Novel MRI-Based CAD System for Early Detection of Thyroid Cancer Using Multi-Input CNN

Can texture analysis of T2-weighted MRI be used to predict extrathyroidal extension in papillary thyroid carcinoma?

Determining the presence of extrathyroidal extension (ETE) is important for established of different surgical protocol and postoperative patient management in patients with papillary thyroid carcinoma (PTC). The correlation relationship between texture features from T2-weighted imaging (T2WI) and ETE has not been explored extensively. This study aimed to explore the value of T2-weighted magnetic resonance imaging - based whole tumor texture analysis in predict extrathyroidal extension with PTC. In this retrospectively study, 76 patients with pathologically proven PTC were recruited, who received surgical resection and underwent preoperative thyroid magnetic resonance imaging. Based on histo-pathologically findings, patients were classified into ETE and no ETE groups. ETE group was further divided into 2 subgroups (minimal ETE and extensive ETE). Whole-tumor texture analysis was independently performed by 2 radiologists on axial T2WI images. Nine histogram and gray-level co-occurrence matrix (GLCM) texture features were automatically extracted. Univariate and multivariate analysis were performed to determine risk factors associated with ETE. Predictive performance was evaluated by receiver operating characteristic (ROC) analysis. Interobserver agreement, confirmed by intraclass correlation coefficients (ICCs) ranging from 0.78 to 0.89, was excellent for texture analysis between 2 radiologists. T2WI image derived entropy, standard deviation, energy and correlation have significant difference between PTC with and without ETE (all P < .05). Among these, entropy showed the best diagnostic efficiency with the area under ROC curve of 0.837, diagnostic threshold of 5.86, diagnostic sensitivity and specificity of 81.5% and 75.6%, respectively. Additionally, the multivariate analysis revealed that high entropy was an independent risk factor of ETE (odds ratio, OR = 19.348; 95%CI, 4.578-81.760; P = .001). The findings indicate a significant association between texture features of the primary tumor based on T2WI and the presence of ETE in PTC. These results have the potential to help predict ETE preoperatively in patients with PTC, offering valuable insights for clinical decision-making.

Special Issue "Computer Aided Diagnosis Sensors"

Sensors used to diagnose, monitor or treat diseases in the medical domain are known as medical sensors [...].

Automated Recognition of Cancer Tissues through Deep Learning Framework from the Photoacoustic Specimen

The fast advancement of biomedical research technology has expanded and enhanced the spectrum of diagnostic instruments. Various research groups have found optical imaging, ultrasonic imaging, and magnetic resonance imaging to create multifunctional devices that are critical for biomedical activities. Multispectral photoacoustic imaging that integrates the ideas of optical and ultrasonic technologies is one of the most essential instruments. At the same time, early cancer identification is becoming increasingly important in order to minimize fatality. Deep learning (DL) techniques have recently advanced to the point where they can be used to diagnose and classify cancer using biological images. This paper describes a hybrid optimization method that combines in-depth transfer learning-based cancer detection with multispectral photoacoustic imaging. The goal of the PS-ACO-RNN approach is to use ultrasound images to detect and classify the presence of cancer. Bilateral filtration (BF) is often used as a noise removal approach in image processing. In addition, lightweight LEDNet models are used to separate the biological images. A feature extractor with particle swarm with ant colony optimization (PS-ACO) paradigm can also be used. Finally, biological images assign appropriate class labels using a recurrent neural network (RNN) model. The effectiveness of the PS-ACO-RNN technique is verified using a benchmark database, and test results show that the PS-ACO-RNN approach works better than current approaches.

Automatic symptoms identification from a massive volume of unstructured medical consultations using deep neural and BERT models

Automatic symptom identification plays a crucial role in assisting doctors during the diagnosis process in Telemedicine. In general, physicians spend considerable time on clinical documentation and symptom identification, which is unfeasible due to their full schedule. With text-based consultation services in telemedicine, the identification of symptoms from a user's consultation is a sophisticated process and time-consuming. Moreover, at Altibbi, which is an Arabic telemedicine platform and the context of this work, users consult doctors and describe their conditions in different Arabic dialects which makes the problem more complex and challenging. Therefore, in this work, an advanced deep learning approach is developed consultations with multi-dialects. The approach is formulated as a multi-label multi-class classification using features extracted based on AraBERT and fine-tuned on the bidirectional long short-term memory (BiLSTM) network. The Fine-tuning of BiLSTM relies on features engineered based on different variants of the bidirectional encoder representations from transformers (BERT). Evaluating the models based on precision, recall, and a customized hit rate showed a successful identification of symptoms from Arabic texts with promising accuracy. Hence, this paves the way toward deploying an automated symptom identification model in production at Altibbi which can help general practitioners in telemedicine in providing more efficient and accurate consultations.

Magnetic Fields and Cancer: Epidemiology, Cellular Biology, and Theranostics

Humans are exposed to a complex mix of man-made electric and magnetic fields (MFs) at many different frequencies, at home and at work. Epidemiological studies indicate that there is a positive relationship between residential/domestic and occupational exposure to extremely low frequency electromagnetic fields and some types of cancer, although some other studies indicate no relationship. In this review, after an introduction on the MF definition and a description of natural/anthropogenic sources, the epidemiology of residential/domestic and occupational exposure to MFs and cancer is reviewed, with reference to leukemia, brain, and breast cancer. The in vivo and in vitro effects of MFs on cancer are reviewed considering both human and animal cells, with particular reference to the involvement of reactive oxygen species (ROS). MF application on cancer diagnostic and therapy (theranostic) are also reviewed by describing the use of different magnetic resonance imaging (MRI) applications for the detection of several cancers. Finally, the use of magnetic nanoparticles is described in terms of treatment of cancer by nanomedical applications for the precise delivery of anticancer drugs, nanosurgery by magnetomechanic methods, and selective killing of cancer cells by magnetic hyperthermia. The supplementary tables provide quantitative data and methodologies in epidemiological and cell biology studies. Although scientists do not generally agree that there is a cause-effect relationship between exposure to MF and cancer, MFs might not be the direct cause of cancer but may contribute to produce ROS and generate oxidative stress, which could trigger or enhance the expression of oncogenes.

Spherical harmonics to quantify cranial asymmetry in deformational plagiocephaly

Cranial deformation and deformational plagiocephaly (DP) in particular affect an important percentage of infants. The assessment and diagnosis of the deformation are commonly carried by manual measurements that provide low interuser accuracy. Another approach is the use of three-dimensional (3D) models. Nevertheless, in most cases, deformation measurements are carried out manually on the 3D model. It is necessary to develop methodologies for the detection of DP that are automatic, accurate and take profit on the high quantity of information of the 3D models. Spherical harmonics are proposed as a new methodology to identify DP from head 3D models. The ideal fitted ellipsoid for each head is computed and the orthogonal distances between head and ellipsoid are obtained. Finally, the distances are modelled using spherical harmonics. Spherical harmonic coefficients of degree 2 and order − 2 are identified as the correct ones to represent the asymmetry characteristic of DP. The obtained coefficient is compared to other anthropometric deformation indexes, such as Asymmetry Index, Oblique Cranial Length Ratio, Posterior Asymmetry Index and Anterior Asymmetry Index. The coefficient of degree 2 and order − 2 with a maximum degree of 4 is found to provide better results than the commonly computed anthropometric indexes in the detection of DP.

Artificial Intelligence and Deep Learning of Head and Neck Cancer

Artificial intelligence (AI) algorithms, particularly deep learning, have developed to the point that they can be applied in image recognition tasks. The use of AI in medical imaging can guide radiologists to more accurate image interpretation and diagnosis in radiology. The software will provide data that we cannot extract from the images. The rapid development in computational capabilities supports the wide applications of AI in a range of cancers. Among those are its widespread applications in head and neck cancer.