A survey, review, and future trends of skin lesion segmentation and classification

Computers in Biology and Medicine, Journal Year: 2023, Volume and Issue: 155, P. 106624 - 106624

Published: Feb. 1, 2023

The Computer-aided Diagnosis or Detection (CAD) approach for skin lesion analysis is an emerging field of research that has the potential to alleviate burden and cost cancer screening. Researchers have recently indicated increasing interest in developing such CAD systems, with intention providing a user-friendly tool dermatologists reduce challenges encountered associated manual inspection. This article aims provide comprehensive literature survey review total 594 publications (356 segmentation 238 classification) published between 2011 2022. These articles are analyzed summarized number different ways contribute vital information regarding methods development systems. include relevant essential definitions theories, input data (dataset utilization, preprocessing, augmentations, fixing imbalance problems), method configuration (techniques, architectures, module frameworks, losses), training tactics (hyperparameter settings), evaluation criteria. We intend investigate variety performance-enhancing approaches, including ensemble post-processing. also discuss these dimensions reveal their current trends based on utilization frequencies. In addition, we highlight primary difficulties evaluating classification systems using minimal datasets, as well solutions difficulties. Findings, recommendations, disclosed inform future automated robust system analysis.

Language: Английский

---

Enhancing Skin Cancer Diagnosis Using Swin Transformer with Hybrid Shifted Window-Based Multi-head Self-attention and SwiGLU-Based MLP

Deleted Journal, Journal Year: 2024, Volume and Issue: unknown

Published: June 5, 2024

Abstract Skin cancer is one of the most frequently occurring cancers worldwide, and early detection crucial for effective treatment. Dermatologists often face challenges such as heavy data demands, potential human errors, strict time limits, which can negatively affect diagnostic outcomes. Deep learning–based systems offer quick, accurate testing enhanced research capabilities, providing significant support to dermatologists. In this study, we Swin Transformer architecture by implementing hybrid shifted window-based multi-head self-attention (HSW-MSA) in place conventional (SW-MSA). This adjustment enables model more efficiently process areas skin overlap, capture finer details, manage long-range dependencies, while maintaining memory usage computational efficiency during training. Additionally, study replaces standard multi-layer perceptron (MLP) with a SwiGLU-based MLP, an upgraded version gated linear unit (GLU) module, achieve higher accuracy, faster training speeds, better parameter efficiency. The modified model-base was evaluated using publicly accessible ISIC 2019 dataset eight classes compared against popular convolutional neural networks (CNNs) cutting-edge vision transformer (ViT) models. exhaustive assessment on unseen test dataset, proposed Swin-Base demonstrated exceptional performance, achieving accuracy 89.36%, recall 85.13%, precision 88.22%, F1-score 86.65%, surpassing all previously reported deep learning models documented literature.

Language: Английский

---

A novel CNN-ViT-based deep learning model for early skin cancer diagnosis

Biomedical Signal Processing and Control, Journal Year: 2025, Volume and Issue: 104, P. 107627 - 107627

Published: Jan. 28, 2025

Language: Английский

---

An Innovative Deep Learning Framework for Skin Cancer Detection Employing ConvNeXtV2 and Focal Self-Attention Mechanisms

Results in Engineering, Journal Year: 2024, Volume and Issue: unknown, P. 103692 - 103692

Published: Dec. 1, 2024

Language: Английский

---

Towards Transparent Healthcare: Advancing Local Explanation Methods in Explainable Artificial Intelligence

Bioengineering, Journal Year: 2024, Volume and Issue: 11(4), P. 369 - 369

Published: April 12, 2024

This paper focuses on the use of local Explainable Artificial Intelligence (XAI) methods, particularly Local Rule-Based Explanations (LORE) technique, within healthcare and medical settings. It emphasizes critical role interpretability transparency in AI systems for diagnosing diseases, predicting patient outcomes, creating personalized treatment plans. While acknowledging complexities inherent trade-offs between model performance, our work underscores significance XAI methods enhancing decision-making processes healthcare. By providing granular, case-specific insights, like LORE enhance physicians’ patients’ understanding machine learning models their outcome. Our reviews significant contributions to healthcare, highlighting its potential improve clinical decision making, ensure fairness, comply with regulatory standards.

Language: Английский

---

A Deep Learning Model Enhances Clinicians' Diagnostic Accuracy to More Than 96% for Anterior Cruciate Ligament Ruptures on Magnetic Resonance Imaging

Arthroscopy The Journal of Arthroscopic and Related Surgery, Journal Year: 2023, Volume and Issue: 40(4), P. 1197 - 1205

Published: Aug. 18, 2023

Language: Английский

---

UCM-Net: A lightweight and efficient solution for skin lesion segmentation using MLP and CNN

Biomedical Signal Processing and Control, Journal Year: 2024, Volume and Issue: 96, P. 106573 - 106573

Published: July 3, 2024

Language: Английский

---

SkinNet-14: a deep learning framework for accurate skin cancer classification using low-resolution dermoscopy images with optimized training time

Neural Computing and Applications, Journal Year: 2024, Volume and Issue: 36(30), P. 18935 - 18959

Published: Aug. 1, 2024

Abstract The increasing incidence of skin cancer necessitates advancements in early detection methods, where deep learning can be beneficial. This study introduces SkinNet-14, a novel model designed to classify types using low-resolution dermoscopy images. Unlike existing models that require high-resolution images and extensive training times, SkinNet-14 leverages modified compact convolutional transformer (CCT) architecture effectively process 32 × pixel images, significantly reducing the computational load duration. framework employs several image preprocessing augmentation strategies enhance input quality balance dataset address class imbalances medical datasets. was tested on three distinct datasets—HAM10000, ISIC PAD—demonstrating high performance with accuracies 97.85%, 96.00% 98.14%, respectively, while time 2–8 s per epoch. Compared traditional transfer models, not only improves accuracy but also ensures stability even smaller sets. research addresses critical gap automated detection, specifically contexts limited resources, highlights capabilities transformer-based are efficient analysis.

Language: Английский

---

FastSAM3D: An Efficient Segment Anything Model for 3D Volumetric Medical Images

Lecture notes in computer science, Journal Year: 2024, Volume and Issue: unknown, P. 542 - 552

Published: Jan. 1, 2024

Language: Английский

---

Advancing Dermatological Diagnostics: Interpretable AI for Enhanced Skin Lesion Classification

Diagnostics, Journal Year: 2024, Volume and Issue: 14(7), P. 753 - 753

Published: April 2, 2024

A crucial challenge in critical settings like medical diagnosis is making deep learning models used decision-making systems interpretable. Efforts Explainable Artificial Intelligence (XAI) are underway to address this challenge. Yet, many XAI methods evaluated on broad classifiers and fail complex, real-world issues, such as diagnosis. In our study, we focus enhancing user trust confidence automated AI systems, particularly for diagnosing skin lesions, by tailoring an method explain model’s ability identify various lesion types. We generate explanations using synthetic images of lesions examples counterexamples, offering a practitioners pinpoint the features influencing classification outcome. validation survey involving domain experts, novices, laypersons has demonstrated that increase decision system. Furthermore, exploration latent space reveals clear separations among most common classes, distinction likely arises from unique characteristics each class could assist correcting frequent misdiagnoses human professionals.

Language: Английский

---