Application Analysis of CycleGAN and ACGAN in Data 
Augmentation of Artificial Intelligence Medical Devices

Runsheng Sha

doi:10.70711/mhr.v2i5.6277

Application Analysis of CycleGAN and ACGAN in Data Augmentation of Artificial Intelligence Medical Devices

Runsheng Sha

Abstract

Objective: To analyze Cycle-Consistent Generative Adversarial Networks, The application value of CycleGAN and Auxiliary Classification Generative Adversarial Network (ACGAN) in data augmentation of artificial intelligence medical devices. Method: From January
2024 to June 2024, 20 pieces of artificial intelligence medical device data were selected as research objects by CycleGAN and ACGAN, and
20 pieces of original image data of artificial intelligence medical devices that had not been data expanded during the same period were selected as reference objects (control group). To compare the excellent rate of image data quality and the accuracy of data model evaluation results
between the two groups. Results: The rate of excellent quality of image data (100.00%) and the accuracy of data model evaluation results
(100.00%) in study group were higher than those in control group (70.00%) and control group (75.00%), and the difference between the two
groups was statistically significant (P < 0.05). Conclusion: CycleGAN and ACGAN have significant application value in the data augmentation of artificial intelligence medical devices, which can not only improve the image data quality of artificial intelligence medical devices, but
also improve the accuracy of data model evaluation results.

Keywords

CycleGAN; ACGAN; Artificial intelligence; Medical devices; Data augmentation

Full Text:

PDF

Included Database

References

[1] Zhang Xiaoyan, Liu Yutian, Wang Tongcai, et al. [1] Analysis of risk factors in clinical use of AI-assisted detection of medical devices [J].

Medical and Health Equipment, 2019, 45(11):77-82. (in Chinese)

[2] Guo Jiajie, Zhang Zhihao, Xiang Lei, et al. Research progress of PET image enhancement methods based on artificial intelligence [J].

China Medical Device Information, 2023, 29(13):25-34.

[3] Wu Xiaona, Chen Hui, He Qian, et al. Application effect of artificial intelligence visual recognition system in foreign medical device

examination [J]. Journal of Advanced Nursing, 2019, 39(02):161-163+203. (in Chinese)

[4] Valkova J, Adam M, Hlavacek J. Artificial intelligence in diabetic retinopathy screening: from idea to a medical device in clinical practice.[J]. Casopis lekaru ceskych, 2, 024, 162(7-8):290-293.

[5] Zhang XC, Bi H, et al. Application of artificial Intelligence-assisted cervical cytology diagnosis and expert consensus on quality control

[J]. China Maternal and Child Health Research, 2023, 34(12):7-13.

[6] Qiu Y P, Wu D, Xiao Y, et al. Definition and classification of artificial intelligence medical technology in China [J]. Journal of Medical

Informatics, 2023, 44(10):11-15.

[7] Line F, Isabelle B, Nicolas M. International Market Access Strategies for Artificial Intelligence-Based Medical Devices: Can We Standardize the Process to Faster Patient Access?[J]. Mayo Clinic Proceedings: Digital Health, 2023, 1(3):406-412.

[8] Yi Tao. Artificial Intelligence "activates" inefficient areas [N]. Qingdao Daily, 2023-08-07(002).

[9] Xiao Feiyi, Li Xue, Li Rui, et al. International experience and enlightenment of medical artificial intelligence technology evaluation and

supervision [J]. China Health Quality Control, 2023, 30(07):58-62. (in Chinese)

[10] Hao P F, Li Y, Chai R, et al. Application of CycleGAN and ACGAN in data augmentation of artificial intelligence medical devices [J].

China Medical Devices, 2019, 39(02):52-56+69.

DOI: http://dx.doi.org/10.70711/mhr.v2i5.6277

Refbacks

There are currently no refbacks.