentar un óptimo cuidado de la salud y la detección a tiempo de muchas patologías reside en la necesidad de someterse con cierta regularidad a un chequeo médico, que se revela como el arma clave para fomentar la prevención.

An immense data volume requires advanced technology

“The amount of information that is generated daily in HM hospitals is huge: the notes of the doctors taken during the consultations, which reflect a history, diagnoses, plans and evolutionary;The images generated in the diagnostic tests, as well as the reports prepared by the different experts about these ... all those clinical data of the patients, who were already addressed and treated by HM hospitals, are evaluated in a new way with the collaboration of Microsoft,which provides us with a new structure to advance in the creation of predictive models, which in the future will help us in the diagnosis and treatment of patients, ”says DR.José María Castellano, scientific director of the HM Hospitals Research Foundation.

REGIONAL | HM Hospitales y Microsoft aplican la Inteligencia Artificial en la investigación contra el cáncer

Within the framework of the collaboration agreement between HM Hospitals and Microsoft, the medical staff is facilitated to access advanced and automated analytical abilities supported by artificial intelligence.Recently, the president of Microsoft in Spain, Alberto Granados, was able to verify in situ how group professionals use this technology.During his visit to the Clara Campal Integral Center (HM CiocC Madrid), located at the HM Sanchinarro University Hospital, he was accompanied by the president of HM Hospitals, the DR.Juan Abarca Cidón, and part of the management team.

The implementation of these models will serve to advance the development of personalized medicine.

AI helps configure the prognosis of cancer patients

New resources built on Microsoft technology can be used in many scenarios.One of the first approached by HM Hospitals has been in the field of oncology.HM CiocC Madrid researchers have finished a project that allows the automatic containe.This is done automatically thanks to an algorithm based on the analysis made by the Inner Eye application of the previous images for treatment planning, which increases precision and saves time to specialists.

Also, applying the latest advances of Microsoft Research in medical image analysis, models are being created that enable automatic identification of structures within CT images and help perform a scale analysis.Based on the results harvested to date - superior to 88% -, it is estimated that it will expedite, among others, the planning work in the area of radiotherapy oncology.

In parallel, the construction of patient databases with rectum adenocarcinoma included in research projects has been automated.Through the training of the Text Analytics for Health application in the identification and capture of the variables of interest in the medical history, the acquisition of data is automated and should only be supervised by the researcher.

Thanks to the processing of natural language through the use of the most innovative Deep Learning techniques, a solution has been implemented that analyzes all available information and extracts the main concepts (symptoms, diagnoses, tests, treatments, genetic information...) Automatically, identified in a contextual way.This will be very useful in any other research or welfare context.

Currently, HM hospitals and Microsoft experts work together in the design of a new project for the identification through AI and predictive variables of a complete clinic response prior to the application of treatment to these patients.

The Integral Center for Neurosciences, located at the Puerta del Sur hospital, world reference in the treatment of Parkinson's

La subtalamotomía mediante HIFU es una técnica eficaz y segura para el tratamiento de las manifestaciones motoras de esta patología y consiste en la realización de una termoablación.

Training of care professionals

In a patient -centered organization, care responsible are key.With the purpose of establishing homogeneous and precise systems that favor rapid training of these profiles, HM Hospitals and Microsoft have initiated a pilot project that aims to evaluate the use of conversational support agents.These have been developed with QNA Maker of Microsoft, which uses Azure cognitive services for the generation of bots without the need to implement code or undertake complex processes.They will be fed by the most common questions and answers database, so that professionals can use this tool to raise issues that arise in the performance of their functions.

A first test has been carried out within the quality area with very satisfactory results, so it is planned to extend it to more areas during the next quarter.From the beginning, with this alliance, HM Hospitals sought to continue its strategic action in the management and control of patient health data.Both are managing to take a great step forward in the implementation of artificial intelligence models, which will serve to advance the development of personalized medicine.