The Development and Application of Artificial Intelligence 
in Medical Image Diagnosis

Chunjie Lv; Biyuan Li*; Gaowei Sun; Bo Yang; Xiuwei Wang

doi:10.70711/mhr.v2i3.4890

The Development and Application of Artificial Intelligence in Medical Image Diagnosis

Chunjie Lv, Biyuan Li*, Gaowei Sun, Bo Yang, Xiuwei Wang

Abstract

This study addresses the challenges in image segmentation and detection within the field of medical image diagnosis by proposing and thoroughly exploring the optimization of the CRTransUNet network structure. While artificial intelligence (AI) has made significant
progress in medical image processing, traditional convolutional neural networks (CNNs), despite their strong performance in segmentation
tasks, have limitations in capturing multi-scale features and handling complex backgrounds. To overcome these issues, this study introduces
Transformer layers and the Cross-Region Multi-Scale Attention Mechanism (CRMSA) into the CRTransUNet network structure. By combining CNNs with Transformers, CRTransUNet effectively captures global contextual information and long-range dependencies, significantly
improving segmentation accuracy and the ability to handle complex structures. Additionally, the CRMSA mechanism enhances feature extraction by establishing attention connections between different regions, deepening the extraction and cross-regional integration of features. This
study not only improves the accuracy of medical image segmentation and detection but also explores the broader applications of AI in medical
diagnosis, particularly in addressing challenges related to data privacy, algorithm transparency, and clinical validation. The findings provide
essential theoretical and practical support for the future development of intelligent and precise medical image diagnosis.

Keywords

Artificial Intelligence (AI); Medical Image Diagnosis; Deep Learning; Diagnostic Accuracy; Clinical Validation

Full Text:

PDF

Included Database

References

[1] Ketkar, Nikhil, Jojo Moolayil, Nikhil Ketkar, and Jojo Moolayil. "Convolutional neural networks." Deep learning with Python: learn

best practices of deep learning models with PyTorch (2021): 197-242.

[2] Lin, Tsung-Yi, et al. "Feature pyramid networks for object detection." Proceedings of the IEEE conference on computer vision and pattern recognition. 2017.

[3] Chen, Jieneng, Jieru Mei, Xianhang Li, Yongyi Lu, Qihang Yu, Qingyue Wei, Xiangde Luo et al. "TransUNet: Rethinking the U-Net architecture design for medical image segmentation through the lens of transformers." Medical Image Analysis 97 (2024): 103280.

[4] Cao, Hu, Yueyue Wang, Joy Chen, Dongsheng Jiang, **aopeng Zhang, Qi Tian, and Manning Wang. "Swin-unet: Unet-like pure transformer for medical image segmentation." In European conference on computer vision, pp. 205-218. Cham: Springer Nature Switzerland,

2022.

[5] Wu, Hang, Yubin Miao, and Ruochong Fu. "Point cloud completion using multiscale feature fusion and cross-regional attention." Image

and Vision Computing 111 (2021): 104193.

[6] Hasany, Syed Nouman, Fabrice Mriaudeau, and Caroline Petitjean. "The Dos and Donts of Grad-CAM in Image Segmentation as

demonstrated on the Synapse multi-organ CT Dataset." In Medical Imaging with Deep Learning. 2024.

DOI: http://dx.doi.org/10.70711/mhr.v2i3.4890

Refbacks

There are currently no refbacks.