SSM-Net:半监督的多任务网络,用于从胰腺EUS图像进行关节损伤细分和分类
Jiajia Li1, Pingping Zhang2, Xia Yang3
1School of Chemistry and Chemical Engineering and National Center for Translational Medicine, Shanghai Jiao Tong University, Shanghai, China.
Artificial intelligence in medicine
|June 28, 2024
概括
这项研究引入了一种新的半监督多任务网络 (SSM-Net),用于改进胰腺癌检测. 该方法通过精确地对内镜超声波 (EUS) 图像中的病变进行分类和细分,提高了早期诊断的性能,优于现有的技术.
科学领域:
- 医疗成像医学成像
- 人工智能的人工智能
- 在瘤学瘤学.
背景情况:
- 由于非特异性症状和使用当前成像方法早期检测的困难,胰腺癌的诊断具有挑战性.
- 内镜超声波 (EUS) 对于胰腺疾病诊断至关重要,但B模式成像易受人工物影响,使损伤分析复杂化.
- 精确细分和分类胰腺病变是必不可少的,但需要专门的专业知识.
研究的目的:
- 开发一个半监督的多任务网络 (SSM-Net) 用于在EUS图像中联合分类和细分胰腺病变.
- 利用标记和未标记的EUS数据来提高诊断准确性并减少专家的负担.
- 通过先进的人工智能技术,提高胰腺癌的早期检测和特征.
主要方法:
- 使用未标记的EUS图像开发了一个突出意识的表示学习模块 (SRLM),用于训练特征提取编码器.
- 一个光谱残余模块 (SRM) 产生语义突出性地图用于对比损失计算.
- 道注意力阻塞 (CAB),合并的全球注意力模块 (MGAM) 和特征相似性损失 (FSL) 用于标记数据上的病变细分和分类.
主要成果:
- 与最先进的方法相比,拟议的SSM-Net在胰腺病变分类和细分方面表现优越.
- 在基于EUS的大型胰腺图像数据集 (LS-EUSPI) 和公共甲状腺数据集上进行了实验.
- 该模型有效地利用了标记和未标记的数据,展示了在医学图像分析中半监督学习的力量.
结论:
- SSM-Net提供了一种有前途的方法,可以提高EUS成像中胰腺病变检测的准确性和效率.
- 这种人工智能驱动的方法有可能帮助临床医生更早,更准确地诊断胰腺癌.
- 该研究强调了在半监督医疗图像分析中整合表示学习和注意力机制的有效性.
相关概念视频
01:18Steps in Outbreak Investigation
In the ever-evolving field of public health, statistical analysis serves as a cornerstone for understanding and managing disease outbreaks. By leveraging various statistical tools, health professionals can predict potential outbreaks, analyze ongoing situations, and devise effective responses to mitigate impact. For that to happen, there are a few possible stages of the analysis:
01:28Infectious Diseases and Their Occurrence
Infectious diseases appear in populations through various transmission patterns, influenced by pathogen characteristics, population immunity, environmental conditions, and social behavior. Understanding these patterns is essential for effective public health surveillance and intervention. These categories—sporadic, outbreak, epidemic, pandemic, and endemic—help frame the nature and scope of disease events.Sporadic diseases occur irregularly and infrequently, without a predictable temporal or...
01:28Rabies
Rabies is a lethal zoonotic disease caused by a single-stranded, negative-sense RNA virus of the Lyssavirus genus, within the family Rhabdoviridae. Its primary mode of transmission to humans is through bites or saliva-contaminated scratches from infected mammals such as dogs, bats, raccoons, or foxes. Transmission can also occur if infectious saliva contacts abraded skin or intact mucous membranes, including the conjunctiva.Viral Entry and Early ReplicationOnce introduced at the bite or scratch...