An attention-based weakly supervised framework for spitzoid melanocytic lesion diagnosis in whole slide images

Amor, Rocío del; Launet, Laëtitia; Colomer, Adrián; Moscardo, Anais; Mosquera-Zamudio, Andres; Monteagudo, Carlos; Naranjo, Valery

doi:10.1016/j.artmed.2021.102197

Cited by 30 publications

(15 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As depicted in Table 2, models trained under a semi-supervision scenario clearly outperform the one trained with annotated samples only (S a ). Note that the proposed model also reaches a better classification accuracy than that of previous works studying the same type of lesions [9].…”

Section: Classificationmentioning

confidence: 63%

“…A notable implementation of this formulation is [3], where [7] have also successfully been applied to WSI, enabling the identification of informative tiles [8]. As for the automatic diagnosis of skin melanocytic lesions, [9] is the only approach so far that leverages the MIL scenario to analyze these tumors on WSIs, combining inductive transfer learning and weakly supervision to propose an end-to-end framework with an attention-based final prediction. Self-training.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

A Self-Training Weakly-Supervised Framework for Pathologist-Like Histopathological Image Analysis

Launet¹,

Colomer²,

Mosquera-Zamudio

et al. 2022

2022 IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

The advent of artificial intelligence-based tools applied to digital pathology brings the promise of reduced workload for pathologists and enhanced patient care, not to mention medical research progress. Yet, despite its great potential, the field is hindered by the paucity of annotated histological data, a limitation for developing robust deep learning models. To reduce the number of expert annotations needed for training, we introduce a novel framework combining self-training and weakly-supervised learning that uses both annotated and unannotated data samples. Inspired by how pathologists examine biopsies, our method considers whole slide images from a bird's eye view to roughly localize the tumor area before focusing on its features at a higher magnification level. Notwithstanding the scarcity of the dataset, the experimental results show that the proposed method outperforms models trained with annotated data only and previous works analyzing the same type of lesions, thus demonstrating the efficiency of the approach.

show abstract

Section: Classificationmentioning

confidence: 63%

Section: Related Workmentioning

confidence: 99%

A Self-Training Weakly-Supervised Framework for Pathologist-Like Histopathological Image Analysis

Launet¹,

Colomer²,

Mosquera-Zamudio

et al. 2022

2022 IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

show abstract

“…However, tissue biopsy is essential to achieve a formal diagnosis. Most deep-learning diagnostic applications for histological images are for the differentiation between melanoma and nevi [ 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 ]. However, multiple studies show applications in the differentiation between melanoma, nevi, and normal skin [ 34 , 35 ]; and differentiation between melanoma and nonmelanoma skin cancers [ 36 , 37 , 38 ].…”

Section: Resultsmentioning

confidence: 99%

“…Using a weakly supervised CNN-based method, Amor et al created a pipeline to identify tiles of tumor regions and then classify WSIs based on the output tiles. This group’s model for ROI extraction achieved an accuracy of 92.31% and a classification model accuracy of 80% [ 28 ].…”

Section: Resultsmentioning

confidence: 99%

Diagnostic and Prognostic Deep Learning Applications for Histological Assessment of Cutaneous Melanoma

Grant

Andrew

Alvarez

et al. 2022

Cancers

View full text Add to dashboard Cite

Melanoma is among the most devastating human malignancies. Accurate diagnosis and prognosis are essential to offer optimal treatment. Histopathology is the gold standard for establishing melanoma diagnosis and prognostic features. However, discrepancies often exist between pathologists, and analysis is costly and time-consuming. Deep-learning algorithms are deployed to improve melanoma diagnosis and prognostication from histological images of melanoma. In recent years, the development of these machine-learning tools has accelerated, and machine learning is poised to become a clinical tool to aid melanoma histology. Nevertheless, a review of the advances in machine learning in melanoma histology was lacking. We performed a comprehensive literature search to provide a complete overview of the recent advances in machine learning in the assessment of melanoma based on hematoxylin eosin digital pathology images. In our work, we review 37 recent publications, compare the methods and performance of the reviewed studies, and highlight the variety of promising machine-learning applications in melanoma histology.

show abstract

“…As the backbone of the teacher and the student model, we choose the SeaNet (with VGG16) proposed in [18]. The use of this framework demonstrated the improvement over standard methods for detecting tumor regions in histological images.…”

Section: Optimization Of the Teacher-student Parametersmentioning

confidence: 99%

A Self-Contrastive Learning Framework for Skin Cancer Detection Using Histological Images

Amor¹,

Colomer²,

Morales³

et al. 2022

2022 IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

Cutaneous spindle cell (CSC) neoplasms are a group of tumors that represent a formidable diagnostic challenge for dermatopathologists. Digital pathology has enabled the application of new methods based on artificial intelligence to reduce the workload of pathologists' daily practice. In this work, we propose a self-learning framework to detect tumor regions in histological images. The use of a teacher-model paradigm increases the annotated database while avoiding manual annotation. The pre-trained latent space of this model is then used in a second stage by another model to differentiate between leiomyomas (benign cases) and leiomyosarcomas (malignant cases). A contrastive learning approach allows separating the latent space of samples from different classes. This framework was tested on an independent database. This novel approach supposes a step forward in the CSC detection as the obtained results suggest (Acc = 0.90 and 0.8451, respectively).

show abstract

An attention-based weakly supervised framework for spitzoid melanocytic lesion diagnosis in whole slide images

Cited by 30 publications

References 34 publications

A Self-Training Weakly-Supervised Framework for Pathologist-Like Histopathological Image Analysis

A Self-Training Weakly-Supervised Framework for Pathologist-Like Histopathological Image Analysis

Diagnostic and Prognostic Deep Learning Applications for Histological Assessment of Cutaneous Melanoma

A Self-Contrastive Learning Framework for Skin Cancer Detection Using Histological Images

Contact Info

Product

Resources

About