Image texture analysis for prostate cancer detection
Kiros, Fetsume Berhe (2019)
2019
Electrical Engineering
Lääketieteen ja terveysteknologian tiedekunta - Faculty of Medicine and Health Technology
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2019-06-04
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tty-201906111871
https://urn.fi/URN:NBN:fi:tty-201906111871
Tiivistelmä
Recent advancement in diagnostic imaging has made the mortality rate associated with prostate cancer (PCa) to decrease drastically. Given to lower threshold cut-off values for prostate-specific antigen (PSA) which is the primary augury aside from clinical signs and symptoms for the presence of cancer and indication for ultrasound-guided biopsy analysis. Due to its availability, low cost, justification and well tolerance of the patient, ultrasound (US) stays to be the golden modality for prostate evaluation and real-time biopsy guidance.
While US stays to be in the first line of a diagnostic prediction method for the presence of PCa it is mainly used for biopsy guidance on account of its lower sensitivity and specificity towards the differential of cancer from other pathologies and relatively high rate of false negative results.
Due to the variable sonographic appearance of the malignancy and its similarity with the non-malignant conditions, it is always possible only to differentiate cancer from the non-cancerous conditions by using biopsy. Moreover, in cases where malignant conditions are apparent together with the non-malignant pathologies there will be the superimposition of echo signals making the differential diagnosis difficult.
Image processing and texture analysis can improve the diagnostic details obtained from digital diagnostic two-dimensional (2D) or three-dimensional (3D) images. We can analyze the image texture parameters quantitatively using the specific MaZda software to ascertain different anatomy for distinguishing normal from the abnormal tissue structures.
In this thesis work prostate, US images texture parameters are analyzed using MaZda software to distinguish the classic texture distribution of PCa. All patients in the study had elevated PSA value and biopsy has confirmed the presence of malignancy.
Keywords: ultrasound, prostate cancer, sensitivity, specificity, pixel, voxel, image processing, texture, texture analysis, MaZda, region of interest (ROI)
While US stays to be in the first line of a diagnostic prediction method for the presence of PCa it is mainly used for biopsy guidance on account of its lower sensitivity and specificity towards the differential of cancer from other pathologies and relatively high rate of false negative results.
Due to the variable sonographic appearance of the malignancy and its similarity with the non-malignant conditions, it is always possible only to differentiate cancer from the non-cancerous conditions by using biopsy. Moreover, in cases where malignant conditions are apparent together with the non-malignant pathologies there will be the superimposition of echo signals making the differential diagnosis difficult.
Image processing and texture analysis can improve the diagnostic details obtained from digital diagnostic two-dimensional (2D) or three-dimensional (3D) images. We can analyze the image texture parameters quantitatively using the specific MaZda software to ascertain different anatomy for distinguishing normal from the abnormal tissue structures.
In this thesis work prostate, US images texture parameters are analyzed using MaZda software to distinguish the classic texture distribution of PCa. All patients in the study had elevated PSA value and biopsy has confirmed the presence of malignancy.
Keywords: ultrasound, prostate cancer, sensitivity, specificity, pixel, voxel, image processing, texture, texture analysis, MaZda, region of interest (ROI)