首页|Friedrich-Alexander-University Erlangen-Nurnberg (FAU) Reports Findings in Machi ne Learning (Features in Backgrounds of Microscopy Images Introduce Biases in Ma chine Learning Analyses)
Friedrich-Alexander-University Erlangen-Nurnberg (FAU) Reports Findings in Machi ne Learning (Features in Backgrounds of Microscopy Images Introduce Biases in Ma chine Learning Analyses)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-New research on Machine Learning is th e subject of a report. According to news reporting from Erlangen, Germany, by Ne wsRx journalists, research stated, "Subvisible particles may be encountered thro ughout the processing of therapeutic protein formulations. Flow imaging microsco py (FIM) and backgrounded membrane imaging (BMI) are techniques commonly used to record digital images of these particles, which may be analyzed to provide part icle size distributions, concentrations, and identities." The news correspondents obtained a quote from the research from Friedrich-Alexan der-University Erlangen-Nurnberg (FAU), "Although both techniques record digital images of particles within a sample, FIM analyzes particles suspended in flowin g liquids, whereas BMI records images of dry particles after collection by filtr ation onto a membrane. This study compared the performance of convolutional neur al networks (CNNs) in classifying images of subvisible particles recorded by bot h imaging techniques. Initially, CNNs trained on BMI images appeared to provide higher classification accuracies than those trained on FIM images. However, attr ibution analyses showed that classification predictions from CNNs trained on BMI images relied on features contributed by the membrane background, whereas predi ctions from CNNs trained on FIM features were based largely on features of the p articles. Segmenting images to minimize the contributions from image backgrounds reduced the apparent accuracy of CNNs trained on BMI images but caused minimal reduction in the accuracy of CNNs trained on FIM images. Thus, the seemingly sup erior classification accuracy of CNNs trained on BMI images compared to FIM imag es was an artifact caused by subtle features in the backgrounds of BMI images."
ErlangenGermanyEuropeCyborgsEmer ging TechnologiesMachine Learning
NETL
NSTL
