DIAGNOSIS

This study tested different ways to diagnose tuberculosis (TB) in over 14,000 people across seven countries. The goal was to find which tests work best in local clinics and for groups that often miss diagnosis, like people with HIV and refugees. Researchers hope to make TB testin…

This study tested whether giving doctors access to an AI tool called Prof. Valmed helps them better identify rheumatic diseases. 82 licensed physicians with training in rheumatology or related fields participated. The AI provided disease suggestions, and researchers compared the …

This study looked at whether two simple tests—a breath test (FeNO) and a blood test (eosinophil count)—can help diagnose asthma more easily than current hospital-based breathing tests. Researchers studied 123 people with suspected asthma to see if these portable tests could accur…

This study tested whether a simple checklist called the YEARS algorithm can safely rule out dangerous blood clots in the lungs (pulmonary embolism) in cancer patients, reducing the need for CT scans. Researchers compared the YEARS approach to standard CT scanning alone in 698 adu…

This study looked at 63 people who had a central retinal artery occlusion (CRAO), also known as an eye stroke. Researchers used advanced MRI scans before and after treatment to find signs that could help diagnose the condition and predict how much vision might return after one mo…

This study looked at 87 heart failure patients aged 60 and older in Tunisia to find out how many also have cardiac amyloidosis, a condition where abnormal proteins build up in the heart. Researchers used heart imaging tests like echocardiograms and bone scans to detect the diseas…

This study tested whether a large language model (a type of AI) can improve how doctors think through medical cases. 249 physicians in internal or family medicine reviewed fictional patient stories, some with AI help and some without. Researchers then scored the accuracy of their…