#Futureshaper

Shikha's journey to the "patient twin"

Computer scientist Shikha Chaganti is doing research work in Princeton, New Jersey, to develop a smartphone app that will enable users to generate a "health avatar" of themselves – created by artificial intelligence. Is this the next step towards the visionary "patient twin"? Learn more in part four of our #Futureshaper series.
7min
Katja Gäbelein
Published on June 14, 2022

A miniature digital Shikha waves at us from the smartphone screen. Though the figure's movements still appear a bit shaky, its visual likeness to the "real" Shikha is unmistakable.   

All the while, the "real" Shikha is moving her scrutinizing gaze back and forth from the smartphone displaying the avatar to the screen of her computer, where programming codes and data evaluations form cryptic-looking shapes. Shikha Chaganti is a research and technology manager on the Digital Technology & Innovation team at the Siemens Healthineers location in Princeton, New Jersey, in the United States. Born in India 34 years ago, Shikha joined Siemens Healthineers in 2019 shortly after completing her doctorate in computer science from Vanderbilt University in Nashville, Tennessee, specializing in medical imaging analysis.

#Futureshaper: Shikha's journey to the "patient twin" - Shikha portrait

The team in Princeton is focusing on artificial intelligence[1]. Scientist Shikha Chaganti has been leading the research project "My Digital Twin" since early 2021. The ambitious goal is to create a mobile phone app that will bring us one step closer to the vision of a digital patient twin.

"Artificial intelligence denotes the capability of computer systems to solve bespoke tasks which, due to their complexity, have until now required human capabilities."

So, why do we need an app like this? And why is patient twinning a technology of the future worth striving for in the first place? 

“In the current system, the patient is often the keeper of all medical data”, explains Shikha in her #Futureshaper interview: "If I'm a patient experiencing some symptoms, I often have to visit several specialists until I get the correct diagnosis. I have to carry reams of paper or CDs with test results and notes from visit to visit. I have to remember the exact symptoms I’m experiencing, when I first noticed them, and pass on the test results that any previous specialist ordered along with their referrals." 

This situation harbors risks, because only physicians with all the relevant information at their disposal are able to make an accurate diagnosis and initiate the right therapeutic measures: “In the worst-case scenario, the patients might not receive the therapy they need on time, which can have far-reaching consequences.”

Whether they are technical developers, creative business managers, or product designers: in our #Futureshaper series, we introduce you to colleagues whose innovative ideas are contributing to pioneer breakthroughs in healthcare.
Read more
#Futureshaper: Shikha's journey to the "patient twin"
The expression on Shikha's face turns very serious when she talks about this. Sadly, in her own private life she has experienced exactly this type of situation, where the correct diagnosis and therapy came too late for the person to be cured. This is just one more reason why she is so strongly committed to developing the app.
Health-related data are collected today – and in real time, too, for example with the help of smartwatches, as Shikha explains: "How active were you today, how many steps have you taken? What's your heart rate?" However, these data remain isolated in their own separate data silos, like many other data, from medical imaging or laboratory tests, for example. Up to now, such data haven't been stored in one central place to enable comprehensive insights into a person's current state of health.

The idea is to establish frictionless means to gather such data in future, so that information from different sources can be correlated to find larger patterns. One day, a person's complete health and wellness data could be collected in that person's digital patient model on the app. The data would then be available anywhere at any time – in emergency situations, too, when speed is of the essence – and could be continuously updated.
The term data silo refers to data repositories stored at different locations to which only specific organizational units or users have access.
On the user interface at the frontend of the app, a digital avatar represents the given person. The avatar is generated by artificial intelligence based on a simple selfie. This digital twin, which looks identical to the "real" person, is more than just a gimmick. "We want the app users to be able to identify with their digital twin, which helps them with consistent usage of the application," notes Shikha, explaining its conceptual intent.
A people-centric innovation method that serves to solve problems and develop new ideas. Interdisciplinary teams pursue their work based on a multi-stage creative process. The goal is to find the best solution for users.
“The patient model can utilize every bit of healthcare data that is available to patients so that they can have personalized care in their health journey,” says Shikha: The app backend will connect with various data silos to create the most complete and comprehensive picture possible of a person's health and wellness data. This will include for example connecting with a person's smartwatch, which will deliver information on their physical activity and vital signs. The app will also support the FHIR®[2] standard. This will allow the app to connect with PACS in hospitals to obtain medical imaging data on the patient.
FHIR stands for "Fast Healthcare Interoperability Resources", a standard that supports data exchange between different software systems within the healthcare system.

This should likewise enable the app to connect with EHR servers so that data from previous examinations, the patient's medication, prior laboratory tests, etc. can also be incorporated into the digital patient. What's more, users will be able to enter their own personal data in the app, like a kind of daily "health diary" recording their current state of health and any symptoms that arise.

EHR stands for "Electronic Health Record". It's a database for storing patients' health data, including their prior medical history, treatments, medications, and allergies, etc.
Using her index finger, Shikha enters new weight data for the avatar in the app's settings – and the AI promptly takes charge of making the avatar look like it has gained a few kilos. The avatar can be modified on the user interface to reflect the real person. The patient model also "grows" continuously in the background, fed by user data from the digital world. The user's personalized digital twin replica is updated with every examination and every new medical dataset. So, every relevant piece of information is dynamically updated right along his or her health journey.
Research work on the app, however, involves much more than just collecting and storing data. It's about deriving new insights from the data for the user so that, if need be, specific medical actions can be taken based on these findings. Artificial intelligence plays the key role here by organizing the data, says Shikha: "It could, for example, intelligently collate and summarize all the medical data, and clearly display the entire history of a patient's illness."
AI will also help to "translate" complex medical information into a visual language the app users can easily understand: "We plan to create patient-friendly visualizations of various types of health data. For example, we could show cinematic rendering of their internal anatomy, which is generated by AI”, explains the scientist.

A new type of 3D visualization technology. It uses data from imaging systems to create photorealistic pictures and videos. It shows the inside of the body with such unprecedented clarity that patients can also understand the images.

Read more
Shikha then leaves for her next meeting. As project manager, she has regular video calls with her international core team to discuss the project status, and to decide on the next steps for developing the app. The division of labor is complex, entailing collaboration for example with internal company experts in medical image analysis, machine learning, software engineering, and natural language processing.
Natural language processing is a subfield of linguistics, computer science, and research in artificial intelligence. Its aim is to facilitate direct communication between humans and computers based on natural human language.
#Futureshaper: Shikha's journey to the "patient twin" - A comprehensive network of experts
Additional specialists are also involved, depending on which aspects of development need addressing, such as the in-house Data Privacy Office that focuses on issues of data protection and privacy. After all, sensitive personal data of all future app users must be securely protected everywhere and at all times. Patent lawyers from the Intellectual Property department are also closely integrated in the process, dedicated to safeguarding intellectual property rights. This is a complex task when it comes to digital twin concepts:
Shikha and her team communicate regularly with both customers and external business partners, because everything developed to create their app must later make good sense in a day-to-day medical context, and meet people's real-world needs and expectations. As Shikha explains, "We meet, for example, with doctors from hospitals and medical centers to talk to them about the technical aspects of our project, and about our shared visions of patient twinning." The opportunities for further developing this concept, which has deliberately been conceived on an open basis, are vast. And customer interest in patient twin solutions is enormous: "Many people agree that this is the future."

Are you also interested in joining our passionate team?

Visit our career portal

Here you'll find an article about the initial successes of digital twinning and challenges on the road to creating digital patient twins.

Many medical technology experts consider patient twinning to be the key to personalized precision medicine. Patients suffering from a wide range of illnesses stand to benefit from such solutions. Supported by artificial intelligence, users can become their own "health managers". Patient twinning solutions could function just as well in rural areas as they do in cities, making the provision of relevant medical information less dependent on place and time. The information comprehensively sorted, processed, and neatly displayed by AI could help physicians make the right diagnosis, and save valuable time. Ideally, it could play a role in preventing users from becoming patients in the first place:
#Futureshaper: Shikha's journey to the "patient twin"
"This is perhaps the greatest benefit that the patient twinning technology offers: by taking effective preventive action, it could help prevent people from falling seriously ill in the first place," says Shikha. After all, continuous monitoring of a user's state of health allows timely intervention if their condition worsens. It is not possible to predict exactly when a solution like the app Shikha Chaganti and her team are developing will be ready for the market. Yet, one day, a digital Shikha like the one waving to us just now rather shakily from her smartphone screen could perhaps help the "real" Shikha get a faster and more precise diagnosis. And that could be of decisive importance to the real Shikha.

By Katja Gäbelein

Katja Gäbelein works as an editor in corporate communications at Siemens Healthineers, and specializes in technology and innovation topics. She writes for text and film media. 

Assistant editor: Guadalupe Sanchez