Acta Laser Biology Sinica, Volume. 29, Issue 2, 142(2020)

Photodynamic Therapy Based on Feature Extraction of Lesion Images

MEI Qisheng, CHEN Yufang, WU Zhencong, and WANG Ming*
Author Affiliations
  • [in Chinese]
  • show less

    The feature recognition and extraction of lesions has always been a classic problem in the field of medical image processing. The feature extraction of lesion images is to transform the original image into a set of characteristic images with obvious physical meaning and statistical significance. Identifying the lesion area and analyzing the condition of the lesion area is an important part of photodynamic therapy. Taking the red spot disease as an example, photodynamic therapy uses light, photosensitizer and molecular oxygen to exert a strong photodynamic effect on the diseased cells, allowing the active oxygen to kill the diseased cells, and finally achieve the purpose of treating the disease. Traditional treatment options have problems such as inability to accurately diagnose the lesion area, inability to accurately treat, and high overall treatment costs. This study designed a narrow-spectrum LED light source for irradiation treatment of port wine stains, and calculated the distribution of lesions and the degree of lesions according to the characteristics of the lesions. Finally, based on the calculation results, the film with non-uniform transmission density was used to precisely control the intensity of the treatment light source. The experimental results show that the image feature extraction technology, the designed light source and the film can solve the problems of inaccurate diagnosis, inaccurate treatment and treatment source uniformity, and have broad application prospects in the treatment of port wine stains described in the experiment.

    Tools

    Get Citation

    Copy Citation Text

    MEI Qisheng, CHEN Yufang, WU Zhencong, WANG Ming. Photodynamic Therapy Based on Feature Extraction of Lesion Images[J]. Acta Laser Biology Sinica, 2020, 29(2): 142

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category:

    Received: Oct. 22, 2019

    Accepted: --

    Published Online: Jan. 27, 2021

    The Author Email: Ming WANG (heqingli@nwu.edu.cn)

    DOI:10.3969/j.issn.1007-7146.2020.02.007

    Topics