By Rose Higgins and Mimi Huizinga, MD, MPH, FACP
Advances in genomics have in recent years accelerated the personalization of medicine, but that was just the beginning.
As researchers have mapped the human genome, clinicians have gained understanding into the effects that even minor abnormalities in the genetic structure can have on patients’ conditions. These insights have enabled providers to make more informed decisions about which treatment options to select over others even in instances when two patients share similar characteristics, such as age, gender, weight, ethnicity, and lifestyle.
Nonetheless – even with this wealth of new genomic data – it has remained a challenge for clinicians to understand how effective treatment plans are overall, as well as gather real-time insights into how a treatment plan is affecting a patient at any given moment – for cancer patients, in particular. The reason is that, despite genetic similarities between different patients and their tumors or lesions, patients often react differently to the same therapy. As a result, clinicians in many instances do not definitively understand the effectiveness of a course of treatment until it is well underway or completed.
Now, however, as a result of advances in radiomics, or advanced imaging analytics, clinicians have the opportunity to gain these insights earlier in the cancer treatment process. The phenotypic data generated by radiomics, when combined with artificial intelligence (AI), provides data concerning the biology of tumors and lesions, enabling researchers and clinicians to accurately predict how a specific tumor or lesion will react to various treatment options, helping to guide the treatment-planning decision process.
Leveraging AI to see below the surface of traditional images
AI-driven analytics are leveraged to extract meaningful, previously unobtainable data from traditional imaging modalities such as CT, MRI, or PET scans. Radiomics solutions then curate this quantitative data to present clinicians with a plethora of data that cannot be gathered from an image via traditional optical analysis.