High breast density or a greater amount of glandular and connective tissue compared to fat are considered as one of the major risk factors for cancer. It is believed, while density may be incorporated into risk assessment, the current prediction models are likely to fail to take complete advantage of the expansive information present in mammograms. Experts allege this information can help to identify women who would benefit from the additional screening with MRI. Karin Dembrower, M.D., breast radiologist and Ph.D. candidate from the Karolinska Institute in Stockholm, Sweden has developed a risk model which is dependent on a wide neural network, a type of AI that can extract vast amounts of information from mammographic images. These have several advantages in comparison to other methods like visual assessment of mammographic density by the radiologists, which are not capable of capturing all risk-relevant information in the image.
Most existing breast cancer screening programs are based on mammography at similar time intervals -- typically, annually or every two years -- for all women. This "one size fits all" approach is not optimized for cancer detection on an individual level and may hamper the effectiveness of screening programs. "Risk prediction is an important building block of an individually adapted screening policy," said study lead author Karin Dembrower, M.D., breast radiologist and Ph.D. candidate from the Karolinska Institute in Stockholm, Sweden. "Effective risk prediction can improve attendance and confidence in screening programs." High breast density, or a greater amount of glandular and connective tissue compared to fat, is considered a risk factor for cancer.
While the outcome for patients with chronic lymphocytic leukemia (CLL) has significantly improved over the last 30 years due to effective treatment options, the number of patients with CLL dying as a result of infection has not particularly decreased; indicating an unmet need in new patients with immune dysfunction. Here, Carsten Niemann, MD, PhD, of Copenhagen University Hospital, Copenhagen, Denmark, points to the statistic that over 25% of patients with CLL will have a severe infection, and how the use of a machine learning algorithm developed in collaboration with the Danish Technical University may help prevent infection-related deaths in patients with CLL. From the 23rd Congress of the European Hematology Association (EHA) 2018, held in Stockholm, Sweden, Dr Niemann explains how this process can help identify those patients with greater than 70% risk of infection, who can then receive either pre-emptive treatment or observation in a new clinical trial currently being designed.
This article is an edited transcript of a lecture given at IJCAI-99, Stockholm, Sweden, on 4 August 1999. The article summarizes concepts, principles, and tools that were found useful in applications involving causal modeling. The principles are based on structural-model semantics in which functional (or counterfactual) relationships representing autonomous physical processes are the fundamental building blocks. The article presents the conceptual basis of this semantics, illustrates its application in simple problems, and discusses its ramifications to computational and cognitive problems concerning causation.