Recent Application of Artificial Intelligence in Non-Gynecological Cancer Cytopathology: A Systematic Review

Improved diagnostic accuracy of thyroid fine-needle aspiration cytology with artificial intelligence technology

BACKGROUND

Artificial intelligence (AI) is increasingly being applied in pathology and cytology, showing promising results. We collected a large dataset of whole slide image of thyroid fine needle aspiration cytology (FNA), incorporating z-stacking, from institutions across the nation to develop an AI model.

METHODS

We conducted a multi-center retrospective diagnostic accuracy study using thyroid FNA dataset from the Open AI Dataset Project that consists of digitalized images samples collected from three university hospitals and 215 Korean institutions through extensive quality check during the case selection, scanning, labeling, and reviewing process. Multiple z-layer images were captured using three different scanners and image patches were extracted from whole slide images and resized after focus-fusion and color normalization. We pre-tested six AI models, determining Inception ResNet v2 as the best model using a subset of dataset, and subsequently tested the final model with total datasets. Additionally, we compared the performance of AI and cytopathologists using randomly selected 1,031 image patches and reevaluated the cytopathologists' performance after reference to AI results.

RESULTS

A total of 10,332 image patches from 306 thyroid FNAs, comprising 78 malignant (Papillary thyroid carcinoma) and 228 benign from 86 institutions were used for the AI training. Inception ResNet v2 achieved highest accuracy of 99.7%, 97.7%, and 94.9% for training, validation, and test dataset, respectively (Sensitivity 99.9%, 99.6%, and 100% and specificity 99.6%, 96.4%, and 90.4% for training, validation, and test dataset, respectively). In the comparison between AI and human, AI model showed higher accuracy and specificity than the average expert cytopathologists beyond the two-standard deviation (Accuracy 99.71% (95% CI, 99.38-100.00%) vs. 88.91% (95% CI, 86.99-90.83%), sensitivity 99.81% (95% CI, 99.54-100.00% ) vs. 87.26% (95% CI, 85.22-89.30%), and specificity 99.61% (95% CI, 99.23-99.99%) vs. 90.58% (95% CI, 88.80-92.36%). Moreover, after referring to the AI results, all the performance of the experts increased (Accuracy 96%, 95%, and 96%, respectively) as well as diagnostic agreement (from 0.64 to 0.84).

CONCLUSIONS

These results suggest that the application of AI technology to thyroid FNA cytology may improve the diagnostic accuracy as well as intra- and inter-observer variability among pathologists. Further confirmatory research is needed.

Prediction of tumor origin in cancers of unknown primary origin with cytology-based deep learning

Cancer of unknown primary (CUP) site poses diagnostic challenges due to its elusive nature. Many cases of CUP manifest as pleural and peritoneal serous effusions. Leveraging cytological images from 57,220 cases at four tertiary hospitals, we developed a deep-learning method for tumor origin differentiation using cytological histology (TORCH) that can identify malignancy and predict tumor origin in both hydrothorax and ascites. We examined its performance on three internal (n = 12,799) and two external (n = 14,538) testing sets. In both internal and external testing sets, TORCH achieved area under the receiver operating curve values ranging from 0.953 to 0.991 for cancer diagnosis and 0.953 to 0.979 for tumor origin localization. TORCH accurately predicted primary tumor origins, with a top-1 accuracy of 82.6% and top-3 accuracy of 98.9%. Compared with results derived from pathologists, TORCH showed better prediction efficacy (1.677 versus 1.265, P < 0.001), enhancing junior pathologists' diagnostic scores significantly (1.326 versus 1.101, P < 0.001). Patients with CUP whose initial treatment protocol was concordant with TORCH-predicted origins had better overall survival than those who were administrated discordant treatment (27 versus 17 months, P = 0.006). Our study underscores the potential of TORCH as a valuable ancillary tool in clinical practice, although further validation in randomized trials is warranted.

Evaluating Urine Cytology Slide Digitization Efficiency: A Comparative Study Using an Artificial Intelligence-Based Heuristic Scanning Simulation and Multiple Z-Plane Scanning

INTRODUCTION

Digitizing cytology slides presents challenges because of their three-dimensional features and uneven cell distribution. While multi-Z-plane scan is a prevalent solution, its adoption in clinical digital cytopathology is hindered by prolonged scanning times, increased image file sizes, and the requirement for cytopathologists to review multiple Z-plane images.

METHODS

This study presents heuristic scan as a novel solution, using an artificial intelligence (AI)-based approach specifically designed for cytology slide scanning as an alternative to the multi-Z-plane scan. Both the 21 Z-plane scan and the heuristic scan simulation methods were used on 52 urine cytology slides from three distinct cytopreparations (Cytospin, ThinPrep, and BD CytoRich™ [SurePath]), generating whole-slide images (WSIs) via the Leica Aperio AT2 digital scanner. The AI algorithm inferred the WSI from 21 Z-planes to quantitate the total number of suspicious for high-grade urothelial carcinoma or more severe cells (SHGUC+) cells. The heuristic scan simulation calculated the total number of SHGUC+ cells from the 21 Z-plane scan data. Performance metrics including SHGUC+ cell coverage rates (calculated by dividing the number of SHGUC+ cells identified in multiple Z-planes or heuristic scan simulation by the total SHGUC+ cells in the 21 Z-planes for each WSI), scanning time, and file size were analyzed to compare the performance of each scanning method. The heuristic scan's metrics were linearly estimated from the 21 Z-plane scan data. Additionally, AI-aided interpretations of WSIs with scant SHGUC+ cells followed The Paris System guidelines and were compared with original diagnoses.

RESULTS

The heuristic scan achieved median SHGUC+ cell coverage rates similar to 5 Z-plane scans across three cytopreparations (0.78-0.91 vs. 0.75-0.88, p = 0.451-0.578). Notably, it substantially reduced both scanning time (137.2-635.0 s vs. 332.6-1,278.8 s, p < 0.05) and image file size (0.51-2.10 GB vs. 1.16-3.10 GB, p < 0.05). Importantly, the heuristic scan yielded higher rates of accurate AI-aided interpretations compared to the single Z-plane scan (62.5% vs. 37.5%).

CONCLUSION

We demonstrated that the heuristic scan offers a cost-effective alternative to the conventional multi-Z-plane scan in digital cytopathology. It achieves comparable SHGUC+ cell capture rates while reducing both scanning time and image file size, promising to aid digital urine cytology interpretations with a higher accuracy rate compared to the conventional single (optimal) plane scan. Further studies are needed to assess the integration of this new technology into compatible digital scanners for practical cytology slide scanning.

Automated Analysis of Nuclear Parameters in Oral Exfoliative Cytology Using Machine Learning

BACKGROUND

As oral cancer remains a major worldwide health concern, sophisticated diagnostic tools are needed to aid in early diagnosis. Non-invasive methods like exfoliative cytology, albeit with the help of artificial intelligence (AI), have drawn additional interest.

AIM

The study aimed to harness the power of machine learning algorithms for the automated analysis of nuclear parameters in oral exfoliative cytology. Further, the analysis of two different AI systems, namely convoluted neural networks (CNN) and support vector machine (SVM), were compared for accuracy.

METHODS

A comparative diagnostic study was performed in two groups of patients (n=60). The control group without evidence of lesions (n=30) and the other group with clinically suspicious oral malignancy (n=30) were evaluated. All patients underwent cytological smears using an exfoliative cytology brush, followed by routine Hematoxylin and Eosin staining. Image preprocessing, data splitting, machine learning, model development, feature extraction, and model evaluation were done. An independent t-test was run on each nuclear characteristic, and Pearson's correlation coefficient test was performed with Statistical Package for the Social Sciences (SPSS) software (IBM SPSS Statistics for Windows, Version 28.0. IBM Corp, Armonk, NY, USA).

RESULTS

 The study found substantial variations between the study and control groups in nuclear size (p<0.05), nuclear shape (p<0.01), and chromatin distribution (p<0.001). The Pearson correlation coefficient of SVM was 0.6472, and CNN was 0.7790, showing that SVM had more accuracy.

CONCLUSION

The availability of multidimensional datasets, combined with breakthroughs in high-performance computers and new deep-learning architectures, has resulted in an explosion of AI use in numerous areas of oncology research. The discerned diagnostic accuracy exhibited by the SVM and CNN models suggests prospective improvements in early detection rates, potentially improving patient outcomes and enhancing healthcare practices.

Multiple serous cavity effusion screening based on smear images using vision transformer

Serous cavity effusion is a prevalent pathological condition encountered in clinical settings. Fluid samples obtained from these effusions are vital for diagnostic and therapeutic purposes. Traditionally, cytological examination of smears is a common method for diagnosing serous cavity effusion, renowned for its convenience. However, this technique presents limitations that can compromise its efficiency and diagnostic accuracy. This study aims to overcome these challenges and introduce an improved method for the precise detection of malignant cells in serous cavity effusions. We have developed a transformer-based classification framework, specifically employing the vision transformer (ViT) model, to fulfill this objective. Our research involved collecting smear images and corresponding cytological reports from 161 patients who underwent serous cavity drainage. We meticulously annotated 4836 patches from these images, identifying regions with and without malignant cells, thus creating a unique dataset for smear image classification. The findings of our study reveal that deep learning models, particularly the ViT model, exhibit remarkable accuracy in classifying patches as malignant or non-malignant. The ViT model achieved an impressive area under the receiver operating characteristic curve (AUROC) of 0.99, surpassing the performance of the convolutional neural network (CNN) model, which recorded an AUROC of 0.86. Additionally, we validated our models using an external cohort of 127 patients. The ViT model sustained its high-level screening performance, achieving an AUROC of 0.98 at the patient level, compared to the CNN model's AUROC of 0.84. The visualization of our ViT models confirmed their capability to precisely identify regions containing malignant cells in multiple serous cavity effusion smear images. In summary, our study demonstrates the potential of deep learning models, particularly the ViT model, in automating the screening process for serous cavity effusions. These models offer significant assistance to cytologists in enhancing diagnostic accuracy and efficiency. The ViT model stands out for its advanced self-attention mechanism, making it exceptionally suitable for tasks that necessitate detailed analysis of small, sparsely distributed targets like cellular clusters in serous cavity effusions.

RARPKB: A knowledge-guide decision support platform for personalized robot-assisted surgery in prostate cancer

BACKGROUND

Robot-assisted radical prostatectomy (RARP) has emerged as a pivotal surgical intervention for the treatment of prostate cancer. However, the complexity of clinical cases, heterogeneity of prostate cancer, and limitations in physician expertise pose challenges to rational decision-making in RARP. To address these challenges, we aimed to organize the knowledge of previously complex cohorts and establish an online platform named the RARP Knowledge Base (RARPKB) to provide reference evidence for personalized treatment plans.

MATERIALS AND METHODS

PubMed searches over the past two decades were conducted to identify publications describing RARP. We collected, classified, and structured surgical details, patient information, surgical data, and various statistical results from the literature. A knowledge-guided decision-support tool was established using MySQL, DataTable, ECharts, and JavaScript. ChatGPT-4 and two assessment scales were used to validate and compare the platform.

RESULTS

The platform comprised 583 studies, 1589 cohorts, 1 911 968 patients, and 11 986 records, resulting in 54 834 data entries. The knowledge-guided decision support tool provide personalized surgical plan recommendations and potential complications on the basis of patients' baseline and surgical information. Compared with ChatGPT-4, RARPKB outperformed in authenticity (100% versus [vs.] 73%), matching (100% vs. 53%), personalized recommendations (100% vs. 20%), matching of patients (100% vs. 0%), and personalized recommendations for complications (100% vs. 20%). Post-use, the average System Usability Scale score was 88.88±15.03, and the Net Promoter Score of RARPKB was 85. The knowledge base is available at http://rarpkb.bioinf.org.cn.

CONCLUSIONS

We introduced the pioneering RARPKB, the first knowledge base for robot-assisted surgery, with an emphasis on prostate cancer. RARPKB can assist in personalized and complex surgical planning for prostate cancer to improve its efficacy. RARPKB provides a reference for the future applications of artificial intelligence in clinical practice.

Artificial-Intelligence-Assisted Detection of Metastatic Colorectal Cancer Cells in Ascitic Fluid

Ascites cytology is a cost-effective test for metastatic colorectal cancer (CRC) in the abdominal cavity. However, metastatic carcinoma of the peritoneum is difficult to diagnose based on biopsy findings, and ascitic aspiration cytology has a low sensitivity and specificity and a high inter-observer variability. The aim of the present study was to apply artificial intelligence (AI) to classify benign and malignant cells in ascites cytology patch images of metastatic CRC using a deep convolutional neural network. Datasets were collected from The OPEN AI Dataset Project, a nationwide cytology dataset for AI research. The numbers of patch images used for training, validation, and testing were 56,560, 7068, and 6534, respectively. We evaluated 1041 patch images of benign and metastatic CRC in the ascitic fluid to compare the performance of pathologists and an AI algorithm, and to examine whether the diagnostic accuracy of pathologists improved with the assistance of AI. This AI method showed an accuracy, a sensitivity, and a specificity of 93.74%, 87.76%, and 99.75%, respectively, for the differential diagnosis of malignant and benign ascites. The diagnostic accuracy and sensitivity of the pathologist with the assistance of the proposed AI method increased from 86.8% to 90.5% and from 73.3% to 79.3%, respectively. The proposed deep learning method may assist pathologists with different levels of experience in diagnosing metastatic CRC cells of ascites.

ChatGPT as an aid for pathological diagnosis of cancer

Diagnostic workup of cancer patients is highly reliant on the science of pathology using cytopathology, histopathology, and other ancillary techniques like immunohistochemistry and molecular cytogenetics. Data processing and learning by means of artificial intelligence (AI) has become a spearhead for the advancement of medicine, with pathology and laboratory medicine being no exceptions. ChatGPT, an artificial intelligence (AI)-based chatbot, that was recently launched by OpenAI, is currently a talk of the town, and its role in cancer diagnosis is also being explored meticulously. Pathology workflow by integration of digital slides, implementation of advanced algorithms, and computer-aided diagnostic techniques extend the frontiers of the pathologist's view beyond a microscopic slide and enables effective integration, assimilation, and utilization of knowledge that is beyond human limits and boundaries. Despite of it's numerous advantages in the pathological diagnosis of cancer, it comes with several challenges like integration of digital slides with input language parameters, problems of bias, and legal issues which have to be addressed and worked up soon so that we as a pathologists diagnosing malignancies are on the same band wagon and don't miss the train.

Diagnostic proficiency test using digital cytopathology and comparative assessment of whole slide images of cytologic samples for quality assurance program in Korea

BACKGROUND

The Korean Society for Cytopathology introduced a digital proficiency test (PT) in 2021. However, many doubtful opinions remain on whether digitally scanned images can satisfactorily present subtle differences in the nuclear features and chromatin patterns of cytological samples.

METHODS

We prepared 30 whole-slide images (WSIs) from the conventional PT archive by a selection process for digital PT. Digital and conventional PT were performed in parallel for volunteer institutes, and the results were compared using feedback. To assess the quality of cytological assessment WSIs, 12 slides were collected and scanned using five different scanners, with four cytopathologists evaluating image quality through a questionnaire.

RESULTS

Among the 215 institutes, 108 and 107 participated in glass and digital PT, respectively. No significant difference was noted in category C (major discordance), although the number of discordant cases was slightly higher in the digital PT group. Leica, 3DHistech Pannoramic 250 Flash, and Hamamatsu NanoZoomer 360 systems showed comparable results in terms of image quality, feature presentation, and error rates for most cytological samples. Overall satisfaction was observed with the general convenience and image quality of digital PT.

CONCLUSIONS

As three-dimensional clusters are common and nuclear/chromatin features are critical for cytological interpretation, careful selection of scanners and optimal conditions are mandatory for the successful establishment of digital quality assurance programs in cytology.

Deep Learning-Based Computational Cytopathologic Diagnosis of Metastatic Breast Carcinoma in Pleural Fluid

A Pleural effusion cytology is vital for treating metastatic breast cancer; however, concerns have arisen regarding the low accuracy and inter-observer variability in cytologic diagnosis. Although artificial intelligence-based image analysis has shown promise in cytopathology research, its application in diagnosing breast cancer in pleural fluid remains unexplored. To overcome these limitations, we evaluate the diagnostic accuracy of an artificial intelligence-based model using a large collection of cytopathological slides, to detect the malignant pleural effusion cytology associated with breast cancer. This study includes a total of 569 cytological slides of malignant pleural effusion of metastatic breast cancer from various institutions. We extracted 34,221 augmented image patches from whole-slide images and trained and validated a deep convolutional neural network model (DCNN) (Inception-ResNet-V2) with the images. Using this model, we classified 845 randomly selected patches, which were reviewed by three pathologists to compare their accuracy. The DCNN model outperforms the pathologists by demonstrating higher accuracy, sensitivity, and specificity compared to the pathologists (81.1% vs. 68.7%, 95.0% vs. 72.5%, and 98.6% vs. 88.9%, respectively). The pathologists reviewed the discordant cases of DCNN. After re-examination, the average accuracy, sensitivity, and specificity of the pathologists improved to 87.9, 80.2, and 95.7%, respectively. This study shows that DCNN can accurately diagnose malignant pleural effusion cytology in breast cancer and has the potential to support pathologists.

Artificial Intelligence in the Advanced Diagnosis of Bladder Cancer-Comprehensive Literature Review and Future Advancement

Artificial intelligence is highly regarded as the most promising future technology that will have a great impact on healthcare across all specialties. Its subsets, machine learning, deep learning, and artificial neural networks, are able to automatically learn from massive amounts of data and can improve the prediction algorithms to enhance their performance. This area is still under development, but the latest evidence shows great potential in the diagnosis, prognosis, and treatment of urological diseases, including bladder cancer, which are currently using old prediction tools and historical nomograms. This review focuses on highly significant and comprehensive literature evidence of artificial intelligence in the management of bladder cancer and investigates the near introduction in clinical practice.

Recent application of artificial intelligence on histopathologic image-based prediction of gene mutation in solid cancers

PURPOSE

Evaluation of genetic mutations in cancers is important because distinct mutational profiles help determine individualized drug therapy. However, molecular analyses are not routinely performed in all cancers because they are expensive, time-consuming and not universally available. Artificial intelligence (AI) has shown the potential to determine a wide range of genetic mutations on histologic image analysis. Here, we assessed the status of mutation prediction AI models on histologic images by a systematic review.

METHODS

A literature search using the MEDLINE, Embase and Cochrane databases was conducted in August 2021. The articles were shortlisted by titles and abstracts. After a full-text review, publication trends, study characteristic analysis and comparison of performance metrics were performed.

RESULTS

Twenty-four studies were found mostly from developed countries, and their number is increasing. The major targets were gastrointestinal, genitourinary, gynecological, lung and head and neck cancers. Most studies used the Cancer Genome Atlas, with a few using an in-house dataset. The area under the curve of some of the cancer driver gene mutations in particular organs was satisfactory, such as 0.92 of BRAF in thyroid cancers and 0.79 of EGFR in lung cancers, whereas the average of all gene mutations was 0.64, which is still suboptimal.

CONCLUSION

AI has the potential to predict gene mutations on histologic images with appropriate caution. Further validation with larger datasets is still required before AI models can be used in clinical practice to predict gene mutations.

Deep Learning-Based Classification and Targeted Gene Alteration Prediction from Pleural Effusion Cell Block Whole-Slide Images

Cytopathological examination is one of the main examinations for pleural effusion, and especially for many patients with advanced cancer, pleural effusion is the only accessible specimen for establishing a pathological diagnosis. The lack of cytopathologists and the high cost of gene detection present opportunities for the application of deep learning. In this retrospective analysis, data representing 1321 consecutive cases of pleural effusion were collected. We trained and evaluated our deep learning model based on several tasks, including the diagnosis of benign and malignant pleural effusion, the identification of the primary location of common metastatic cancer from pleural effusion, and the prediction of genetic alterations associated with targeted therapy. We achieved good results in identifying benign and malignant pleural effusions (0.932 AUC (area under the ROC curve)) and the primary location of common metastatic cancer (0.910 AUC). In addition, we analyzed ten genes related to targeted therapy in specimens and used them to train the model regarding four alteration statuses, which also yielded reasonable results (0.869 AUC for ALK fusion, 0.804 AUC for KRAS mutation, 0.644 AUC for EGFR mutation and 0.774 AUC for NONE alteration). Our research shows the feasibility and benefits of deep learning to assist in cytopathological diagnosis in clinical settings.

Trends in AI-powered Classification of Thyroid Neoplasms Based on Histopathology Images - a Systematic Review

Background

Assessment of thyroid nodules histopathology using AI is crucial for an accurate diagnosis. This systematic review analyzes recent works employing deep learning approaches for classifying thyroid nodules based on histopathology images, evaluating their performance, and identifying limitations.

Methods

Eligibility criteria focused on peer-reviewed English papers published in the last 5 years, applying deep learning to categorize thyroid histopathology images. The PubMed database was searched using relevant keyword combinations.

Results

Out of 103 articles, 11 studies met inclusion criteria. They used convolutional neural networks to classify thyroid neoplasm. Most studies aimed for basic tumor subtyping; however, 3 studies targeted the prediction of tumor-associated mutation. Accuracy ranged from 77% to 100%, with most over 90%.

Discussion

The findings from our analysis reveal the effectiveness of deep learning in identifying discriminative morphological patterns from histopathology images, thus enhancing the accuracy of thyroid nodule histopathological classification. Key limitations were small sample sizes, subjective annotation, and limited dataset diversity. Further research with larger diverse datasets, model optimization, and real-world validation is essential to translate these tools into clinical practice.