Abstract:Unsupervised segmentation techniques, which do not require labeled data for training and can be more easily integrated into the clinical routine, represent a valid solution especially from a clinical feasibility perspective. Indeed, large-scale annotated datasets are not always available, undermining their immediate implementation and use in the clinic. Breast cancer is the most common cause of cancer death in women worldwide. In this study, breast lesion delineation in Dynamic Contrast Enhanced MRI (DCE-MRI) series was addressed by means of four popular unsupervised segmentation approaches: Split-and-Merge combined with Region Growing (SMRG), k-means, Fuzzy C-Means (FCM), and spatial FCM (sFCM). They represent well-established pattern recognition techniques that are still widely used in clinical research. Starting from the basic versions of these segmentation approaches, during our analysis, we identified the shortcomings of each of them, proposing improved versions, as well as developing ad hoc pre- and post-processing steps. The obtained experimental results, in terms of area-based—namely, Dice Index (DI), Jaccard Index (JI), Sensitivity, Specificity, False Positive Ratio (FPR), False Negative Ratio (FNR)—and distance-based metrics—Mean Absolute Distance (MAD), Maximum Distance (MaxD), Hausdorff Distance (HD)—encourage the use of unsupervised machine learning techniques in medical image segmentation. In particular, fuzzy clustering approaches (namely, FCM and sFCM) achieved the best performance. In fact, for area-based metrics, they obtained DI = 78.23% ± 6.50 (sFCM), JI = 65.90% ± 8.14 (sFCM), sensitivity = 77.84% ± 8.72 (FCM), specificity = 87.10% ± 8.24 (sFCM), FPR = 0.14 ± 0.12 (sFCM), and FNR = 0.22 ± 0.09 (sFCM). Concerning distance-based metrics, they obtained MAD = 1.37 ± 0.90 (sFCM), MaxD = 4.04 ± 2.87 (sFCM), and HD = 2.21 ± 0.43 (FCM). These experimental findings suggest that further research would be useful for advanced fuzzy logic techniques specifically tailored to medical image segmentation.

A Performance Comparison of Supervised and Unsupervised Image Segmentation Methods

Semi-Supervised Convolutional Vision Transformer with Bi-Level Uncertainty Estimation for Medical Image Segmentation

Two Approaches to Supervised Image Segmentation

A Semi-Supervised Color Image Segmentation Method

Dataset and Performance Metrics towards Semantic Segmentation

A comprehensive survey of image segmentation: clustering methods, performance parameters, and benchmark datasets

Review on machine learning techniques for image segmentation

Image Segmentation Algorithms Overview

Unsupervised Salient Object Segmentation From Color Images

Performance Analysis of Clustering Based Image Segmentation Techniques

Image segmentation evaluation: a survey of methods

A framework for evaluating image segmentation algorithms

A Comprehensive Review of Modern Object Segmentation Approaches

Diagnostics in Semantic Segmentation

On Unsupervised Methods for Medical Image Segmentation: Investigating Classic Approaches in Breast Cancer DCE-MRI

Image Segmentation Using Deep Learning: A Survey

Recent Progress in Semantic Image Segmentation

A survey of semi- and weakly supervised semantic segmentation of images

Image-driven Unsupervised 3D Model Co-Segmentation

An Efficient Data Analytics and Optimized Algorithm for Enhancing the Performance of Image Segmentation Using Deep Learning Model

Supervised Classification Performance of Multispectral Images