archyworldys

AI, the big push in breast cancer screening

Artificial intelligence as support for radiologists in breast cancer screening would allow a greater number of cancers to be detected, which would mean a better prognosis for the patient and the use of less aggressive treatments, according to Dr. Esperanza Elías, a radiologist specializing in Intelligence Artificial applied to screening …

Artificial intelligence as support for radiologists in breast cancer screening would allow a greater number of cancers to be detected, which would mean a better prognosis for the patient and the use of less aggressive treatments, according to the Dr. Esperanza Elias, radiologist specialist in Artificial Intelligence applied to breast screening and speaker at the 36th Congress of the Spanish Society of Medical Radiology (SERAM) and XXXI CIR.

The development of new artificial intelligence systems with “deep learning” technology has improved traditional CAD algorithms. Therefore, they are capable of detecting suspicious lesions of breast cancer both in digital mammography like in tomosynthesis, assigning them a score based on the probability of malignancy. In addition, they can be used as support for the radiologist’s readings, making the task easier, achieving a reduction in delay, increasing cancer detection and decreasing false positives and negatives.

Artificial intelligence is also capable of classifying mammograms based on the probability of malignancy, allowing the radiologist to focus their actions on those with the highest probability of cancer.

Although more prospective studies carried out in real clinical settings are necessary, in the short term artificial intelligence will have a very prominent role in mammography” says Dr. Elías.

There are studies that show that it could be reduce the workload in screening programs by up to 70% without reducing sensitivity. It also increases the positive predictive value of referred patients, especially in patients classified by the system as high risk.

According to Dr. Elijah “In addition, artificial intelligence would allow the generalization of the use of tomosynthesis in screening programs, where an increase in the detection of breast cancer has also been shown, since artificial intelligence would allow reducing the increase in the workload that they entail. tomosynthesis readings, because the radiologist takes more than twice as long to read a tomosynthesis than a digital mammogram”.

Radiogenomics

Radiogenomics is a new artificial intelligence technique, which applied to medical images (breast resonance, digital mammography, tomosynthesis, ultrasound or PET) studies the relationship between image phenotypes and the tumor genome. This implies that, through artificial intelligence, radiogenomics tries to obtain image biomarkers that relate the characteristics of the image that cancer presents (image phenotypes), with respect to the surrounding healthy tissue, with the processes that occur at the genetic and molecular in tumor tissue.

The goal of radiogenomics is develop imaging biomarkers incorporating phenotypic and genotypic alterations that can predict risk and outcomes, allowing patients to be better stratified for more precise treatment.

These new techniques will allow a precision medicine through a complex study, with computer and mathematical models, which assess the interaction of genes, metabolites, proteins and other biological components, being able to carry out a focused treatment of the disease and prevention strategies for groups of individuals, according to their disease, genetics, factors environments and lifestyles. Biopsies can be avoided, thus avoiding comorbidity and associated complications, which only provide us with information on the biopsied part of the tumor, while radiogenomics studies the entire tumor and the patient as a whole” points out Dr. Elías.

Trending