Artificial intelligence for cervical cancer screening: Scoping review, 2009-2022.

BACKGROUND: The intersection of artificial intelligence (AI) with cancer research is increasing, and many of the advances have focused on the analysis of cancer images. OBJECTIVES: To describe and synthesize the literature on the diagnostic accuracy of AI in early imaging diagnosis of cervical cancer following Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews (PRISMA-ScR). SEARCH STRATEGY: Arksey and O'Malley methodology was used and PubMed, Scopus, and Google Scholar databases were searched using a combination of English and Spanish keywords. SELECTION CRITERIA: Identified titles and abstracts were screened to select original reports and cross-checked for overlap of cases. DATA COLLECTION AND ANALYSIS: A descriptive summary was organized by the AI algorithm used, total of images analyzed, data source, clinical comparison criteria, and diagnosis performance. MAIN RESULTS: We identified 32 studies published between 2009 and 2022. The primary sources of images were digital colposcopy, cervicography, and mobile devices. The machine learning/deep learning (DL) algorithms applied in the articles included support vector machine (SVM), random forest classifier, k-nearest neighbors, multilayer perceptron, C4.5, Naïve Bayes, AdaBoost, XGboots, conditional random fields, Bayes classifier, convolutional neural network (CNN; and variations), ResNet (several versions), YOLO+EfficientNetB0, and visual geometry group (VGG; several versions). SVM and DL methods (CNN, ResNet, VGG) showed the best diagnostic performances, with an accuracy of over 97%. CONCLUSION: We concluded that the use of AI for cervical cancer screening has increased over the years, and some results (mainly from DL) are very promising. However, further research is necessary to validate these findings.

Women's critical experiences with the pap smear for the development of cervical cancer screening devices.

Global cervical cancer incidence and mortality have remained a major public health problem. Depending on the quality and coverage of preventive programs, and the capacity of health care systems, different screening tests are used, with the Pap smear being the most widely implemented. Several difficulties have been reported in accessing timely detection, causing late cervical cancer diagnosis. Trying to close these gaps, new screening devices have been developed in recent years; however, there is a lack of knowledge about whether or not women perspective has been included in the design process and technological development of these devices. This scoping review aimed to describe and synthesize scientific literature on women's critical experiences with Pap smears to prospectively contribute to the design, development, and scale-up of cervical cancer screening devices. The electronic databases Web of Science, Scopus, PubMed, PsycINFO and SciELO were searched for relevant studies published between 2012 and 2021; finally, 18 qualitative studies were included. Experiences were classified into four categories: fear and embarrassment, speculum pain and discomfort, outcome distress and health service barriers. Critical experiences before, during, and after the test were analyzed in turn. In particular, during the test, women reported pain associated with the use of the speculum. The acceptability of new screening devices will largely depend on comfort during the test and timely delivery of results. This review provides a useful qualitative synthesis, not only to advance the design of novel devices but also for future implementation research in cervical screening services.

Sparse representation of MER signals for localizing the Subthalamic Nucleus in Parkinson's disease surgery.

Deep brain stimulation (DBS) of Subthalamic Nucleus (STN) is the best method for treating advanced Parkinson's disease (PD), leading to striking improvements in motor function and quality of life of PD patients. During DBS, online analysis of microelectrode recording (MER) signals is a powerful tool to locate the STN. Therapeutic outcomes depend of a precise positioning of a stimulator device in the target area. In this paper, we show how a sparse representation of MER signals allows to extract discriminant features, improving the accuracy in identification of STN. We apply three techniques for over-complete representation of signals: Method of Frames (MOF), Best Orthogonal Basis (BOB) and Basis Pursuit (BP). All the techniques are compared to classical methods for signal processing like Wavelet Transform (WT), and a more sophisticated method known as adaptive Wavelet with lifting schemes (AW-LS). We apply each processing method in two real databases and we evaluate its performance with simple supervised classifiers. Classification outcomes for MOF, BOB and BP clearly outperform WT and AW-LF in all classifiers for both databases, reaching accuracy values over 98%.

Unsupervised learning applied in MER and ECG signals through Gaussians mixtures with the Expectation-Maximization algorithm and Variational Bayesian Inference.

Automatic identification of biosignals is one of the more studied fields in biomedical engineering. In this paper, we present an approach for the unsupervised recognition of biomedical signals: Microelectrode Recordings (MER) and Electrocardiography signals (ECG). The unsupervised learning is based in classic and bayesian estimation theory. We employ gaussian mixtures models with two estimation methods. The first is derived from the frequentist estimation theory, known as Expectation-Maximization (EM) algorithm. The second is obtained from bayesian probabilistic estimation and it is called variational inference. In this framework, both methods are used for parameters estimation of Gaussian mixtures. The mixtures models are used for unsupervised pattern classification, through the responsibility matrix. The algorithms are applied in two real databases acquired in Parkinson's disease surgeries and electrocardiograms. The results show an accuracy over 85% in MER and 90% in ECG for identification of two classes. These results are statistically equal or even better than parametric (Naive Bayes) and nonparametric classifiers (K-nearest neighbor).

NEUROZONE: on-line recognition of brain structures in stereotactic surgery--application to Parkinson's disease.

The success of stereotactic surgery for Deep Brain Stimulation depends critically on the exact positioning of a microelectrode recording in a target area of the brain. This paper presents the software system NEUROZONE composed of two main applications: first, it allows online recognition of brain structures by the analysis of signals from microelectrode recordings (MER), and second, it processes and analyses off-line databases allowing the inclusion of new trained classifiers for automatic identification. The software serves as a support to the analysis done by a medical specialist during surgery, and seeks to reduce the adverse side effects that may occur because of inadequate identification of the target areas. The software also allows the specialists to label recordings obtained during surgery, in order to generate a new off-line database or increase the amount of records in an already existing off-line database. NEUROZONE has been tested for Deep Brain Stimulation performed at the Institute for Epilepsy and Parkinson of the Eje Cafetero (Colombia), achieving positive identifications of the Subthalamic Nucleus (STN) over to 85% using a naive Bayes classifier.

Multi-patient learning increases accuracy for Subthalamic Nucleus identification in deep brain stimulation.

Establishing the exact position of basal ganglia is key in several brain surgeries, particularly in deep brain stimulation for patients suffering from Parkinson's disease. There have been recent attempts to introduce automatic systems with the ability to localize, with high accuracy, specific brain regions. These systems usually follow the classical supervised learning paradigm, in which training data from different patients are employed to construct a classifier that is patient-independent. In this paper, we show how by sharing information from different patients, it is possible to increase accuracy for targeting the Subthalamic Nucleus. We do this in the context of multi-task learning, where different but related tasks are used simultaneously to leverage the performance of a learning system. Results show that the multitask framework can outperform the traditional patient-independent scenario in two different real datasets.