Artificial Intelligence Classification Research Articles

Artificial intelligence-based triage of large bowel biopsies can improve workflow

BackgroundLarge bowel biopsies are one of the commonest types of biopsy specimen. We describe a service evaluation study to test the feasibility of using artificial intelligence (AI) to triage large bowel biopsies from a reporting backlog and prioritize those that require more urgent reporting. MethodsThe pathway was developed in the UK by National Health Service (NHS) laboratory staff working in a medium-sized general hospital. The AI platform was interfaced with the slide scanner software and the reporting platform’s software, so that pathologists could correct the AI label and reinforce the training set as they reported the cases. ResultsThe AI classifier achieved a sensitivity of 97.56% and specificity of 93.02% for the case-level-diagnosis of neoplasia (adenoma and adenocarcinoma) and for an AI diagnosis of any significant pathology (i.e., adenomas, adenocarcinomas, inflammation, hyperplastic polyps, and sessile serrated lesions) sensitivity was 95.65% and specificity 92.96%. The automated AI diagnostic classification pathway took approximately 175 s per slide to download and process the scanned whole slide image (WSI) and return an AI diagnostic classification. Biopsies with an AI diagnosis of neoplasia or inflammation were prioritized for reporting while the remainder followed the routine reporting pathway. The AI triaged pathway resulted in a significantly shorter reporting turnaround time for pathologist verified neoplastic cases (P < 0.001) and inflammation (P < 0.05). The project’s costs amounted to £14800, excluding laboratory staff salaries. More time and resources were spent on developing the interface between the AI platform and laboratory IT systems than on the development of the AI platform itself. ConclusionsNHS laboratory staff were able to implement an AI solution to accurately triage large bowel biopsies into several diagnostic classes and this improved reporting turnaround times for cases with neoplasia or with inflammation.

School Mode Choice Classification Model Exploitation Though Artificial Intelligence Classification Application

Human behavior pattern recognition exploited using Artificial Neural Networks (ANN) is at the core of Artificial Intelligence (AI) classification applications. School trip transport forecasting mode selection based upon ANN forecasting ability provided an easy-to-use scientific toolkit for sustainable urban policies designing and implementation. The paper capitalized acknowledged forecasting ability of ANN to recognize and classify behavior patterns leading to parental decisions of school mode choice. Main stages of this research included the conduction of an extended questionnaire survey in 512 parents of school students in Thessaloniki (northern Greece) and the investigation whether ANN forecasting model could classify school mode choice. Research provided promising results (forecasting ability ranging from 76% to 93%) on school mode parental selection forecasting models based on ANN classifier, providing a solid proof of concept for further investigation.

Using attention methods to predict judicial outcomes

The prediction of legal judgments is one of the most recognized fields in Natural Language Processing, Artificial Intelligence, and Law combined. By legal prediction, we mean intelligent systems capable of predicting specific judicial characteristics such as the judicial outcome, the judicial class, and the prediction of a particular case. In this study, we used an artificial intelligence classifier to predict the decisions of Brazilian courts. To this end, we developed a text crawler to extract data from official Brazilian electronic legal systems, consisting of two datasets of cases of second-degree murder and active corruption. We applied various classifiers, such as Support Vector Machines, Neural Networks, and others, to predict judicial outcomes by analyzing text features from the dataset. Our research demonstrated that Regression Trees, Gated Recurring Units, and Hierarchical Attention Networks tended to have higher metrics across our datasets. As the final goal, we searched the weights of one of the algorithms, Hierarchical Attention Networks, to find samples of the words that might be used to acquit or convict defendants based on their relevance to the algorithm.

Improving quality control in the routine practice for histopathological interpretation of gastrointestinal endoscopic biopsies using artificial intelligence.

Colorectal and gastric cancer are major causes of cancer-related deaths. In Korea, gastrointestinal (GI) endoscopic biopsy specimens account for a high percentage of histopathologic examinations. Lack of a sufficient pathologist workforce can cause an increase in human errors, threatening patient safety. Therefore, we developed a digital pathology total solution combining artificial intelligence (AI) classifier models and pathology laboratory information system for GI endoscopic biopsy specimens to establish a post-analytic daily fast quality control (QC) system, which was applied in clinical practice for a 3-month trial run by four pathologists. Our whole slide image (WSI) classification framework comprised patch-generator, patch-level classifier, and WSI-level classifier. The classifiers were both based on DenseNet (Dense Convolutional Network). In laboratory tests, the WSI classifier achieved accuracy rates of 95.8% and 96.0% in classifying histopathological WSIs of colorectal and gastric endoscopic biopsy specimens, respectively, into three classes (Negative for dysplasia, Dysplasia, and Malignant). Classification by pathologic diagnosis and AI prediction were compared and daily reviews were conducted, focusing on discordant cases for early detection of potential human errors by the pathologists, allowing immediate correction, before the pathology report error is conveyed to the patients. During the 3-month AI-assisted daily QC trial run period, approximately 7-10 times the number of slides compared to that in the conventional monthly QC (33 months) were reviewed by pathologists; nearly 100% of GI endoscopy biopsy slides were double-checked by the AI models. Further, approximately 17-30 times the number of potential human errors were detected within an average of 1.2 days. The AI-assisted daily QC system that we developed and established demonstrated notable improvements in QC, in quantitative, qualitative, and time utility aspects. Ultimately, we developed an independent AI-assisted post-analytic daily fast QC system that was clinically applicable and influential, which could enhance patient safety.

Application of Artificial Neural Network to Nucleic Acid Analysis: Accurate Discrimination for Untypical Real-Time Fluorescence Curves With High Specificity and Sensitivity

Abstract As a division of polymerase chain reaction (PCR), convective PCR (CPCR) is able to achieve highly efficient thermal cycling based on free thermal convection with pseudo-isothermal heating, which could be beneficial to point-of-care (POC) nucleic acid analysis. Similar to traditional PCR or isothermal amplification, due to a couple of issues, e.g., reagent, primer design, reactor, reaction dynamics, amplification status, temperature and heating condition, and other reasons, in some cases of CPCR tests, untypical real-time fluorescence curves with positive or negative tests will show up. Especially, when parts of the characteristics between untypical low-positive and negative tests are mixed together, it is difficult to discriminate between them using traditional cycle threshold (Ct) value method. To handle this issue which may occur in CPCR, traditional PCR or isothermal amplification, as an example, instead of using complicated mathematical modeling and signal processing strategy, an artificial intelligence (AI) classification method with artificial neural network (ANN) modeling is developed to improve the accuracy of nucleic acid detection. It has been proven that both the detection specificity and sensitivity can be significantly improved even with a simple ANN model. It can be estimated that the developed method based on AI modeling can be adopted to solve similar problem with PCR or isothermal amplification methods.

Artificial-Intelligence-Based Open-Circuit Fault Diagnosis in VSI-Fed PMSMs and a Novel Fault Recovery Method

Artificial intelligence (AI) techniques are widely used in fault diagnosis because they are superior in detection and prediction. The detection of faults in power systems containing electronic components is critical. The switch faults of the voltage source inverter (VSI) have a severe impact on the driving system. Short-circuit switches increase the thermal stress due to their fast and high stator currents. Additionally, open-circuit switches cause unstable motor operation. However, these issues are not sufficiently addressed or accurately predicted for VSI switch faults in the literature. Thus, this paper investigates the use of different AI classifiers for three-phase VSI fault diagnosis. Various AI methods are used, such as naïve Bayes, support vector machine (SVM), artificial neural network (ANN), and decision tree (DT) techniques. These methods are applied to a VSI-fed permanent magnet synchronous motor (PMSM) to detect the faults in the inverter switches. These methods use the drain–source voltage and PWM signals to decide whether the switch is healthy or unhealthy. In addition, they are compared in terms of their detection accuracy. In this regard, the comparative results show that the DT method has the highest accuracy as compared to other methods in the fault diagnosis process. Moreover, this paper proposes a novel and universal voltage compensation loop to compensate for the absence of the voltage portion due to the open switch fault. Thus, the driving system is assisted in operating under its normal operating conditions. The universal term is used because the proposed voltage compensation loop can be implemented in any type of inverter. To validate the results, the proposed system is implemented using two software programs, LTSPICE XVII-USA, WEKA 3.9-New Zealand.

Виявлення бонет-трафіку на основі потоків, використовуючи ШІ

This paper outlines the generalized framework for building end-to-end botnet network activity detection systems using artificial intelligence (AI) techniques. The paper describes network flows reconstruction as a primary feature-extraction method and considers different AI classifiers for achieving a better detection rate. The results of the latest research by other authors in the field are incorporated to implement a more efficient approach for botnet discovery. The described intrusion detection pipeline was tested on a dataset with real botnet activity traces. The performance metrics for different AI classification models were obtained and analyzed in detail. Different data preprocessing techniques were tried and described which helped improve the results even further. Some options for future enhancement of network feature selection were proposed as well. The comparison of the obtained performance metrics was drawn against the results provided by other researchers in this field.

Machine learning-supported interpretation of kidney graft elementary lesions in combination with clinical data

Interpretation of kidney graft biopsies using the Banff classification is still heterogeneous. In this study, extreme gradient boosting classifiers learned from two large training datasets (n=631 and 304 cases) where the "reference diagnoses" were not strictly defined following the Banff rules but from central reading by expert pathologists and further interpreted consensually by experienced transplant nephrologists, in light of the clinical context. In three external validation datasets (n=3744, 589, and 360), the classifiers yielded a mean ROC curve AUC (95%CI) of: 0.97 (0.92-1.00), 0.97 (0.96-0.97), and 0.95 (0.93-0.97) for antibody-mediated rejection (ABMR); 0.94 (0.91-0.96), 0.94 (0.92-0.95), and 0.91 (0.88-0.95) for T cell-mediated rejection; >0.96 (0.90-1.00) with all three for interstitial fibrosis-tubular atrophy. We also developed a classifier to discriminate active and chronic active ABMR with 95% accuracy. In conclusion, we built highly sensitive and specific artificial intelligence classifiers able to interpret kidney graft scoring together with a few clinical data and automatically diagnose rejection, with excellent concordance with the Banff rules and reference diagnoses made by a group of experts. Some discrepancies may point toward possible improvements that could be made to the Banff classification.

Artificial intelligence to understand fluctuation of fetal brain activity by recognizing facial expressions.

To examine whether artificial intelligence can achieve discoveries regarding fetal brain activity. In this observational study, the authors collected images of fetal faces using a four-dimensional ultrasound technique obtained from singleton pregnancies of outpatients in routine practice at 27 to 37 weeks of gestation between February 1 and December 31, 2021. The authors developed an artificial intelligence classifier to recognize seven facial expressions of fetuses, then applied it to video files of fetal facial images to generate the probabilities, as confidence scores, of each expression category. Discrete Fourier transform and chaotic analysis were used to investigate the scores. Mann-Whitney test, t test, variance test, and one-way analysis of variance were used for statistical analysis. Facial expression changes were observed in cycles averaging 66 to 73 s. The power spectrum showed that mouthing and neutral expressions were the most prevalent. There was a difference between categories for the spectrum (p=0.004). Two different states--dense and sparse--of confidence scores were discovered. The correlation dimension was 1.19 ± 0.22 and 1.33 ± 0.27 for dense and sparse, respectively (p=0.047). This method objectively and quantitatively demonstrated fetal brain activity and may provide insight into how the fetus spends its time in utero.

Abstract 13550: Electrocardiographic-Based Artificial Intelligence Model in Prediction of 1-year Mortality in Heart Failure With Reduced Ejection Fraction

Objectives: To assess whether electrocardiography (ECG)-based artificial intelligence (AI) algorithm developed to detect 1-year mortality in patients with HF with reduced ejection fraction (HFrEF). Methods and results: Data from two hospital visits (2016 Oct to 2021 Apr) was used. Participants with good quality baseline ECGs and HFrEF were included. We tested the hypothesis that the application of AI to the ECG could identify 1-yr mortality in HFrEF patients. A total of 16,228 (64% male, mean age of 58.5) were eligible. When tested on an independent set of 16,228 patients, the network model yielded values for the area under the curve, sensitivity, specificity, and accuracy of 0.83, 83.3%, and 64.2%, respectively ( Figure a ). Those with a positive AI screen were at 2.6 times the risk (p&lt;0.0001) of developing future mortality compared with those with a negative screen ( Figure b ). We used a sensitivity map to visualize the ECG region used in the AI model to detect high 1-yr mortality in HFrEF ( Figure c ). The map shows that the AI model focused on the QRS complex, particularly the R-wave, in most patients. Conclusions: The application of AI to the ECG permits the ECG to serve as a powerful screening tool for detecting 1-yr mortality in HFrEF. Figure a. The area under the curve of the convolutional neural network used to identify 1-year mortality of heart failure with reduced ejection fraction. Figure b. Long-term incidence of mortality in patients with initially 1-year mortality stratified by artificial intelligence classification. Figure c. Sensitivity map of a deep learning model for detecting 1-year mortality in heart failure with reduced ejection fraction.

Assessment of Radiology Artificial Intelligence Software: A Validation and Evaluation Framework.

Artificial intelligence (AI) software in radiology is becoming increasingly prevalent and performance is improving rapidly with new applications for given use cases being developed continuously, oftentimes with development and validation occurring in parallel. Several guidelines have provided reporting standards for publications of AI-based research in medicine and radiology. Yet, there is an unmet need for recommendations on the assessment of AI software before adoption and after commercialization. As the radiology AI ecosystem continues to grow and mature, a formalization of system assessment and evaluation is paramount to ensure patient safety, relevance and support to clinical workflows, and optimal allocation of limited AI development and validation resources before broader implementation into clinical practice. To fulfil these needs, we provide a glossary for AI software types, use cases and roles within the clinical workflow; list healthcare needs, key performance indicators and required information about software prior to assessment; and lay out examples of software performance metrics per software category. This conceptual framework is intended to streamline communication with the AI software industry and provide healthcare decision makers and radiologists with tools to assess the potential use of these software. The proposed software evaluation framework lays the foundation for a radiologist-led prospective validation network of radiology AI software. Learning Points: The rapid expansion of AI applications in radiology requires standardization of AI software specification, classification, and evaluation. The Canadian Association of Radiologists' AI Tech & Apps Working Group Proposes an AI Specification document format and supports the implementation of a clinical expert evaluation process for Radiology AI software.

Structural Model Based on Genetic Algorithm for Inhibiting Fatty Acid Amide Hydrolase

The fatty acid amide hydrolase (FAAH) is an enzyme responsible for the degradation of anandamide, an endocannabinoid. Pharmacologically blocking this target can lead to anxiolytic effects; therefore, new inhibitors can improve therapy in this field. In order to speed up the process of drug discovery, various in silico methods can be used, such as molecular docking, quantitative structure–activity relationship models (QSAR), and artificial intelligence (AI) classification algorithms. Besides architecture, one important factor for an AI model with high accuracy is the dataset quality. This issue can be solved by a genetic algorithm that can select optimal features for the prediction. The objective of the current study is to use this feature selection method in order to identify the most relevant molecular descriptors that can be used as independent variables, thus improving the efficacy of AI algorithms that can predict FAAH inhibitors. The model that used features chosen by the genetic algorithm had better accuracy than the model that used all molecular descriptors generated by the CDK descriptor calculator 1.4.6 software. Hence, carefully selecting the input data used by AI classification algorithms by using a GA is a promising strategy in drug development.

Numerical simulation of gears for fault detection using artificial intelligence models

Artificial intelligence (AI) models are widely used in every field of life to classify and predict the abnormal conditions. The key to activating AI models is to obtain enough training samples. The lumped parameter model is the one of the best choices to provide sufficient samples. In this study, the lumped parameter model is used in the numerical simulation of gears to obtain enough training samples to detect their running state. The aim is to replace the traditional sample generation method with the modified lumped parameter numerical method to realize the fault diagnosis of gears. Firstly, the lumped parameter model of gears is constructed and further updated. Secondly, failure types are inserted into the updated lumped parameter model and generate samples. The set composed of measured signals and simulated signals is equally divided, and the corresponding time-domain signal indexes are calculated to form a sample matrix for AI training and classification. Finally, CNN, RNN and LSTM are selected as the representatives of AI model, and then the trained models are used to classify the unknown measurement fault samples. The classification model trained by the simulation fault samples has almost 100% accuracy for the current gear fault classification, which shows that it is a feasible way to generate sample data for artificial intelligence fault classification by using the lumped parameter models.

AI‐driven network softwarization scheme for efficient message exchange in IoT environment beyond 5G

SummaryInternet of things (IoT) has massively adopted the market due to the current era demanding fully intelligent and autonomous services. However, efficient network management and message exchange are challenging in a dynamic and energy‐constrained IoT environment. Hence, for efficient message passing in IoT applications, softwarization of the IoT network is essential, wherein the logical control plane is decoupled from the data plane consisting of hardware devices such as routers and switches. Using softwarization, a centralized software‐defined networking (SDN) controller is responsible for routing data packets from source to destination in a dynamic environment. However, to reduce the computational overhead of filtering malicious and nonmalicious packets, artificial intelligence (AI) classifiers prove beneficial. Moreover, such challenges have trade‐offs with network requirements such as ultralow latency, high reliability, and higher data rates. Motivated by this, we propose an AI‐enabled network softwarization scheme for efficient message exchange under the 6G network. Lastly, the performance of the proposed scheme is evaluated with different performance metrics, such as accuracy, precision, f1‐score, packet drop ratio, and latency. The empirical result revealed that the proposed scheme outperforms in terms of accuracy and controller efficiency, that is, 81.64% and 82.2%, respectively.

Use of digital pathology and artificial intelligence for the diagnosis of Helicobacter pylori in gastric biopsies.

A common source of concern about digital pathology (DP) is that limited resolution could be a reason for an increased risk of malpractice. A frequent question being raised about this technology is whether it can be used to reliably detect Helicobacter pylori (HP) in gastric biopsies, which can be a significant burden in routine work. The main goal of this work is to show that a reliable diagnosis of HP infection can be made by DP even at low magnification. The secondary goal is to demonstrate that artificial intelligence (AI) algorithms can diagnose HP infections on virtual slides with sufficient accuracy. The method we propose is based on the Warthin-Starry (W-S) silver stain which allows faster detection of HP in virtual slides. A software tool, based on regular expressions, performed a specific search to select 679 biopsies on which a W-S stain was done. From this dataset 185 virtual slides were selected to be assessed by WSI and compared with microscopy slide readings. To determine whether HP infections could be accurately diagnosed with machine learning. AI was used as a service (AIaaS) on a neural network-based web platform trained with 468 images. A test dataset of 210 images was used to assess the classifier performance. In 185 gastric biopsies read with DP we recorded only 4 false positives and 4 false negatives with an overall agreement of 95.6%. Compared with microscopy, defined as the "gold standard" for the diagnosis of HP infections, WSI had a sensitivity and specificity of 0.95 and 0.96, respectively. The ROC curve of our AI classifier generated on a testing dataset of 210 images had an AUC of 0.938. This study demonstrates that DP and AI can be used to reliably identify HP at 20X resolution.

Mathematical Ranking of AI Classifiers Using Confusion Matrix and Matthews Correlation Coefficient

This research primarily deals with mathematically ranking the level of accuracy of various Artificial Intelligence (AI) based machine learning classifiers, using Mathews Correlation Coefficient (MCC), leveraging Confusion Matrices. A detailed Literature survey was done to gather the existing knowledge. This knowledge was used as foundational basis to further build the scope of this research project. The classifiers used were Support Vector Machine (SVM), K-Nearest Neighbors (KNN), and Convolutional Neural Network (CNN). These classifiers were used for Face recognition of individuals. A total of 33 adult test subjects with 10 images per subject resulting in a total of 330 distinct images were part of this research project. These test subjects were from the Labeled Faces in The Wild publicly available database along with teachers from James B. Conant High School who voluntarily participated in this research. Python based face recognition program wrappers and associated environment was built around pre-existing classifiers and the image data was passed to these wrappers. These wrappers recognized the faces of individuals in the images over 10 trials wherein each trial consisted of 33 distinct images, with varying degree of accuracy and presented that as outputs. These outputs were used to determine the Confusion Matrices which in turn were used to calculate the MCC scores. The MCC scores were plotted, and these results showed that the SVM AI classifier had the highest level of relative accuracy for face recognition, followed by KNN and CNN classifiers.

Cooperation between artificial intelligence and endoscopists for diagnosing invasion depth of early gastric cancer

Background and study aimsThe diagnostic ability of endoscopists to determine invasion depth of early gastric cancer is not favorable. We designed an artificial intelligence (AI) classifier for differentiating intramucosal and submucosal gastric cancers and examined it to establish a diagnostic method based on cooperation between AI and endoscopists.Patients and methodsWe prepared 500 training images using cases of mainly depressed-type early gastric cancer from 250 intramucosal cancers and 250 submucosal cancers. We also prepared 200 test images each of 100 cancers from another institution. We designed an AI classifier to differentiate between intramucosal and submucosal cancers by deep learning. We examined the performance of the AI classifier and the majority vote of the endoscopists as high confidence and low confidence diagnostic probability, respectively, and cooperatively combined them to establish a diagnostic method providing high accuracy.ResultsInternal evaluation of the training images showed that accuracy, sensitivity, specificity, and F1 measure by the AI classifier were 77%, 76%, 78%, and 0.768, and those of the majority vote of the endoscopists were 72.6%, 53.6%, 91.6%, and 0.662, respectively. A diagnostic method based on cooperation between AI and the endoscopists showed that the respective values were 78.0%, 76.0%, 80.0%, and 0.776 for the test images. The value of F1 measure was especially higher than those by AI or the endoscopists alone.ConclusionsCooperation between AI and endoscopists improved the diagnostic ability to determine invasion depth of early gastric cancer.

Effect of Story Structure Instruction Based on Visual Analysis on Reading Comprehension Intervention for Dyslexic Students.

The application of artificial intelligence (AI) technology in the field of dyslexia is becoming increasingly abundant. However, the content of related literature shows that there is still a lack of systematic and comprehensive research in this field at home and abroad. By outlining the development of AI technology, the meaning, causes, and classification of learning disabilities, the most representative studies on the application of AI technology in dyslexia education, including four aspects of diagnosis, intervention, assessment, and services are analyzed. The study finds that AI technology can improve the conditions suffering from dyslexia and dysgraphia, and can serve the education of dyslexic children as a technical tool to overcome dyslexia. By summarizing the effect of story structure teaching based on visual analysis on reading comprehension intervention for students with dyslexia, it can provide useful references and references for related research.

Machine learning applied to big data from marine cabled observatories: A case study of sablefish monitoring in the NE Pacific

Ocean observatories collect large volumes of video data, with some data archives now spanning well over a few decades, and bringing the challenges of analytical capacity beyond conventional processing tools. The analysis of such vast and complex datasets can only be achieved with appropriate machine learning and Artificial Intelligence (AI) tools. The implementation of AI monitoring programs for animal tracking and classification becomes necessary in the particular case of deep-sea cabled observatories, as those operated by Ocean Networks Canada (ONC), where Petabytes of data are now collected each and every year since their installation. Here, we present a machine-learning and computer vision automated pipeline to detect and count sablefish (Anoplopoma fimbria), a key commercially exploited species in the N-NE Pacific. We used 651 hours of video footage obtained from three long-term monitoring sites in the NEPTUNE cabled observatory, in Barkley Canyon, on the nearby slope, and at depths ranging from 420 to 985 m. Our proposed AI sablefish detection and classification pipeline was tested and validated for an initial 4.5 month period (Sep 18 2019-Jan 2 2020), and was a first step towards validation for future processing of the now decade-long video archives from Barkley Canyon. For the validation period, we trained a YOLO neural network on 2917 manually annotated frames containing sablefish images to obtain an automatic detector with a 92% Average Precision (AP) on 730 test images, and a 5-fold cross-validation AP of 93% (± 3.7%). We then ran the detector on all video material (i.e., 651 hours from a 4.5 month period), to automatically detect and annotate sablefish. We finally applied a tracking algorithm on detection results, to approximate counts of individual fishes moving on scene and obtain a time series of proxy sablefish abundance. Those proxy abundance estimates are among the first to be made using such a large volume of video data from deep-sea settings. We discuss our AI results for application on a decade-long video monitoring program, and particularly with potential for complementing fisheries management practices of a commercially important species.

Possible Bias in Supervised Deep Learning Algorithms for CT Lung Nodule Detection and Classification.

Simple SummaryArtificial Intelligence (AI) algorithms can assist clinicians in their daily tasks by automatically detecting and/or classifying nodules in chest CT scans. Bias of such algorithms is one of the reasons why implementation of them in clinical practice is still not widely adopted. There is no published review on the bias that these algorithms may contain. This review aims to present different types of bias in such algorithms and present possible ways to mitigate them. Only then it would be possible to ensure that these algorithms work as intended under many different clinical settings.Artificial Intelligence (AI) algorithms for automatic lung nodule detection and classification can assist radiologists in their daily routine of chest CT evaluation. Even though many AI algorithms for these tasks have already been developed, their implementation in the clinical workflow is still largely lacking. Apart from the significant number of false-positive findings, one of the reasons for that is the bias that these algorithms may contain. In this review, different types of biases that may exist in chest CT AI nodule detection and classification algorithms are listed and discussed. Examples from the literature in which each type of bias occurs are presented, along with ways to mitigate these biases. Different types of biases can occur in chest CT AI algorithms for lung nodule detection and classification. Mitigation of them can be very difficult, if not impossible to achieve completely.