From the clinic to imaging, the sense of AI for healthcare
[ad_1]
Just 5.80 seconds. It is the time taken, on average, by an artificial intelligence platform to perform a clinical audit, compared to the 10 days required for a human operator to achieve the same result. Numbers provided in 2020 by the G20 Task Force on the digital economy and the impacts of AI and which, given the recent developments in information technology, risk further widening the gap. And yet, man and machine can still get along in a healthcare scenario where cuts and reductions put at risk not only the punctuality of assistance but also the quality of the service.
“The potential of artificial intelligence in healthcare is enormous,” underlines Andrea Celli, managing director of Philips Italy, Israel and Greece. «The solutions on the market make it possible to obtain strategic information in an ever faster and more precise way, for predictive and more personalized management». This is the case of Philips SmartSpeed, which applies AI to magnetic resonance (MRI) instruments. The platform is based on two AI radiology technologies, Compressed Sense and Adaptive-Cs-Net, which together speed up scan time, without compromising image quality. SmartSpeed aims to deliver fast, high-quality imaging for a broad range of patients, from those experiencing pain to those who have difficulty holding their breath during an exam or who wear MR Conditional implants. But it’s not just technology: «For Philips, the value of Ai is based on human experience, on the possibility of combining the power of algorithms with profound clinical knowledge. Solutions are born to be integrated into the workflows of healthcare professionals and in the patient care path, without difficulty, indeed with extreme naturalness. Technological innovation in healthcare can be humane and sustainable, with people at the centre”.
The Sienese hospital-university company demonstrates that Ai tools are already a reality, even in Italy. Here, AI is used for the improvement of clinical-diagnostic pathways, in particular for the analysis of images obtained with MRI and CT (computed tomography). The end? “Evaluating the extent of strokes and making treatment decisions in a more timely and effective way,” says Antonio Davide Barretta, general manager of the company. “In the case of acute ischemic stroke, with the help of AI, clinicians can view information on location, extent and severity, fundamental factors for decision-making and to ensure the best therapy”.
The attention to innovation is confirmed by the local Saihub network, Siena Artificial Intelligence Hub, a design glue between companies, the world of research and centers of excellence, and by the presence of the hospital company in the Aioti alliance, a European network that brings together entities involved in the development and implementation of the Internet of Things and AI also in the healthcare field. Moreover, the hospital has a study project on Rett syndrome, also based on the use of AI to predict the onset of symptoms and the associated phenotypic variability.
But if the current trend is represented by chatbots, healthcare is not exempt. Microsoft, which has invested heavily in OpenAi, has cobbled together various AI tools to improve specific scenarios, such as understanding analytics trends and spotting inconsistencies over time. There are four pillars of this vision of “intelligent healthcare”: the Gpt large language model, the ChatGpt chatbot, the Codex programming language and the Dall-E 2 platform, already famous for being the engine that drives Bing Image Creator, website that everyone can use to create graphic works starting from a few textual indications. By putting these elements together, with the first three integrated into the recent Gpt-4, Microsoft manages not only to feed its algorithms with hundreds of scientific publications, which form the background for AI training activities, but also to speed up patient monitoring, with simple chat windows where digital assistants provide tips, guidance and best practices to follow. Thus the “augmented” follow-up does not detract from the doctor’s work, on the contrary it enhances it, letting intuitions be guided by data.
[ad_2]
Source link
