Industry News.
A curated collection of the most timely and relevant news, updates and information from the world of AI and machine learning.
Novel Machine Learning Tool Predicts Gain- and Loss-of-function Variants Across The Genome
A new machine learning tool called LoGoFunc has been developed to predict whether genetic variations will have harmful effects. LoGoFunc can distinguish between harmful mutations that make proteins less active and those that make them more active. This is important because different types of mutations can lead to different diseases. LoGoFunc uses information from protein structures and human protein interactions to make its predictions. The researchers who created LoGoFunc believe it could be used to improve personalized medicine, drug discovery, and genetic counseling.
AI Predicts Malignancy on Breast Ultrasound
A new AI model has been developed to help radiologists avoid unnecessary biopsies and follow-up exams for breast cancer. The model can accurately predict malignancy on breast ultrasound based on BI-RADS assessment. In a study, the AI model was able to potentially avoid 11% of benign lesion biopsies and 46.2% of follow-ups. It also downgraded 29.4% of suspicious lesions to benign or probably benign. The authors suggest that AI could be a valuable tool for improving the accuracy and efficiency of breast cancer diagnosis.
Enabling Early Detection of Cancer With AI-Based Chromatin Biomarkers
Researchers have developed a new method for using AI to detect cancer early. The method analyzes changes in blood cell chromatin, the material that packages DNA, to identify tumors. In a study, the method was able to distinguish between healthy and sick people with 85% accuracy. It can also identify the type of tumor and monitor the success of treatment. This research is still in the early stages, but it has the potential to improve cancer diagnosis and treatment.
AI Study Reveals The Tongue's Unique Surface
Researchers used AI and 3D imaging to analyze the unique patterns of papillae on the human tongue. These papillae are the small bumps that give our tongue its texture and help us taste and touch. The study found that the patterns of papillae are unique to each person, which could have implications for personalized food preferences and early diagnosis of oral cancers.
AI Improves X-Ray Reads for Rads With Varying Experience
A new study reveals the promising potential of AI in assisting radiologists of all experience levels with chest x-ray interpretation. Researchers tested a commercially available AI algorithm and found it boosted the ability of radiologists to detect abnormalities, regardless of their years of practice. The AI also shaved 31% off reading time, suggesting improved efficiency alongside accuracy. However, the study focused on a specific set of abnormalities and lacked patient clinical history context for the radiologists. These findings pave the way for AI as a valuable tool to enhance radiologists' diagnostic capabilities and potentially streamline workflows.
New AI Algorithm Identifies Risk of Falls and Fracture for Older Women
A new AI algorithm is being developed to help identify older women at risk of falls and fractures. The algorithm analyzes abdominal aortic calcification (AAC) during routine bone density scans. AAC is a marker for advanced vascular disease, and studies have shown that women with moderate to extensive AAC are more likely to experience falls and fractures. The AI algorithm can assess AAC in under a minute, while it takes an experienced reader 5-6 minutes. This could allow for automatic AAC assessment during bone density scans, potentially helping to prevent falls and fractures in older women.