$3.1 million NIH grant funds AI-generated pediatric head imaging database

The University of Chicago has received a $3.1 million grant from the National Institutes of Health to create innovative AI tools to generate a database of pediatric head images that will advance research on cranial development while protecting patient privacy.

The project will be led by Antonio Porras, PhD, Associate Professor of Neurological Surgery, and his team, in partnership with Children's National Hospital in Washington, DC, Children's Hospital Los Angeles and the University of Southern California, and Children's Hospital Colorado and the University of Colorado Anschutz Medical Campus. The research team will utilize one of the largest curated datasets of clinical images from pediatric patients and develop new generative AI tools to create realistic synthetic images at all ages and stages of cranial development.

Hospitals cannot always share craniofacial images from real patients with other institutions for research purposes because of concerns about patient privacy. The project will democratize the study of cranial development by developing a publicly available synthetic data platform, with representative images of both healthy children and those with craniosynostosis, a condition where sutures in an infant's skull fuse prematurely before the brain is fully formed.

We have not been able to share this data in the past, but our advances in new pediatric-centric AI methodologies are starting to allow us to generate synthetic datasets that do not compromise privacy. We are learning to synthesize scientifically accurate and realistic datasets so that others will be able to develop their own research projects to help advance our understanding about craniofacial development."

Antonio Porras, PhD, Associate Professor of Neurological Surgery, University of Chicago

A window into the patient's condition

The grant is one of the first projects underway in the newly launched Section of Computational Neurosurgery at UChicago that brings together clinical and scientific expertise in neurosurgery, artificial intelligence, and data science to develop new approaches to the diagnosis and treatment of neurological disease.

Porras is scientific director of the section, the first of its kind in the nation, in partnership with clinical director Rohaid Ali, MD, Assistant Professor of Neurological Surgery. They collaborate with neurosurgeons, computer scientists, engineers, data scientists and other researchers across the university to translate computational methods into clinical care and scientific discovery. 

"As we look at where neurosurgery and neuroscience are heading, AI and computational capabilities are important tools that advance our treatment of pediatric and adult diseases of the brain, spine, and peripheral nerves," said Mohamad Bydon, MD, Ginni & Mark Rometty Chair of the Department of Neurological Surgery and Stahl Professor of Neuroscience. "Images are a window into a patient's condition. This project creates a paradigm that enables us to study and predict outcomes for our patients and ultimately have more effective surgeries and treatments."

Translatable potential across medicine

Surgeons can treat craniosynostosis with procedures to reopen prematurely fused skull bones in infants, giving the brain enough room to develop normally. Such procedures require careful planning, assessing the current shape and structure of the skull and then projecting how it will develop after surgery.

Because children are growing so quickly at this age, their skulls may change even in the few weeks between initial imaging scans and when they can have surgery. Access to a robust dataset of how a child at that same stage is expected to develop will allow surgeons at more hospitals, large and small, to better prepare for procedures by comparing their patients to reference data without having to take additional images and expose kids to more radiation unnecessarily.

"Not everyone works at a large, high-volume academic medical center or has access to experienced surgeons who have done thousands of these procedures. So, these datasets are going to help guide surgeons everywhere to perform procedures in the most effective way possible," Ali said. "I think the implications are hugely beneficial for all medical fields, not just neurosurgery. The creation of reliable, synthetic data based on imaging is going to have a translatable potential across a whole host of surgical and disease pathologies and allow more people to learn from others' experience."

Planning for the right thing

Porras said the first goal of the project is to develop a complete dataset of images for healthy children and next build a repository of images for patients with craniosynostosis, both before and after surgery. The team will then develop a publicly available data-sharing platform with a graphical interface that other surgeons can use to evaluate their own patients, without the need for specialized technical data skills that may not be available at all institutions.

"Now we will be able to objectively and quantitatively evaluate the specific anomalies of these children when they are diagnosed, and forecast how they will develop," Porras said. "This will allow surgeons to evaluate the differences and similarities between the outcomes from different surgical approaches, so for the first time, we can help them plan for the optimal thing to do for each patient."

Comments

The opinions expressed here are the views of the writer and do not necessarily reflect the views and opinions of News Medical.
Post a new comment
Post

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.

You might also like...
Leicester's cardiovascular research hub shines spotlight on life-saving innovations