相关实验视频
Updated: Feb 13, 2026

13:19
Deep Neural Networks for Image-Based Dietary Assessment
Published on: March 13, 2021
10.0K
基于使用超声波图像的卷积神经网络模型,开发了胆囊瘤多的术前预测模型
Yongyi Zhu1, Yi Lu2, Qingjin Zeng1
1Department of Ultrasound, The Third Affiliated Hospital, Sun Yat-sen University, Guangzhou, China.
Translational gastroenterology and hepatology
|February 12, 2026
概括
一个卷积神经网络 (CNN) 模型准确地使用超声波图像预测胆囊中的瘤多. 这个AI工具有助于选择合适的治疗方法,用于胆囊的多性病变 (PLGs).
科学领域:
- 医疗成像医学成像
- 人工智能在医学中的应用
- 胃肠病学 胃肠病学
背景情况:
- 胆囊的多性病变 (PLG) 是常见的,需要区分瘤和非瘤类型.
- 新生态息肉需要手术切除,使得准确的手术前鉴定对于患者管理至关重要.
- 使用当前的方法,区分瘤与非瘤息肉仍然是一个重大的临床挑战.
研究的目的:
- 开发和评估一个卷积神经网络 (CNN) 模型,用于使用超声波图像在胆囊中预先预测胆囊中瘤多.
- 评估CNN模型在识别瘤胆囊息肉中的可靠性和诊断性能.
主要方法:
- 一项多中心的回顾性研究,涉及380例病例 (921张超声波图像).
- 一个CNN模型 (Inception-V3) 使用专用训练套的超声波图像进行训练.
- CNN模型的预测性能在内部和外部测试集上得到验证,并与超声波仪和名ogram模型进行了比较.
主要成果:
- 在CNN模型下,曲线下的面积 (AUC) 达到0.896 (内部) 和0.852 (外部测试集).
- 与三种超声波相比,CNN模型显示出更高的诊断效率 (AUC 0.687-0.803).
- 该CNN模型的性能与基于超声波特征的诺莫格拉姆模型 (AUC 0.880) 相当.
结论:
- 开发的CNN模型显示,在术前识别瘤胆囊聚体时,具有有前途的预测性能.
- 基于超声波的CNN分析对于协助选择适合PLG的治疗策略非常有价值.
- 这种人工智能方法提供了一种可靠的工具,可以提高瘤胆囊息肉的诊断准确度.
相关概念视频
Gallbladder
2.4K
The gallbladder is a small, pear-shaped organ that plays a crucial role in our digestive system. Measuring about 10 cm in length, it is comparable in size to a kiwi fruit and is located in a hollow area on the lower surface of the liver. The gallbladder's primary function is to store and concentrate bile, a fluid produced by the liver that aids in digestion.
The gallbladder's anatomy consists of three regions: the fundus, body, and neck. Extending from the neck, the cystic duct joins...
The gallbladder's anatomy consists of three regions: the fundus, body, and neck. Extending from the neck, the cystic duct joins...
2.4K
Protein Networks
4.6K
An organism can have thousands of different proteins, and these proteins must cooperate to ensure the health of an organism. Proteins bind to other proteins and form complexes to carry out their functions. Many proteins interact with multiple other proteins creating a complex network of protein interactions.
These interactions can be represented through maps depicting protein-protein interaction networks, represented as nodes and edges. Nodes are circles that are representative of a protein,...
These interactions can be represented through maps depicting protein-protein interaction networks, represented as nodes and edges. Nodes are circles that are representative of a protein,...
4.6K
Convolution Properties II
594
The important convolution properties include width, area, differentiation, and integration properties.
The width property indicates that if the durations of input signals are T1 and T2, then the width of the output response equals the sum of both durations, irrespective of the shapes of the two functions. For instance, convolving two rectangular pulses with durations of 2 seconds and 1 second results in a function with a width of 3 seconds.
The area property asserts that the area under the...
The width property indicates that if the durations of input signals are T1 and T2, then the width of the output response equals the sum of both durations, irrespective of the shapes of the two functions. For instance, convolving two rectangular pulses with durations of 2 seconds and 1 second results in a function with a width of 3 seconds.
The area property asserts that the area under the...
594
Diseases of the Liver and Gallbladder
2.0K
Liver and gallbladder diseases are a significant health concern, with prominent conditions including cirrhosis, hepatitis, non-alcoholic fatty liver disease (NAFLD), and gallstones. Jaundice is a common manifestation of liver and biliary disease.
Cirrhosis is characterized by the scarring of hepatic lobules in the liver, which are replaced by fibrous tissue, affecting the liver's normal functioning. NAFLD, on the other hand, is caused by an excessive build-up of fat in the liver, not...
Cirrhosis is characterized by the scarring of hepatic lobules in the liver, which are replaced by fibrous tissue, affecting the liver's normal functioning. NAFLD, on the other hand, is caused by an excessive build-up of fat in the liver, not...
2.0K
Stereotype Content Model
15.5K
The Stereotype Content Model (SCM) was first proposed by Susan Fiske and her colleagues (Fiske, Cuddy, Glick & Xu, 2002; see also Fiske, 2012 and Fiske, 2017). The SCM specifies that when someone encounters a new group, they will stereotype them based on two metrics: warmth—or that group’s perceived intent, and how likely they are to provide help or inflict harm—and competence—or their ability to carry out that objective. Depending on the warmth-competence...
15.5K
Convolution Properties I
621
Convolution computations can be simplified by utilizing their inherent properties.
The commutative property reveals that the input and the impulse response of an LTI (Linear Time-Invariant) system can be interchanged without affecting the output:
The commutative property reveals that the input and the impulse response of an LTI (Linear Time-Invariant) system can be interchanged without affecting the output:
621

