http://journals.andromedapublisher.com/index.php/JMLFS/issue/feed 2022-02-02T22:19:28+00:00 Professor Stefano Moretti stefano.moretti@andromedapublisher.com Open Journal Systems <p><em><span style="font-size: 11.0pt;">JMLFS is a peer-reviewed, open access journal which specialises on ML driven advances in any of the <strong><span style="font-weight: normal;">fundamental sciences</span></strong>, here intended as mathematics, <strong><span style="font-weight: normal;">physics</span></strong>, chemistry and biology. The latter are organised in four&nbsp; subsections, called A, B, C &amp; D, each with at least one editor, with no distinction between theoretical and experimental content. JMLFS publishes submitted articles of letter type, occasional special/topical issues (by invitation) and will consider publishing proceedings (upon enquiry).</span></em></p> http://journals.andromedapublisher.com/index.php/JMLFS/article/view/229 2022-02-02T22:19:27+00:00 Ammu Raju ammu@iiitb.ac.in Neelam Sinha neelam.sinha@iiitb.ac.in

<p>Automatic image segmentation is a critical component of medical image analysis, and hence quantifying<br>segmentation performance is crucial. Challenges in medical image segmentation are mainly due to spatial<br>variations of regions to be segmented and an imbalance in distribution of the classes. Commonly used<br>metrics treat all detected pixels indiscriminately. However, pixels in smaller segments must be treated differently from pixels in larger segments, as smaller ones aid in early treatment of associated disease and are<br>also easier to miss. To address this, we propose a novel evaluation metric for segmentation performance,<br>emphasizing smaller segments, by assigning higher weightage to smaller segment pixels. “Small SEGment<br>Emphasized Performance” (SSEGEP) metric, based on weighted false positives, was proposed and evaluated using synthetic images and 4 publicly available clinical datasets of eye (fundus imaging, n = 33),<br>breast (mammogram, n = 108), liver (CT, n = 131), and pancreas cancer (CT, n = 107), where n refers<br>to the number of images from the dataset that were used in the study. Mean opinion score (MOS) was<br>calculated from the scores (5-scale scores) of 33-fundus-image segmentation as assigned by 15 researchers<br>(2–5 years of experience in image analysis). Statistical analysis was performed for the other datasets to<br>quantify the relevance of the proposed approach. Across 33 fundus images, where the largest exudate is<br>1.41% and the smallest is 0.0002% of the image, the proposed metric is 30% closer to MOS, as compared<br>to Dice Similarity Coefficient (DSC). Statistical significance testing resulted in a p value of order 10−18 that<br>shows the significance of SSEGEP for hepatic tumors compared to DSC. The proposed metric is found to<br>perform better for the images having multiple segments of a single class.</p>

2022-02-02T22:17:27+00:00 ##submission.copyrightStatement##