SNMMI Mars Shot: Digital Twin/Data Science

Today, treatment of patients requires a cohesive team of practitioners working in concert. With a limited number of trained medical practitioners and allied health professionals facing a rapidly expanding pool of patients, research, and potential applications for novel therapies, the field of nuclear medicine is full of clinical data gaps and even greater learning opportunities. More research is needed to address the massive projected advances of data science and machine learning in the field of nuclear medicine and in the medical profession as a whole.

SOLUTION

In view of the rapid integration of the digital age into medical practice, collection and sharing of image and patient data to drive machine learning approaches will strengthen clinical capabilities¹. Advanced data science, including machine learning methods, is primed to progress the efforts of clinicians through use of a digital representation, or twin, of the patient’s anatomy and physiology. This digital twin is based on models whose parameters can be learned automatically from real or simulated medical images and additional clinical, biological, behavioral, and environmental data.

APPLICATION

Advanced algorithms can use learned parameters to serve the three pillars of digital medicine: computerized assistance to diagnosis, prognosis, and therapy². Digital twins enable learning and discovering knowledge, generating and testing hypotheses, and performing in silico (computer modeling or simulation) experiments and comparisons. They are poised to play a key future role in formulating highly personalized treatments and interventions³.

IMPACT, OUTCOMES & CHANGE

Continuing advances in imaging equipment and in reconstruction and analysis tools allow better visualization, quantification, and interpretation of images today, opening the door to deeper discoveries tomorrow. Digital twins offer a risk-free playground on which to explore innovations, strategize for possible futures, and test limitless what-if scenarios. Digital twins can connect the right data, artificial intelligence models, and human workers to explore possibilities, futures, and strategies⁴.


1. https://www.snmmi.org/NewsPublications/NewsDetail.aspx?ItemNumber=39568, 2. https://hal.inria.fr/hal-02063234/file/Ayache_invited_talk_MCA_FY2019.pdf,
3. https://www.mdpi.com/2075-4426/11/8/745/htm, 4. https://www.accenture.com/us-en/insights/health/mirrored-world