Effective Diagnosis and Treatment through Content-Based Medical Image Retrieval (CBMIR) by Using Artificial Intelligence

Abstract

Medical-image-based diagnosis is a tedious task‚ and small lesions in various medical images can be overlooked by medical experts due to the limited attention span of the human visual system, which can adversely affect medical treatment. However, this problem can be resolved by exploring similar cases in the previous medical database through an efficient content-based medical image retrieval (CBMIR) system. In the past few years, heterogeneous medical imaging databases have been growing rapidly with the advent of different types of medical imaging modalities. Recently, a medical doctor usually refers to various types of imaging modalities all together such as computed tomography (CT), magnetic resonance imaging (MRI), X-ray, and ultrasound, etc of various organs in order for the diagnosis and treatment of specific disease. Accurate classification and retrieval of multimodal medical imaging data is the key challenge for the CBMIR system. Most previous attempts use handcrafted features for medical image classification and retrieval, which show low performance for a massive collection of multimodal databases. Although there are a few previous studies on the use of deep features for classification, the number of classes is very small. To solve this problem, we propose the classification-based retrieval system of the multimodal medical images from various types of imaging modalities by using the technique of artificial intelligence, named as an enhanced residual network (ResNet). Experimental results with 12 databases including 50 classes demonstrate that the accuracy and F1.score by our method are respectively 81.51% and 82.42% which are higher than those by the previous method of CBMIR (the accuracy of 69.71% and F1.score of 69.63%).

Ultrasound Image-Based Diagnosis of Malignant Thyroid Nodule Using Artificial Intelligence

“…In the case of imbalance in the training data of the two classes such as the medical image processing system that is normally faced with the problem of data collection due to special characteristics of medical images, the binary cross-entropy function can produce bias in the trained classifier. To solve this problem, our proposed method uses a modified version of the binary cross-entropy, called the Weighted Binary Cross-Entropy (wBCE), as shown in Equation (6). As shown in this equation, we assigned different weight values to the losses caused by samples in each class in the binary cross-entropy function.…”

“…As a result, the diagnostic performance varies and is limited. With the development of digital technology, image-based diagnosis techniques have been widely used to help doctors investigate problems with organs that are underneath the skin and/or deep inside the human body [1][2][3][4][5][6][7][8][9][10][11]. For example, doctors have used X-ray imaging to capture lung and/or bone images that can help to indicate whether a disease/injury exists in these organs [9,10].…”

“…However, the use of captured images is still dependent on personal knowledge and experiences of doctors. To overcome this problem, Computer-Aided Diagnosis systems (CAD) have been developed to assist doctors in the diagnosis and treatment processes [1][2][3][4][5][6][7][8][9][10]. As indicated by its name, the CAD systems can serve as an additional expert in the double screening process that aims to enhance the human diagnostic performance based on a computer program [11].…”

Increasing Complexity to Simplify Clinical Care: High Resolution Mass Spectrometry as an Enabler of AI Guided Clinical and Therapeutic Monitoring

“…Lastly, large curated image databases, such as The Cancer Imaging Archive (TCIA), [ 26 ] provide quality control and regulatory clearance of cancer‐related medical images under standardized Digital Imaging and Communications in Medicine (DICOM) standards [ 27 ] for a wide range of data including chests X‐rays, computed tomography, ultrasound and magnetic resonance imaging. [ 28–32 ] Thus, for deep learning applications, a facilitated path from technical feasibility to medical outcomes adjudication has underpinned much of the current success. A recent clinical trial studying the use of deep learning to identify images of diabetic retinopathy has resulted in the first FDA approved use of an autonomous AI diagnostic.…”

Convolutional Neural Networks in Image Retrieval System

“…Recent years have shown that some advances in machine learning algorithms called 'Deep Learning (DL)' [10] is introduced for solving the above problems. The conventional machine learning methods have used "shallow" architectures, while deep learning act like human brain where it is organized as deep architecture and processes data in multiple stages of representation and transformation [11] [12]. There are many open-source libraries such as TensorFlow (https://www.tensorflow.org) and PyTorch (https://pytorch.org/ ) that allow the creation of deep neural nets in Python .…”