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

Abstract

Early detection of thyroid nodules can greatly contribute to the prediction of cancer burdening and the steering of personalized management. We propose a novel multimodal MRI-based computer-aided diagnosis (CAD) system that differentiates malignant from benign thyroid nodules. The proposed CAD is based on a novel convolutional neural network (CNN)-based texture learning architecture. The main contribution of our system is three-fold. Firstly, our system is the first of its kind to combine T2-weighted MRI and apparent diffusion coefficient (ADC) maps using a CNN to model thyroid cancer. Secondly, it learns independent texture features for each input, giving it more advanced capabilities to simultaneously extract complex texture patterns from both modalities. Finally, the proposed system uses multiple channels for each input to combine multiple scans collected into the deep learning process using different values of the configurable diffusion gradient coefficient. Accordingly, the proposed system would enable the learning of more advanced radiomics with an additional advantage of visualizing the texture patterns after learning. We evaluated the proposed system using data collected from a cohort of 49 patients with pathologically proven thyroid nodules. The accuracy of the proposed system has also been compared against recent CNN models as well as multiple machine learning (ML) frameworks that use hand-crafted features. Our system achieved the highest performance among all compared methods with a diagnostic accuracy of 0.87, specificity of 0.97, and sensitivity of 0.69. The results suggest that texture features extracted using deep learning can contribute to the protocols of cancer diagnosis and treatment and can lead to the advancement of precision medicine.

Spherical harmonics to quantify cranial asymmetry in deformational plagiocephaly

“…However, as stated above, the maximum differentiation in absolute values of f 2 –2 is maximum for lmax = 4. Minimizing the RMSE with higher lmax degrees, confirms why other researchers opt for using maximum l-values of 100 43 , 1024 29 or even an infinite set of harmonic functions 26 , 27 . Our understanding for reaching a global complexity value, instead of regional or local complexities 29 , confirms the expections of differentiating between regular heads and heads with DP using a relatively low lmax value of 4.…”

“…In the field of medicine, spherical harmonics are frequently used as part of recent efforts to develop computer-aided, non-invasive methodologies to predict tumour growth behaviour, on the hypothesis that malignant tumours tend to have more irregular shapes in comparison to benign tumours. An example of this application is the use of spherical harmonics for the development of a computer-aided methodology for the diagnosis of thyroid cancer 26 , 27 . The contrast between the approach used in the proposed paper for cranial asymmetry quantification and the approach for diagnosis of malignant tumours can be of great value, due to the different spherical harmonics approach undertaken herein.…”

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

“…In contrast, and at the algorithmic level, various deep learning models were widely used in medical NLP. Such as: The BiLSTM/LSTM in [14] , [16] , [20] , [21] , [22] , [23] , [24] , the convolutional neural network (CNN) in [19] , [25] , the capsule network [26] , the transformers [27] , the ResNet-34 network [28] , and the generative adversarial network (GAN) [29] . Medical symptom extraction is a well-known problem in health or medical-related NLP tasks.…”

Automatic Symptoms Identification from Massive Volume of Unstructured Medical Consultations Using Deep Neural and BERT Models

“…In contrast, and at the algorithmic level, various deep learning models were widely used in medical NLP. Such as: The BiLSTM/LSTM in [14,16,20,21,22,23,24], the convolutional neural network (CNN) in [19,25], the capsule network [26], the transformers [27], the ResNet-34 network [28], and the generative adversarial network (GAN) [29]. Medical symptom extraction is a wellknown problem in health or medical-related NLP tasks.…”