Yasunari Matsuzaka

This review summarizes the current advances, applications, and research prospects of computer vision in advancing medical imaging. Computer vision in healthcare has revolutionized medical practice by increasing diagnostic accuracy, improving patient care, and increasing operational efficiency. Likewise, deep learning algorithms have advanced medical image analysis, significantly improved healthcare outcomes and transforming diagnostic processes. Specifically, convolutional neural networks are crucial for modern medical image segmentation, enabling the accurate, efficient analysis of various imaging modalities and helping enhance computer-aided diagnosis and treatment planning. Computer vision algorithms have demonstrated remarkable capabilities in detecting various diseases. Artificial intelligence (AI) systems can identify lung nodules in chest computed tomography scans at a sensitivity comparable to that of experienced radiologists. Computer vision can analyze brain scans to detect problems such as aneurysms and tumors or areas affected by diseases such as Alzheimer's. In summary, computer vision in medical imaging is significantly improving diagnostic accuracy, efficiency, and patient outcomes across a range of medical specialties.

Cancer gene therapy is attracting considerable attention as a new treatment method for overcoming intractable cancers. CAR-T cell therapy has already achieved remarkable results, particularly for hematological tumors. Because CAR-T cells can increase within the body, they have the advantage of requiring only a single administration. In addition, CAR-T cell therapy targeting the CD19 antigen has been established for relapsed or refractory disease in young people with CD19-positive acute B-cell leukemia (B-acute lymphoblastic leukemia, B-ALL) and diffuse large B-cell lymphoma (DLBCL). In addition to CAR-T cell therapy, oncolytic viruses represent a promising approach for cancer treatment, with some already in clinical use and others being researched for their potential benefits. These viruses infect and kill cancer cells, triggering an immune response that helps the body recognize and fight cancer. Oncolytic virus therapy is a form of immunotherapy that uses modified viruses to target and destroy tumor cells while potentially stimulating antitumor immune responses. These viruses have shown promising activity in clinical trials, with some approved for specific cancers like melanoma. Research is ongoing to improve their efficacy, expand their use to other cancer types, and overcome the logistical challenges associated with their delivery. Gene therapy can potentially treat diseases caused by recessive gene disorders like cystic fibrosis, hemophilia, muscular dystrophy, and sickle cell anemia, as well as acquired genetic diseases, such as cancer and viral infections like acquired immunodeficiency syndrome (AIDS).

The molecular regulation and therapeutic applications of brain-derived neurotrophic factor (BDNF)–tropomyosin-related kinase B (TrkB) signaling in major depressive disorder (MDD) through interaction with vascular endothelial growth factor (VEGF) and N-methyl-D-aspartic acid (NMDA) receptors show promise. While BDNF-TrkB signaling is implicated in antidepressant action, the association between BDNFs and depression has not yielded conclusive results. Some studies show decreased BDNF levels in depression, while others indicate that increased BDNF expression in certain brain regions can induce depression susceptibility. The role of BDNFs varies across different brain regions, necessitating further study of individual mechanisms. This regional variability complicates the development of targeted therapies. The antidepressant-like and neurotrophic actions of BDNFs require VEGF signaling, but there is also a reciprocal interdependence, as VEGF actions are dependent on BDNFs. This complex relationship complicates the development of targeted therapies.

Computer vision and artificial intelligence have revolutionized the field of pathological image analysis, enabling faster and more accurate diagnostic classification. Deep learning architectures like convolutional neural networks (CNNs), have shown superior performance in tasks such as image classification, segmentation, and object detection in pathology. Computer vision has significantly improved the accuracy of disease diagnosis in healthcare. By leveraging advanced algorithms and machine learning techniques, computer vision systems can analyze medical images with high precision, often matching or even surpassing human expert performance. In pathology, deep learning models have been trained on large datasets of annotated pathology images to perform tasks such as cancer diagnosis, grading, and prognostication. While deep learning approaches show great promise in diagnostic classification, challenges remain, including issues related to model interpretability, reliability, and generalization across diverse patient populations and imaging settings.

Human Leukocyte Antigen (HLA) is like a device that monitors the internal environment of the body. T lymphocytes immediately recognize the HLA molecules that are expressed on the surface of the cells of the different individual, attacking it defeats microorganisms that is one of the causes of rejection in organ transplants performed between people with unmatched HLA types. Over 2850 and 3580 different polymorphisms have been reported for HLA-A and HLA-B respectively, around the world. HLA genes are associated with the risk of developing a variety of diseases, including autoimmune diseases, and play an important role in pathological conditions. By using a deep learning method called multi-task learning to simultaneously predict the gene sequences of multiple HLA genes, it is possible to improve accuracy and shorten execution time. Some new systems use a model called convolutional neural network (CNNs) in deep learning, which uses neural networks consisting of many layers and can learn complex correlations between SNP information and HLA gene sequences based on reference data for HLA imputation, which serves as training data. The learned model can output predicted values of HLA gene sequences with high accuracy using SNP information as input. To investigate which part of the input information surrounding the HLA gene is used to make learning predictions, predictions were made using not only a small number of nearby SNP information but also many SNP information distributed over a wider area by visualizing the learning information of the model. While conventional methods are strong at learning using nearly SNP information and not good at learning using SNP information located at distant locations, some new systems are thought that prediction accuracy may have improved because this problem was overcome. HLA genes are involved in the onset of a variety of diseases and are attracting attention. As an important area from the perspective of elucidating pathological conditions and realizing personalized medicine. The applied multi-task learning to two different HLA imputation reference panels—a Japanese panel (n = 1118) and type I diabetes genetics consortium panel (n = 5122). Through 10-fold cross-validation on these panels, the multi-task learning achieved higher imputation accuracy than conventional methods, especially for imputing low-frequency and rare HLA alleles. The increased prediction accuracy of HLA gene sequences is expected to increase the reliability of HLA analysis, including integrated analysis between different racial populations, and is expected to greatly contribute to the identification of HLA gene sequences associated with diseases and further elucidation of pathological conditions.

Adeno-associated virus (AAV) is characterized by non-pathogenicity, long-term infection, and broad tropism and is actively developed as a vector virus for gene therapy products. AAV is classified into more than 100 serotypes based on differences in the amino acid sequence of the capsid protein. Endocytosis involves the uptake of viral particles by AAV and accessory receptors during AAV infection. After entry into the cell, they are transported to the nucleus through the nuclear pore complex. AAVs mainly use proteoglycans as receptors to enter cells, but the types of sugar chains in proteoglycans that have binding ability are different. Therefore, it is necessary to properly evaluate the primary structure of receptor proteins, such as amino acid sequences and post-translational modifications, including glycosylation, and the higher-order structure of proteins, such as the folding of the entire capsid structure and the three-dimensional (3D) structure of functional domains, to ensure the efficacy and safety of biopharmaceuticals. To further enhance safety, it is necessary to further improve the efficiency of gene transfer into target cells, reduce the amount of vector administered, and prevent infection of non-target cells.

In vaccine development, many use the spike protein (S protein), which has multiple “spike-like” structures protruding from the spherical structure of the coronavirus, as an antigen. However, there are concerns about its effectiveness and toxicity. When S protein is used in a vaccine, its ability to attack viruses may be weak, and its effectiveness in eliciting immunity will only last for a short period of time. Moreover, it may cause “antibody-dependent immune enhancement”, which can enhance infections. In addition, the three-dimensional (3D) structure of epitopes is essential for functional analysis and structure-based vaccine design. Additionally, during viral infection, large amounts of extracellular vesicles (EVs) are secreted from infected cells, which function as a communication network between cells and coordinate the response to infection. Under conditions where SARS-CoV-2 (severe acute respiratory syndrome coronavirus 2) molecular vaccination produces overwhelming SARS-CoV-2 spike glycoprotein, a significant proportion of the overproduced intracellular spike glycoprotein is transported via EVs. Therefore, it will be important to understand the infection mechanisms of SARA-CoV-2 via EV-dependent and EV-independent uptake into cells and to model the infection processes based on 3D structural features at interaction sites.

Mesenchymal stem cells (MSCs) have emerged as a promising approach for drug delivery strategies because of their unique properties. These strategies include stem cell membrane-coated nanoparticles, stem cell-derived extracellular vesicles, immunomodulatory effects, stem cell-laden scaffolds, and scaffold-free stem cell sheets. MSCs offer advantages such as low immunogenicity, homing ability, and tumor tropism, making them ideal for targeted drug delivery systems. Stem cell-derived extracellular vesicles have gained attention for their immune properties and tumor-homing abilities, presenting a potential solution for drug delivery challenges. The relationship between MSC-based drug delivery and the self-renewal and differentiation capabilities of MSCs lies in the potential of engineered MSCs to serve as effective carriers for therapeutic agents while maintaining their intrinsic properties. MSCs exhibit potent immunosuppressive functions in MSC-based drug delivery strategies. Stem cell-derived EVs have low immunogenicity and strong therapeutic potential for tissue repair and regeneration. Scaffold-free stem cell sheets represent a cutting-edge approach in regenerative medicine, offering a versatile platform for tissue engineering and regeneration across different medical specialties. MSCs have shown great potential for clinical applications in regenerative medicine because of their ability to differentiate into various cell types, secrete bioactive factors, and modulate immune responses. Researchers are exploring these innovative approaches to enhance drug delivery efficiency and effectiveness in treating various diseases.

Heparan sulfate proteoglycans are highly glycosylated proteins in which heparan sulfate, a glycosaminoglycan sugar chain, is an acidic sugar chain consisting of a repeating disaccharide structure of glucuronic acid and N-acetylglucosamine is locally sulfated. Syndecan, one of the transmembrane HSPGs, functions as a receptor that transmits signals from the extracellular microenvironment to the inside of the cell. In the vascular system, heparan sulfate proteoglycans, a major component of the glycocalyx, enable the binding of various plasma-derived molecules due to their diversity, epimerization of glycosaminoglycans chains, long chains, and sulfation. Heparan sulfate proteoglycans present in the extracellular matrix serve as a reservoir for bioactive molecules such as chemokines, cytokines, and growth factors. Aberrant expression of heparan sulfate proteoglycans, heparanase, and sulfatase is observed in many pathological conditions. Therefore, it can be applied to therapeutic strategies for a wide range of fields including Alzheimer’s disease, heart failure, cancer, organ transplants, diabetes, chronic inflammation, aging, and autoimmune diseases.

Computational intelligence (CI) uses applied computational methods for problem-solving inspired by the behavior of humans and animals. Biological systems are used to construct software to solve complex problems, and one type of such system is an artificial immune system (AIS), which imitates the immune system of a living body. AISs have been used to solve problems that require identification and learning, such as computer virus identification and removal, image identification, and function optimization problems. In the body’s immune system, a wide variety of cells work together to distinguish between the self and non-self and to eliminate the non-self. AISs enable learning and discrimination by imitating part or all of the mechanisms of a living body’s immune system. Certainly, some deep neural networks have exceptional performance that far surpasses that of humans in certain tasks, but to build such a network, a huge amount of data is first required. These networks are used in a wide range of applications, such as extracting knowledge from a large amount of data, learning from past actions, and creating the optimal solution (the optimization problem). A new technique for pre-training natural language processing (NLP) software ver.9.1by using transformers called Bidirectional Encoder Representations (BERT) builds on recent research in pre-training contextual representations, including Semi-Supervised Sequence Learning, Generative Pre-Training, ELMo (Embeddings from Language Models), which is a method for obtaining distributed representations that consider context, and ULMFit (Universal Language Model Fine-Tuning). BERT is a method that can address the issue of the need for large amounts of data, which is inherent in large-scale models, by using pre-learning with unlabeled data. An optimization problem involves “finding a solution that maximizes or minimizes an objective function under given constraints”. In recent years, machine learning approaches that consider pattern recognition as an optimization problem have become popular. This pattern recognition is an operation that associates patterns observed as spatial and temporal changes in signals with classes to which they belong. It involves identifying and retrieving predetermined features and rules from data; however, the features and rules here are not logical information, but are found in images, sounds, etc. Therefore, pattern recognition is generally conducted by supervised learning. Based on a new theory that deals with the process by which the immune system learns from past infection experiences, the clonal selection of immune cells can be viewed as a learning rule of reinforcement learning.