Preliminary findings on diagnostic performance of computed tomography perfusion images for intracranial arterial stenosis: a retrospective study.

OBJECTIVES: Computed tomographic perfusion (CTP) can play an auxiliary role in the selection of patients with acute ischemic stroke for endovascular treatment. However, data on CTP in non-stroke patients with intracranial arterial stenosis are scarce. We aimed to investigate images in patients with asymptomatic intracranial arterial stenosis to determine the detection accuracy and interpretation time of large/medium-artery stenosis or occlusion when combining computed tomographic angiography (CTA) and CTP images. METHODS: We retrospectively reviewed 39 patients with asymptomatic intracranial arterial stenosis from our hospital database from January 2021 to August 2023 who underwent head CTP, head CTA, and digital subtraction angiography (DSA). Head CTA images were generated from the CTP data, and the diagnostic performance for each artery was assessed. Two readers independently interpreted the CTA images before and after CTP, and the results were analyzed. RESULTS: After adding CTP maps, the accuracy (area under the curve) of diagnosing internal carotid artery (R1: 0.847 vs. 0.907, R2: 0.776 vs. 0.887), middle cerebral artery (R1: 0.934 vs. 0.933, R2: 0.927 vs. 0.981), anterior cerebral artery (R1: 0.625 vs. 0.750, R2: 0.609 vs. 0.750), vertebral artery (R1: 0.743 vs. 0.764, R2: 0.748 vs. 0.846), and posterior cerebral artery (R1: 0.390 vs. 0.575, R2: 0.390 vs. 0.585) occlusions increased for both readers (p < 0.05). Mean interpretation time (R1: 72.4 ± 6.1 s vs. 67.7 ± 6.4 s, R2: 77.7 ± 3.8 s vs. 72.6 ± 4.7 s) decreased when using a combination of both images both readers (p < 0.001). CONCLUSIONS: The addition of CTP images improved the accuracy of interpreting CTA images and reduced the interpretation time in asymptomatic intracranial arterial stenosis. These findings support the use of CTP imaging in patients with asymptomatic intracranial arterial stenosis.

Is Telemedicine Suitable for Patients with Chronic Inflammatory Skin Conditions? A Systematic Review.

Telemedicine, the provision of remote healthcare, has gained prominence, accelerated by the COVID-19 pandemic. It has the potential to replace routine in-person follow-up visits for patients with chronic inflammatory skin conditions. However, it remains unclear whether telemedicine can effectively substitute in-person consultations for this patient group. This systematic review assessed the effectiveness and safety of telemedicine compared with traditional in-person care for chronic inflammatory skin diseases. A comprehensive search in various databases identified 11 articles, including 5 randomized controlled trials (RCTs) and 1 clinical controlled trial (CCT). These studies evaluated telemedicine's impact on patients with psoriasis and atopic dermatitis, with varying methods like video consultations and digital platforms. The findings tentatively suggest that telemedicine does not seem to be inferior compared with in-person care, particularly in terms of condition severity and quality of life for patients with chronic inflammatory skin diseases. However, these results should be interpreted with caution due to the inherent uncertainties in the evidence. There are indications that telemedicine can offer benefits such as cost-effectiveness, time savings, and reduced travel distances, but it is important to recognize these findings as preliminary, necessitating further validation through more extensive research.

A novel pencil graphite electrode modified with an iron-based conductive metal-organic framework exhibited good ability in simultaneous sensing bisphenol A and bisphenol S.

Bisphenol A (BPA) and its substitute bisphenol S (BPS) are desirable materials widely used in manufacturing plastic products but can pose carcinogenic risks to humans. A new conductive iron-based metal-organic framework (Fe-HHTP)-modified pencil graphite electrode (PGE) for electrochemically sensing BPA and BPS was prepared and fully characterized by SEM, TEM, FT-IR, XRD, and XPS. Results showed that the optimal conditions for preparing Fe-HHTP/PGE were a pH of 6.5, a Fe-HHTP concentration of 2 mg·mL-1, a deposition potential of 0 V, and a deposition time of 100 s. The Fe-HHTP/PGE prepared under such conditions harbored a significant electrocatalytic activity with a detection limit of 0.8 nM for BPA and 1.7 nM for BPS (S/N = 3). Correspondingly, the electrochemical response current was linearly correlated to BPA and BPS, ranging from 0.01 to 100 µM. Fe-HHTP/PGE also obtained satisfactory recoveries by 93.8-102.1% and 96.0-101.3% for detecting BPA and BPS in plastic food packaging samples. Our work has provided a novel electrochemical tool to simultaneously detect BPA and BPS in food packaging samples and environmental matrixes.

An intelligent early warning system of analyzing Twitter data using machine learning on COVID-19 surveillance in the US.

The World Health Organization (WHO) declared on 11th March 2020 the spread of the coronavirus disease 2019 (COVID-19) a pandemic. The traditional infectious disease surveillance had failed to alert public health authorities to intervene in time and mitigate and control the COVID-19 before it became a pandemic. Compared with traditional public health surveillance, harnessing the rich data from social media, including Twitter, has been considered a useful tool and can overcome the limitations of the traditional surveillance system. This paper proposes an intelligent COVID-19 early warning system using Twitter data with novel machine learning methods. We use the natural language processing (NLP) pre-training technique, i.e., fine-tuning BERT as a Twitter classification method. Moreover, we implement a COVID-19 forecasting model through a Twitter-based linear regression model to detect early signs of the COVID-19 outbreak. Furthermore, we develop an expert system, an early warning web application based on the proposed methods. The experimental results suggest that it is feasible to use Twitter data to provide COVID-19 surveillance and prediction in the US to support health departments' decision-making.

Deletion at 12q12 increases the risk of developmental delay and intellectual disability.

Single-nucleotide polymorphism (SNP) arrays have been widely used to identify novel genomic imbalances. Many of these genomic imbalances have been confirmed to interact with developmental delays, intellectual disabilities (IDs), and congenital defects. Here, we identify a Chinese girl with a 3.18-Mb deletion at 12q12 (human genome build 19: 43,418,911-46,601,627) who showed postnatal growth delay, low-set ears, small hands and feet, widely spaced nipples, and blue sclerae. Deletions at 12q12 are extremely rare chromosomal imbalances; only four cases involving a deletion of this type have previously been reported. In these five sporadic cases, all of the patients exhibited developmental issues accompanied by different degrees of ID. A review of DECIPHER patient data revealed an additional six cases involving genomic deletion at 12q12. Many of the patients in these cases exhibited developmental delay and ID. When these patients were included, 91% and 73% of individuals with a deletion in this chromosomal region presented with developmental retardation and ID, respectively. Database searches indicated that this copy number variant (CNV) has not been found in normal humans. Therefore, we suggest that a CNV in this region is a risk factor for developmental retardation and ID.

High-Frequency Stimulation of the Subthalamic Nucleus Activates Motor Cortex Pyramidal Tract Neurons by a Process Involving Local Glutamate, GABA and Dopamine Receptors in Hemi-Parkinsonian Rats.

Deep brain stimulation (DBS) is widely used to treat advanced Parkinson's disease (PD). Here, we investigated how DBS applied on the subthalamic nucleus (STN) influenced the neural activity in the motor cortex. Rats, which had the midbrain dopaminergic neurons partially depleted unilaterally, called the hemi-Parkinsonian rats, were used as a study model. c-Fos expression in the neurons was used as an indicator of neural activity. Application of high-frequency stimulation (HFS) upon the STN was used to mimic the DBS treatment. The motor cortices in the two hemispheres of hemi-Parkinsonian rats were found to contain unequal densities of c-Fos-positive (Fos+) cells, and STN-HFS rectified this bilateral imbalance. In addition, STN-HFS led to the intense c-Fos expression in a group of motor cortical neurons which exhibited biochemical and anatomical characteristics resembling those of the pyramidal tract (PT) neurons sending efferent projections to the STN. The number of PT neurons expressing high levels of c-Fos was significantly reduced by local application of the antagonists of non-N-methyl-D-aspartate (non-NMDA) glutamate receptors, gammaaminobutyric acid A (GABAA) receptors and dopamine receptors in the upper layers of the motor cortex. The results indicate that the coincident activations of synapses and dopamine receptors in the motor cortex during STN-HFS trigger the intense expression of c-Fos of the PT neurons. The implications of the results on the cellular mechanism underlying the therapeutic effects of STN-DBS on the movement disorders of PD are also discussed.

Drop-by-drop chemical reaction and sample introduction for capillary electrophoresis.

In this paper, we report a novel sample introduction and chemical reaction strategy by drop-by-drop inkjet injection for an electrophoretically mediated microanalysis (EMMA). This method makes it possible to achieve an on-line introduction of reactant solutions by alternately ejecting small plugs, with an overlapping region of the plugs for mixing the reactants by electrophoresis, supporting chemical reactions, followed by electrophoretic separation of the final compounds. As a proof-of-concept of the method, the EMMA of an inkjetted mixture of 4-fluoro-7-nitrobenzofurazan (NBD-F) and amino acids was carried out as a model chemical reaction. The product NBD-amino acids were quantified by detection with laser induced fluorescence. The optimal conditions for the procedure were: inkjet driving voltage: +40-44 V; pulse width: 20-24 µs; drop-by-drop injection of reactant solutions: alternately 2 drops × 25 times for the amino acid solution and the NBD-F solution; zone overlapping voltage and time: 3 kV and 2 s; incubation time after overlapping: 5 min; separation voltage: 18 kV. Under the optimized conditions, a significant enhancement in sensitivity and a sensitive quantitative analysis were realized. The results obtained were comparable with those using the off-line labeling method. This method is rapid, cost-effective, and readily automated for EMMA.

Recovering Euclidean structure by conics and spheres: application to camera calibration.

A novel method of two-dimensional Euclidean structure recovery in one view from the projections of N parallel conics is proposed, which can be applied to camera calibration. Without considering the conic dual to the absolute points, we transform conic features from the homogeneous coordinates to the lifted coordinates. In the lifted space, the conic features have similar properties to the point or line features, which especially means that the homography can also be deduced by conic features directly. Our work gives a generic framework of recovering the Euclidean structure from conic features. A series of experiments with simulated and real data are conducted. The experiment results show that the proposed method has its validity in practical applications to camera calibration.

Recovering Euclidean structure from principal-axes paralleled conics: applications to camera calibration.

We focus on recovering the 2D Euclidean structure further for camera calibration from the projections of N parallel similar conics in this paper. This work demonstrates that the conic dual to the absolute points (CDAP) is the general form of the conic dual to the circular points, so it encodes the 2D Euclidean structure. However, the geometric size of the conic should be known if we utilize the CDAP. Under some special conditions (concentric conics), we proposed the rank-1 and rank-2 constraints. Our work relaxes the problem conditions and gives a more general framework than before. Experiments with simulated and real data are carried out to show the validity of the proposed algorithm.

A comparative study of GNN and MLP based machine learning for the diagnosis of Alzheimer's Disease involving data synthesis.

Alzheimer's Disease (AD) is a neurodegenerative disease that commonly occurs in older people. It is characterized by both cognitive and functional impairment. However, as AD has an unclear pathological cause, it can be hard to diagnose with confidence. This is even more so in the early stage of Mild Cognitive Impairment (MCI). This paper proposes a U-Net based Generative Adversarial Network (GAN) to synthesize fluorodeoxyglucose - positron emission tomography (FDG-PET) from magnetic resonance imaging - T1 weighted imaging (MRI-T1WI) for further usage in AD diagnosis including its early-stage MCI. The experiments have displayed promising results with Structural Similarity Index Measure (SSIM) reaching 0.9714. Furthermore, three types of classifiers are developed, i.e., one Multi-Layer Perceptron (MLP) based classifier, two Graph Neural Network (GNN) based classifiers where one is for graph classification and the other is for node classification. 10-fold cross-validation has been conducted on all trials of experiments for classifier comparison. The performance of these three types of classifiers has been compared with the different input modalities setting and data fusion strategies. The results have shown that GNN based node classifier surpasses the other two types of classifiers, and has achieved the state-of-the-art (SOTA) performance with the best accuracy at 90.18% for 3-class classification, namely AD, MCI and normal control (NC) with the synthesized fluorodeoxyglucose - positron emission tomography (FDG-PET) features fused at the input level. Moreover, involving synthesized FDG-PET as part of the input with proper data fusion strategies has also proved to enhance all three types of classifiers' performance. This work provides support for the notion that machine learning-derived image analysis may be a useful approach to improving the diagnosis of AD.

Bilirubin influences the predictive effect of body mass index on hospital mortality in critically ill patients.

Introduction: Body mass index (BMI) can predict mortality in critically ill patients. Moreover, mortality is related to increased bilirubin levels. Thus, herein, we aimed to investigate the effect of bilirubin levels on the usefulness of BMI in predicting mortality in critically ill patients. Methods: Data were extracted from the Medical Information Mart for Intensive Care (MIMIC IV) database. Patients were divided into two groups according to their total bilirubin levels within 24 h. Cox proportional hazard regression models were applied to obtain adjusted hazard ratios and 95 % confidence intervals for the correlation between BMI categories and hospital mortality. The dose-response relationship was flexibly modeled using a restricted cubic spline (RCS) with three knots. Results: Of the 14376 patients included, 3.4 % were underweight, 29.3 % were of normal body weight, 32.2 % were overweight, and 35.1 % were obese. For patients with total bilirubin levels <2 mg/dL, hospital mortality was significantly lower in patients with obesity than in normal body weight patients (p < 0.05). However, the opposite results were observed for patients with total bilirubin levels ≥2 mg/dL. The Cox proportional hazard regression models suggested that the risk of death was lower in patients with overweightness and obesity than in normal body weight patients when the total bilirubin levels were <2 mg/dL, but not in the other case (total bilirubin levels ≥2 mg/dL). RCS analyses showed that, for patients with total bilirubin levels <2 mg/dL, the risk of death gradually decreased with increasing BMI. Conversely, for patients with total bilirubin levels ≥2 mg/dL, this risk did not decrease with increasing BMI until reaching obesity, after which it increased rapidly. Conclusion: BMI predicted the risk of death differently in critically ill patients with different bilirubin levels.

Nobiletin Regulates Polyinosinic-polycytidylic Acid-induced Inflammation in Macrophages Partially via the PPAR-γ Signaling Pathway.

INTRODUCTION: Macrophage dysregulation is a common pathogenic feature of viruses that provides extensive targets for antiviral therapy. Nobiletin, a polymethoxylated flavonoid found in citrus fruits, has a multitude of effects. METHODS: We investigated the effect of nobiletin on polyinosinic-polycytidylic acid (poly(I:C))-induced inflammation in RAW264.7 cells. Nobiletin inhibited the production of poly(I:C)-induced inflammatory factors, including tumor necrosis factor (TNF)-α, interleukin (IL)-6, and CXCL10. High-throughput sequencing revealed that nobiletin inhibited the expression of TNF-α, IL-6, and CXCL10 and promoted the expression of CD206, Chil3, and Vcam1. In the Kyoto Encyclopedia of Genes and Genomes enrichment analyses, the upregulated differential genes were significantly enriched in the peroxisome proliferator-activated receptor (PPAR) signaling pathway. RESULTS: The PPAR-γ inhibitor T0070907 significantly reversed the inhibitory effects of nobiletin on IL-6 and CXCL10 but had no significant effect on TNF-α secretion. CONCLUSION: Thus, nobiletin regulated poly(I:C)-induced inflammatory responses in RAW264.7 cells partially via the PPAR-γ signaling pathway.

Increased serum carboxylesterase-1 levels are associated with metabolic dysfunction associated steatotic liver disease and metabolic syndrome in children with obesity.

BACKGROUND: Carboxylesterase 1(CES1) is expressed mainly in the liver and adipose tissue and is highly hypothesized to play an essential role in metabolism. Our study aimed to investigate the association between CES1 and metabolic syndrome (MetS) and metabolic dysfunction associated steatotic liver disease (MASLD) in children with obesity in China. METHODS: This study included 72 children with obesity aged 6-13years (including 25(35%) diagnosed as MetS and 36(50%) diagnosed as MASLD). All subjects were measured in anthropometry, serum level of biochemical parameters related to obesity, circumstance levels of insulin-like growth factor1, adipokines (adiponectin, leptin and growth differentiation factor 15) and CES1. RESULTS: Higher serum CES1 level were found in the MetS group (P = 0.004) and the MASLD group (P < 0.001) of children with obesity. Serum CES1 levels were positively correlated with alanine aminotransferase, aspartate aminotransferase, triglyceride, cholesterol, low-density lipoprotein cholesterol, GDF15, Leptin and negatively correlated with high-density lipoprotein cholesterol, adiponectin and IGF1. We also found a multivariable logistic regression analysis of MASLD and MetS predicted by CES1 significantly (MASLD P < 0.01, MetS P < 0.05). The combination of CES1, sex, age and BMI Z-score showed a sensitivity and specificity of 92.7% for the identification of MASLD and 78.6% for the identification of MetS. The cutoff for CES1 of MASLD is 56.30 ng/mL and of MetS is 97.79 ng/mL. CONCLUSIONS: CES1 is associated with an increasing risk of MetS and MASLD and can be established as a biomarker for metabolic syndrome and MASLD of children with obesity.

A Systematic Review of Application Progress on Machine Learning-Based Natural Language Processing in Breast Cancer over the Past 5 Years.

Artificial intelligence (AI) has been steadily developing in the medical field in the past few years, and AI-based applications have advanced cancer diagnosis. Breast cancer has a massive amount of data in oncology. There has been a high level of research enthusiasm to apply AI techniques to assist in breast cancer diagnosis and improve doctors' efficiency. However, the wise utilization of tedious breast cancer-related medical care is still challenging. Over the past few years, AI-based NLP applications have been increasingly proposed in breast cancer. In this systematic review, we conduct the review using preferred reporting items for systematic reviews and meta-analyses (PRISMA) and investigate the recent five years of literature in natural language processing (NLP)-based AI applications. This systematic review aims to uncover the recent trends in this area, close the research gap, and help doctors better understand the NLP application pipeline. We first conduct an initial literature search of 202 publications from Scopus, Web of Science, PubMed, Google Scholar, and the Association for Computational Linguistics (ACL) Anthology. Then, we screen the literature based on inclusion and exclusion criteria. Next, we categorize and analyze the advantages and disadvantages of the different machine learning models. We also discuss the current challenges, such as the lack of a public dataset. Furthermore, we suggest some promising future directions, including semi-supervised learning, active learning, and transfer learning.

Natural Language Processing Applications for Computer-Aided Diagnosis in Oncology.

In the era of big data, text-based medical data, such as electronic health records (EHR) and electronic medical records (EMR), are growing rapidly. EHR and EMR are collected from patients to record their basic information, lab tests, vital signs, clinical notes, and reports. EHR and EMR contain the helpful information to assist oncologists in computer-aided diagnosis and decision making. However, it is time consuming for doctors to extract the valuable information they need and analyze the information from the EHR and EMR data. Recently, more and more research works have applied natural language processing (NLP) techniques, i.e., rule-based, machine learning-based, and deep learning-based techniques, on the EHR and EMR data for computer-aided diagnosis in oncology. The objective of this review is to narratively review the recent progress in the area of NLP applications for computer-aided diagnosis in oncology. Moreover, we intend to reduce the research gap between artificial intelligence (AI) experts and clinical specialists to design better NLP applications. We originally identified 295 articles from the three electronic databases: PubMed, Google Scholar, and ACL Anthology; then, we removed the duplicated papers and manually screened the irrelevant papers based on the content of the abstract; finally, we included a total of 23 articles after the screening process of the literature review. Furthermore, we provided an in-depth analysis and categorized these studies into seven cancer types: breast cancer, lung cancer, liver cancer, prostate cancer, pancreatic cancer, colorectal cancer, and brain tumors. Additionally, we identified the current limitations of NLP applications on supporting the clinical practices and we suggest some promising future research directions in this paper.

Fractional Tensor Recurrent Unit (fTRU): A Stable Forecasting Model With Long Memory.

The tensor recurrent model is a family of nonlinear dynamical systems, of which the recurrence relation consists of a p -fold (called degree- p ) tensor product. Despite such models frequently appearing in advanced recurrent neural networks (RNNs), to this date, there are limited studies on their long memory properties and stability in sequence tasks. In this article, we propose a fractional tensor recurrent model, where the tensor degree p is extended from the discrete domain to the continuous domain, so it is effectively learnable from various datasets. Theoretically, we prove that a large degree p is essential to achieve the long memory effect in a tensor recurrent model, yet it could lead to unstable dynamical behaviors. Hence, our new model, named fractional tensor recurrent unit (fTRU), is expected to seek the saddle point between long memory property and model stability during the training. We experimentally show that the proposed model achieves competitive performance with a long memory and stable manners in several forecasting tasks compared to various advanced RNNs.

4-Octyl itaconate inhibits poly(I:C)-induced interferon-ß secretion in mouse bone marrow-derived macrophages partially by activating Nrf2.

Viruses have become a major threat to human health. Interferon-ß (IFN-ß) has a key role in the antivirus process, as it can increase the expression of antivirus-associated genes. Itaconate and its derivatives can regulate the immune response, secretion of inflammatory factors, and pyroptosis of macrophages. The effect of itaconate on IFN-ß secretion of double-stranded RNA-induced macrophages are not well known. A derivative of itaconate, 4-octoyl itaconate (4-OI), was used to treat mouse bone marrow-derived macrophages (BMDM) induced with 100 µg/mL poly(I:C). The IFN-ß concentration was detected through ELISA, and IFN-ß mRNA expression was detected through quantitative PCR. High-throughput transcriptome sequencing was used to analyze changes in the BMDM transcriptome after 4-OI treatment. The Nrf2 expression was knocked down with siRNA.4-OI inhibited poly(I:C)-induced IFN-ß secretion and mRNA expression in BMDM. Results of transcriptome sequencing revealed that 4-OI downregulated 1047 genes and upregulated 822 genes. GO and KEGG enrichment of differently expressed genes revealed that many downregulated genes were related to the anti-virus process, whereas many upregulated genes were related to metabolism. The Nrf2 inhibitor ML385 and Nrf2 siRNA could partially reverse the inhibitory effect of 4-OI. In conclusion, 4-octyl itaconate could inhibit the poly(I:C)-induced interferon-ß secretion in BMDM partially by regulating Nrf2.

Accurate and highly reproducible picoliter injection system for capillary electrophoresis.

A novel, highly accurate sample injection system for capillary electrophoresis (CE) was developed based on an inkjet microchip capable of reproducing exact introduction volumes at the picoliter level. The difficulty in analyte discrimination using electrokinetic injection was also overcome using this injection method. The injection system consisted of an XY stage, an inkjet droplet ejection microchip, and a reservoir with a plug-in septum. To evaluate the precision of the system, a mixture of NBD-labeled amino acids consisting of Gly, L-Phe, L-Asp, and L-Ser was separated, and the performance was compared with that of traditional hydrodynamic and electrokinetic injection methods. The results demonstrated that the introduced volume highly relied on the number of droplets with low relative standard derivation (RSD) and good linear correction coefficient in the proposed injection method. In addition, a urine sample was analyzed via CE coupled with the inkjet injection system for the detection of the amino acid taurine. The concentration of urinary taurine was determined to be 2.42 ± 0.08 µM (confidence level, 95%; RSD, 1.05%; n = 4) with a recovery of 98.92-109.54% (n = 3). These results demonstrate the inkjet injection system we developed has the potential to revolutionize capillary electrophoretic separation in practical and commercial applications that require an automated accurate injection system.

Efficiency of Multifunctional Antibacterial Hydrogels for Chronic Wound Healing in Diabetes: A Comprehensive Review.

Diabetic chronic wounds or amputation, which are complications of diabetes mellitus (DM), are a cause of great suffering for diabetics. In addition to the lack of oxygen, elevated reactive oxygen species (ROS) and reduced vascularization, microbial invasion is also a critical factor that induces non-healing chronic diabetic wounds, ie, wounds still remaining in the stage of inflammation, after which the wound tissue begins to age and becomes necrotic. To clear up the infection, alleviate the inflammation in the wound and prevent necrosis, many kinds of hydrogel have been fabricated to eliminate infections with pathogens. The unique properties of hydrogels make them ideally suited to wound dressings because they provide a moist environment for wound healing and act as a barrier against bacteria. This review article will mainly cover the recent developments and innovations of antibacterial hydrogels for diabetic chronic wound healing.

Applications of Explainable Artificial Intelligence in Diagnosis and Surgery.

In recent years, artificial intelligence (AI) has shown great promise in medicine. However, explainability issues make AI applications in clinical usages difficult. Some research has been conducted into explainable artificial intelligence (XAI) to overcome the limitation of the black-box nature of AI methods. Compared with AI techniques such as deep learning, XAI can provide both decision-making and explanations of the model. In this review, we conducted a survey of the recent trends in medical diagnosis and surgical applications using XAI. We have searched articles published between 2019 and 2021 from PubMed, IEEE Xplore, Association for Computing Machinery, and Google Scholar. We included articles which met the selection criteria in the review and then extracted and analyzed relevant information from the studies. Additionally, we provide an experimental showcase on breast cancer diagnosis, and illustrate how XAI can be applied in medical XAI applications. Finally, we summarize the XAI methods utilized in the medical XAI applications, the challenges that the researchers have met, and discuss the future research directions. The survey result indicates that medical XAI is a promising research direction, and this study aims to serve as a reference to medical experts and AI scientists when designing medical XAI applications.