In the grand scheme of things, it’s about incorporating AI into every procedure of a surgery/ treatment.
By Srinath Srinivasan
Recently, Siemens Healthineers, the healthcare division of Siemens, invested Rs 200 million in a manufacturing facility in Bengaluru to integrate R&D and manufacturing. The company says it will put together state-of-the-artCT scan machines and C-arm radiology systems for emerging markets such as India, and provide employment locally. While there are business gains certainly, the company is headed towards having Artificial Intelligence (AI) as one of the primary technology pillars in their products. As a market leader, it aims to be realistic in its approach to incorporate AI in its products. It says AI will pan out in phases, with hybrid devices coming to the market first.
As of now, there are certain use cases where AI is involved and these are mostly in the backend. According to Peter Seitz, executive vice president – surgery, Siemens Healthineers, there are a few stages before the technology could get predictive . “Today the technology is at the level of visualisation. For instance, we can do a real time temperature mapping with an MRI in ablation procedures. We can use this to eliminate the perfusion of tumor. In the long term, it’s about warning. For instance, an algorithm which monitors the anatomy can automatically segment various parts of the body and warn us that we would end up operating the wrong part if we continue what we are doing,” he says.
In the grand scheme of things, it’s about incorporating AI into every procedure of a surgery/ treatment. Currently, in cardiology, AI is used to predict the intensity of scar tissues that are induced during ablation procedures for treating atrial fibrillation/ arrhythmia. This requires a well-trained AI system which can take into account the various aspects of the anatomy and severity of the arrthymia in real-time.
The idea of AI seeping into every procedure raises the question of machine autonomy and whether a surgeon’s competency would be compromised. “If one of our machines today were to perform a procedure autonomously, we won’t be able to get this product cleared as a medical device. Today, the responsibility for the outcome of a procedure rests with doctors,” says Seitz.
Seitz believes that there will be devices which will increasingly assist doctors in making decisions and if there were to be a tipping point where machines could perform such procedures, then the responsibility would fall on to the machine’s manufacturers. “I don’t see this yet. Today, the best surgeon out there is the one who has the most experience and is agile with his hands,” asserts Seitz. However, he thinks that the procedures could be mechanised and their control could rest with surgeons. “One thing we could see coming in early is, surgeons may no longer manually do all the procedures but can make robots do them. AI and robotics can precisely enable this,” he adds.
This changes the dynamics of employment in a medical ecosystem. A comparitively young, digital native person who can control machines precisely and has the domain knowledge may take the place of the most experienced surgeon at some point.
With new tech, new business models and challenges come together. When the machines manufactured and deployed in India start communicating with a global network, data becomes the currency of this communication. “We have an installed base of 6,00,000 equipment around the world. It is not correct to think that the data collected from these can be fed into an AI system. We need the permission of the hospitals who own these equipment to even begin research activities,” says Seitz.
AI will enable the devices to handle data and secure them. Hospitals will be able to rethink their revenue channels with the data. The centre in Bengaluru is also vested in training software engineers and houses 1900 R&D personnel. It has developed a bleeding edge imaging platform called Syngo that can construct high-resolution life-like images of human anatomy and has over 2,00,000 deployments globally.