We are creating some awesome events for you. Kindly bear with us.

Machine Learning Model Finds COVID-19 Risks for Cancer Patients

When it comes to whether cancer patient tests positive for COVID-19 is more likely to become hospitalised from the disease, that on certain risk factors, according to a new study by researchers at Lawrence Livermore National Laboratory (LLNL) and the University of California, San Francisco (UCSF), who looked to identify cancer-related risks for poor outcomes from COVID-19.

The scientists discovered previously unreported associations between a rare type of cancer — as well as two cancer treatment-related medicines — and an elevated risk of hospitalisation from COVID-19 after analysing one of the largest databases of cancer patients.

There is still a lot of concern about the influence of COVID-19 on the risk of cancer patients. The strength of the dataset, which allowed us to look at this particular sector and was large enough to identify some statistically significant cancer-related and medicine-related viewpoints, was critical to the success of this work. These are potentially actionable items for clinicians to share with their patients in order to raise awareness, increase precautions, or perhaps consider alternative therapies.

The researchers looked at de-identified Electronic Health Record (EHR) data from the UC Health COVID Research Data Set (UC CORDS) on nearly a half-million patients who had COVID-19 testing at all 17 UC-affiliated institutions using a logistical regression approach. The database contains information on patient demographics, comorbidities, lab work, cancer kinds, and various cancer therapies for approximately 50,000 cancer patients, more than 17,000 of whom also tested positive for COVID.

The researchers looked at a variety of factors and disease outcomes, such as hospitalisation, ventilation, and death, and discovered a link between COVID-19 and a specific group of rarer blood cancers, as well as two cancer medications: venetoclax (used to treat leukaemia) and methotrexate (used to treat cancer) (an immune suppressant used in chemotherapy).

The ability to evaluate the potential risk of many different factors with some reasonable confidence that the findings are statistically significant and relevant to a population outside the cohort under study is made possible by having access to such a large database with detailed medical histories for each patient.

Such unexpected results highlight one of the difficulties in using EHRs to model disease: many confounding factors that complicate the research and must be accounted for. More research is needed to better understand and confirm the decreased risk of COVID-19 test positivity in cancer patients.

The research is part of a pilot study, and the researchers are looking for external funding to continue their investigation. To find better ways to treat patients, Ray said the researchers would like to integrate time-varying genetic and imaging data, as well as more complex AI and other techniques. They seek to learn not only about the disease’s mechanisms but also about how socioeconomic factors like money and insurance can influence it.

As reported by OpenGov Asia, creating smarter, more accurate systems requires a hybrid human-machine approach, according to researchers at the University of California, Irvine. In a study published this month in Proceedings of the National Academy of Sciences, they present a new mathematical model that can improve performance by combining human and algorithmic predictions and confidence scores.

To test the framework, researchers conducted an image classification experiment in which human participants and computer algorithms worked separately to correctly identify distorted pictures of animals and everyday items—chairs, bottles, bicycles, trucks. The human participants ranked their confidence in the accuracy of each image identification as low, medium or high, while the machine classifier generated a continuous score. The results showed large differences in confidence between humans and AI algorithms across images.

Send this to a friend