How much and where to use manual guidance in the computational detection of contours for histopathological images?

Stoean, Cătălin; Stoean, Ruxandra; Săndiță, Adrian; Meşină, Cristian; Gruia, Corina Lavinia; Ciobanu, Daniela

doi:10.1007/s00500-018-3029-9

Cited by 5 publications

(4 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this context, ensemble methods have been proposed in order to assist the pathologist in cancer grading, by providing an additional degree of confidence that highlights samples deserving further study [8]. Image segmentation by machine learning methods, and its relation to human experts' annotation is another remarkable direction [9]. Breast cancer multi-classification on the public BreakHis data set has been tackled by appointing an approach that performs semantic hierarchical feature learning followed by a convolutional neural network (CNN) [10].…”

Section: Plos Onementioning

confidence: 99%

A hybrid unsupervised—Deep learning tandem for electrooculography time series analysis

Stoean

Becerra-García

et al. 2020

PLoS ONE

Self Cite

View full text Add to dashboard Cite

Medical data are often tricky to get mined for patterns even by the generally demonstrated successful modern methodologies of deep learning. This paper puts forward such a medical classification task, where patient registers of two of the categories are sometimes hard to be distinguished because of samples showing characteristics of both labels in turn in several repetitions of the screening procedure. To this end, the current research appoints a pre-processing clustering step (through self-organizing maps) to group the data based on shape similarity and relabel it accordingly. Subsequently, a deep learning approach (a tandem of convolutional and long short-term memory networks) performs the training classification phase on the 'cleaned' samples. The dual methodology was applied for the computational diagnosis of electrooculography tests within spino-cerebral ataxia of type 2. The accuracy obtained for the discrimination into three classes was of 78.24%. The improvement that this duo brings over the deep learner alone does not stem from significantly higher accuracy results when the performance is considered for all classes. The major finding of this combination is that half of the presymptomatic cases were correctly found, in opposition to the single deep model, where this category was sacrificed by the learner in favor of a good accuracy overall. A high accuracy in general is desirable for any medical task, however the correct identification of cases before the symptoms become evident is more important.

show abstract

Section: Plos Onementioning

confidence: 99%

A hybrid unsupervised—Deep learning tandem for electrooculography time series analysis

Stoean

Becerra-García

et al. 2020

PLoS ONE

Self Cite

View full text Add to dashboard Cite

show abstract

“…Hematoxylin and eosin staining of tissues preserves important region-to-region morphology and allows microscopic evaluation of neighboring cell morphology, structure, and organization to derive a diagnosis. Despite using simplified algorithmic decision trees interobserver variation persists among experts [21,22]. To improve the accuracy and reduce discordance among specialists artificial intelligence (AI) algorithms based on deep neural networks (DNNs) have been recently developed and become a valuable tool for the generation and implementation of complex multi-parametric decision algorithms for pathological analysis [23] (Box 1).…”

Section: Intraoperative Histologymentioning

confidence: 99%

Cancer Surgery 2.0: Guidance by Real-Time Molecular Technologies

Ogrinc

Saudemont

Takáts

et al. 2021

Trends in Molecular Medicine

View full text Add to dashboard Cite

In vivo cancer margin delineation during surgery remains a major challenge. Despite the availability of several image guidance techniques and intraoperative assessment, clear surgical margins and debulking efficiency remain scarce. For this reason, there is particular interest in developing rapid intraoperative tools with high sensitivity and specificity to help guide cancer surgery in vivo. Recently, several emerging technologies including intraoperative mass spectrometry have paved the way for molecular guidance in a clinical setting. We evaluate these techniques and assess their relevance for intraoperative surgical guidance and how they can transform the future of molecular cancer surgery, diagnostics, patient management and care. HighlightsCancer surgery is the first treatment in many solid tumors. Quality of the initial surgery has a strong influence on the patient outcome. Tumors must be excised at best, but accurate delineation of their edges and margin definition is a huge challenge.

show abstract

“…The goal of automated medical image processing often involves their enhancement so as to facilitate the inspection of the images by medical staff. Moreover, computer aided diagnosis techniques can also be employed to help practitioners to diagnose a disease [1]. Therefore, an automated intelligent system is able to determine whether a patient is likely to suffer from the disease by analyzing the acquired medical image data.…”

Section: Introductionmentioning

confidence: 99%

Histopathological image analysis for breast cancer diagnosis by ensembles of convolutional neural networks and genetic algorithms

Molina‐Cabello

Rodríguez-Rodríguez

Thurnhofer‐Hemsi

et al. 2021

2021 International Joint Conference on Neural Networks (IJCNN)

View full text Add to dashboard Cite

One of the most invasive cancer types which affect women is breast cancer. Unfortunately, it exhibits a high mortality rate. Automated histopathological image analysis can help to diagnose the disease. Therefore, computer aided diagnosis by intelligent image analysis can help in the diagnosis tasks associated with this disease. Here we propose an automated system for histopathological image analysis that is based on deep learning neural networks with convolutional layers. Rather than a single network, an ensemble of them is built so as to attain higher recognition rates, which are obtained by computing a consensus decision from the individual networks of the ensemble. A final step involves the optimization of the set of networks that are included in the ensemble by a genetic algorithm. Experimental results are provided with a set of benchmark images, with favorable outcomes.Index Terms-convolutional neural networks, image classification, breast cancer, medical image processing, genetic algorithms This work is partially supported by the following Spanish grants: TIN2016-75097-P, RTI2018-094645-B-I00 and UMA18-FEDERJA-084. All of them include funds from the European Regional Development Fund (ERDF). The authors acknowledge the funding from the Universidad de Málaga. Karl Thurnhofer-Hemsi is funded by a Ph.D. scholarship from the Spanish Ministry of Education, Culture and Sport under the FPU program (FPU15/06512).

show abstract

How much and where to use manual guidance in the computational detection of contours for histopathological images?

Cited by 5 publications

References 17 publications

A hybrid unsupervised—Deep learning tandem for electrooculography time series analysis

A hybrid unsupervised—Deep learning tandem for electrooculography time series analysis

Cancer Surgery 2.0: Guidance by Real-Time Molecular Technologies

Histopathological image analysis for breast cancer diagnosis by ensembles of convolutional neural networks and genetic algorithms

Contact Info

Product

Resources

About