CB2 Helps Researchers Transform Biomedical Data into Knowledge

UA scientists working with clinical data have an inhouse ally who can help them make the most of their biomedical research. The UA Center for Biomedical Informatics & Biostatistics, or CB2, transforms researchers’ data into knowledge, knowledge that benefits patients and the public.

When UA dermatologist Clara Curiel, MD, needed to harness data relating to tissue samples for future research to better understand the biology and treatment of skin cancer, she collaborated with CB2’s data scientists.  

Dr. Curiel was looking for a way to efficiently and confidentially link patient data with data from tissue specimens and images to advance the diagnosis, prognosis and treatment of skin cancers. According to the Centers for Disease Control and Prevention, skin cancer is the most common form of cancer in the United States. In Arizona alone, 2,504 melanomas of the skin were reported in 2020.

Dr. Curiel is now using the Patient Registry, Imaging Database and Tissue Bank (PRIT) to expand and elevate her research. PRIT connects three discrete software programs with one another to securely and efficiently manage data gleaned from her clinical research, thus facilitating prognosis and treatment of skin cancer. 

With the help of CB2 Director, Electronic Data Capture, Terry Smith, the team launched an IRB-approved patient registry. The registry houses vital demographic information about each study participant.

“CB2 assisted Curiel’s  team  to efficiently capture and organize that information, but more importantly, what information needs to be gathered about the disease and treatment, so that it becomes a template and standard,” says Nirav Merchant, former CB2 Interim Director. “If another researcher in Arizona or elsewhere is doing similar work, the underlying data includes the key data elements which are essential to perform large scale analysis.”

Dr. Curiel also needed to link clinical images of melanoma lesions with patient data contained in the registry. Vern Pilling, CB2 Director of Biomedical Informatics Services, created software automation to do just that by leveraging the imaging database, Omero. Omero, part of UA’s strategic initiative, fills a technological gap between the university and outside resources. 

“Lesions need to be automatically linked to patient information, specifically, the meta information about that patient,” says Merchant. “Omero figures out what meta data corresponds to a patient’s name. It’s smart. So, two systems that are not technically connected can go back and forth and exchange information, which makes the data more meaningful.” 

A third system, Open Specimen, a Biobanking Laboratory Management Information System managed by CB2, allows researchers to track patients’ physical specimens stored in a tissue bank, or biorepository, from collection to utilization. In essence, software, administered by Manuel Snyder, CB2 Associate Director, Electronic Data Capture, tracks what specimens come into the tissue bank and what specimens go out. 

“Open Specimen empowers researchers to efficiently use these tissue banks not just within UA but outside the university, as well,” says Merchant. “You can go to Open Specimen if you’re a researcher from somewhere else, and you can make a formal request for tissue to be used in your own lab. Open Specimen empowers researchers like Curiel to make the most of their data.”

Another part of empowering researchers like Curiel is making sure that they receive rigorous training in the use of CB2’s resources. Doug Cromey, co-manager of the RII Imaging Cores – Optical division at UA, heads up that endeavor.

“Investing in these things is one thing, but teaching people how to use it is another,” says Merchant. “Robust, hands-on training is necessary to teach people how to maximize these resources.”

But not only do people need hands-on training, says Merchant. So does AI. CB2 is making preparations to maximize data through machine learning readiness, that is, training computers to use copious, high-quality data to analyze data and build models.

“One way to train AI is to give computers information that is thorough and connected,” says Merchant. For example, in the case of skin cancer, if a computer is given copious images of lesions and corresponding background information, it uses that accumulated knowledge to yield robust diagnostic and prognostic information, such as treatment options, aggressiveness of the disease and staging.

“The linking of that data and consistency of that data make AI more powerful,” says Merchant.

Curiel and CB2 are working with other health centers with the aim of pooling their respective data to make the research and clinical work richer and more effective.

“If you don’t have a unit like CB2, you cannot empower people to do better science,” says Merchant. CB2 helps scientists do something above and beyond what they were able to do before.”