This website requires JavaScript.

LMFLOSS: A Hybrid Loss For Imbalanced Medical Image Classification

Abu Adnan SadiLabib ChowdhuryNursrat Jahan ...+3 Nabeel Mohammed
Dec 2022
摘要
Automatic medical image classification is a very important field where theuse of AI has the potential to have a real social impact. However, there arestill many challenges that act as obstacles to making practically effectivesolutions. One of those is the fact that most of the medical imaging datasetshave a class imbalance problem. This leads to the fact that existing AItechniques, particularly neural network-based deep-learning methodologies,often perform poorly in such scenarios. Thus this makes this area aninteresting and active research focus for researchers. In this study, wepropose a novel loss function to train neural network models to mitigate thiscritical issue in this important field. Through rigorous experiments on threeindependently collected datasets of three different medical imaging domains, weempirically show that our proposed loss function consistently performs wellwith an improvement between 2%-10% macro f1 when compared to the baselinemodels. We hope that our work will precipitate new research toward a moregeneralized approach to medical image classification.
展开全部
图表提取

暂无人提供速读十问回答

论文十问由沈向洋博士提出,鼓励大家带着这十个问题去阅读论文,用有用的信息构建认知模型。写出自己的十问回答,还有机会在当前页面展示哦。

Q1论文试图解决什么问题?
Q2这是否是一个新的问题?
Q3这篇文章要验证一个什么科学假设?
0
被引用
笔记
问答