TY - JOUR AU - Sandino Garzon, Alvaro Andres AU - Herrera García, Rodrigo Javier PY - 2019/05/26 Y2 - 2024/03/28 TI - Detección de Candidatos a Microcalcificaciones Mamarias Agrupadas en Mamografías JF - Ingeniería JA - Ing. VL - 24 IS - 2 SE - Electrical, Electronic and Telecommunications Engineering DO - 10.14483/23448393.12512 UR - https://revistas.udistrital.edu.co/index.php/reving/article/view/12512 SP - 159-170 AB - <div><strong>Context</strong>: Mammary microcalcifications are not-palpable lesions that are present in approximately 55 %</div><div>of breast cancer. These are a frequent findings in mammograms and may be an indicator of the disease</div><div>in its early stages.</div><div>&nbsp;</div><div><strong>Method</strong>: A method was implemented in order to get mammary microcalcifications enhancement based</div><div>on multi-resolution analysis with Wavelet transform. Then, candidates were segmented using threshol-</div><div>ding, in this technique, the threshold was determined with statistical parameters from Wavelet distribu-</div><div>tion coefficients. Later, a couple of Support Vector Machines models was used to classify images that</div><div>contains mammary microcalcifications.</div><div>&nbsp;</div><div><strong>Results</strong>: Classification task was performed using Support Vector Machines (SVM). The following</div><div>evaluation metrics was achieved: AUC of 93.6 %, accuracy of 89.4 %, sensivity of 88.4 % and specificity</div><div>of 90.5 %</div><div>&nbsp;</div><div><strong>Conclusions</strong>: In this approach the length and distribution of microcalcifications was used as features</div><div>to select candidates. These features are also used as criteria in clinical evaluation to detect mammary</div><div>cancer in early stages. The proposed method to image enhancement can unmask microcalcifications</div><div>that are not visible at naked eye. In most mammographies the proposed algorithm classify correctly</div><div>microcalcifications in different distributions.</div> ER -