Current Trends of Artificial Intelligence for Colorectal Cancer Pathology Image Analysis: A Systematic Review

Colorectal cancer (CRC) is one of the most common cancers requiring early pathologic diagnosis using colonoscopy biopsy samples. Recently, artificial intelligence (AI) has made significant progress and shown promising results in the field of medicine despite several limitations. We performed a systematic review of AI use in CRC pathology image analysis to visualize the state-of-the-art. Studies published between January 2000 and January 2020 were searched in major online databases including MEDLINE (PubMed, Cochrane Library, and EMBASE). Query terms included "colorectal neoplasm," "histology," and "artificial intelligence." Of 9000 identified studies, only 30 studies consisting of 40 models were selected for review. The algorithm features of the models were gland segmentation (n = 25, 62%), tumor classification (n = 8, 20%), tumor microenvironment characterization (n = 4, 10%), and prognosis prediction (n = 3, 8%). Only 20 gland segmentation models met the criteria for quantitative analysis, and the model proposed by Ding et al. (2019) performed the best. Studies with other features were in the elementary stage, although most showed impressive results. Overall, the state-of-the-art is promising for CRC pathological analysis. However, datasets in most studies had relatively limited scale and quality for clinical application of this technique. Future studies with larger datasets and high-quality annotations are required for routine practice-level validation.

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

State-of-the-art artificial intelligence (AI) has recently gained considerable interest in the healthcare sector and has provided solutions to problems through automated diagnosis. Cytological examination is a crucial step in the initial diagnosis of cancer, although it shows limited diagnostic efficacy. Recently, AI applications in the processing of cytopathological images have shown promising results despite the elementary level of the technology. Here, we performed a systematic review with a quantitative analysis of recent AI applications in non-gynecological (non-GYN) cancer cytology to understand the current technical status. We searched the major online databases, including MEDLINE, Cochrane Library, and EMBASE, for relevant English articles published from January 2010 to January 2021. The searched query terms were: "artificial intelligence", "image processing", "deep learning", "cytopathology", and "fine-needle aspiration cytology." Out of 17,000 studies, only 26 studies (26 models) were included in the full-text review, whereas 13 studies were included for quantitative analysis. There were eight classes of AI models treated of according to target organs: thyroid (n = 11, 39%), urinary bladder (n = 6, 21%), lung (n = 4, 14%), breast (n = 2, 7%), pleural effusion (n = 2, 7%), ovary (n = 1, 4%), pancreas (n = 1, 4%), and prostate (n = 1, 4). Most of the studies focused on classification and segmentation tasks. Although most of the studies showed impressive results, the sizes of the training and validation datasets were limited. Overall, AI is also promising for non-GYN cancer cytopathology analysis, such as pathology or gynecological cytology. However, the lack of well-annotated, large-scale datasets with Z-stacking and external cross-validation was the major limitation found across all studies. Future studies with larger datasets with high-quality annotations and external validation are required.

Recent Applications of Artificial Intelligence from Histopathologic Image-Based Prediction of Microsatellite Instability in Solid Cancers: A Systematic Review

Cancers with high microsatellite instability (MSI-H) have a better prognosis and respond well to immunotherapy. However, MSI is not tested in all cancers because of the additional costs and time of diagnosis. Therefore, artificial intelligence (AI)-based models have been recently developed to evaluate MSI from whole slide images (WSIs). Here, we aimed to assess the current state of AI application to predict MSI based on WSIs analysis in MSI-related cancers and suggest a better study design for future studies. Studies were searched in online databases and screened by reference type, and only the full texts of eligible studies were reviewed. The included 14 studies were published between 2018 and 2021, and most of the publications were from developed countries. The commonly used dataset is The Cancer Genome Atlas dataset. Colorectal cancer (CRC) was the most common type of cancer studied, followed by endometrial, gastric, and ovarian cancers. The AI models have shown the potential to predict MSI with the highest AUC of 0.93 in the case of CRC. The relatively limited scale of datasets and lack of external validation were the limitations of most studies. Future studies with larger datasets are required to implicate AI models in routine diagnostic practice for MSI prediction.

Deep learning-based histopathological segmentation for whole slide images of colorectal cancer in a compressed domain

Automatic pattern recognition using deep learning techniques has become increasingly important. Unfortunately, due to limited system memory, general preprocessing methods for high-resolution images in the spatial domain can lose important data information such as high-frequency information and the region of interest. To overcome these limitations, we propose an image segmentation approach in the compressed domain based on principal component analysis (PCA) and discrete wavelet transform (DWT). After inference for each tile using neural networks, a whole prediction image was reconstructed by wavelet weighted ensemble (WWE) based on inverse discrete wavelet transform (IDWT). The training and validation were performed using 351 colorectal biopsy specimens, which were pathologically confirmed by two pathologists. For 39 test datasets, the average Dice score, the pixel accuracy, and the Jaccard score were 0.804 ± 0.125, 0.957 ± 0.025, and 0.690 ± 0.174, respectively. We can train the networks for the high-resolution image with the large region of interest compared to the result in the low-resolution and the small region of interest in the spatial domain. The average Dice score, pixel accuracy, and Jaccard score are significantly increased by 2.7%, 0.9%, and 2.7%, respectively. We believe that our approach has great potential for accurate diagnosis.

Current Trend of Artificial Intelligence Patents in Digital Pathology: A Systematic Evaluation of the Patent Landscape

The integration of digital pathology (DP) with artificial intelligence (AI) enables faster, more accurate, and thorough diagnoses, leading to more precise personalized treatment. As technology is advancing rapidly, it is critical to understand the current state of AI applications in DP. Therefore, a patent analysis of AI in DP is required to assess the application and publication trends, major assignees, and leaders in the field. We searched five major patent databases, namely, those of the USPTO, EPO, KIPO, JPO, and CNIPA, from 1974 to 2021, using keywords such as DP, AI, machine learning, and deep learning. We discovered 6284 patents, 523 of which were used for trend analyses on time series, international distribution, top assignees; word cloud analysis; and subject category analyses. Patent filing and publication have increased exponentially over the past five years. The United States has published the most patents, followed by China and South Korea (248, 117, and 48, respectively). The top assignees were Paige.AI, Inc. (New York City, NY, USA) and Siemens, Inc. (Munich, Germany) The primary areas were whole-slide imaging, segmentation, classification, and detection. Based on these findings, we expect a surge in DP and AI patent applications focusing on the digitalization of pathological images and AI technologies that support the vital role of pathologists.

Strong association of Torque teno virus/Torque teno-like minivirus to Kikuchi-Fujimoto lymphadenitis (histiocytic necrotizing lymphadenitis) on quantitative analysis

INTRODUCTION

Kikuchi-Fujimoto disease (KFD) is a rare benign lymphadenitis that mainly involves cervical lymph nodes of young Asian women with unknown etiology. Recently, we experienced a case of KFD found with Torque teno virus/Torque teno-like mini virus (TTV/TTMV) from a 26-year-old woman. TTV/TTMV is a genus of Circoviridae that causes necrotizing lymphadenitis in pigs, which shares the key histologic finding of KFD. The purpose of this study is to investigate the pathogenic role of TTV/TTMV in KFD by quantitative polymerase chain reaction (qPCR) analysis.

METHOD

We performed two-step qPCR specific to TTV/TTMV with formalin-fixed paraffin-embedded tissue of sequentially selected 100 KFD patients and 50 randomly selected, matched normal controls. Consequent direct sequencing was done for confirmation with PCR products.

RESULTS

PCR amplification of TTV and TTMV was found in a significantly higher proportion in KFDs than normal controls (TTV, 85% vs. 18%, p < 0.000; TTMV, 91% vs. 24%, p < 0.000). After the sequencing, KFD samples showed more sequence matching than control samples for TTMV (94% vs. 30%, p < 0.000).

CONCLUSION

This finding strongly suggests the possible implication of TTV/TTMV in the pathogenesis of KFD. Animal or in vivo experimental design should be followed in the future.Key Points• Kikuchi-Fujimoto disease (KFD) is rare and its etiology is still unclear.• Torque teno/Torque teno-like minivirus (TTV/TTMV) is a recently introduced virus in the Circoviridae family that causes necrotizing lymphadenitis in pigs, histologically similar to KFD.• We discovered the significantly increased TTV/TTMV viral loads in the KFD patients than normal controls, which implicates TTV/TTMV in the pathogenesis of KFD.

Calculated Tumor-Associated Neutrophils Are Associated with the Tumor-Stroma Ratio and Predict a Poor Prognosis in Advanced Gastric Cancer

The tumor-associated neutrophils (TANs) value and tumor-stroma ratio (TSR) are promising prognostic parameters in the tumor microenvironment. We aimed to evaluate the prognostic role and relationship of TANs and TSR in gastric cancer. Our study comprised 157 patients who underwent gastrectomy for advanced gastric cancer. TANs were assessed by immunohistochemical staining (CD15 and CD66b) and were analyzed with an image analyzer. TANs have been known to have different functional subpopulations of N1 (anti-tumor) and N2 (pro-tumor). We developed "calculated TANs with pro-tumor function (cN2; CD15 minus CD66b)". The TSR was evaluated using hematoxylin and eosin staining. High-grade CD15-positive, cN2 in the tumor center, and TSR were significantly related to poor disease-free survival (DFS). TSR and cN2 were independent prognostic factors for DFS (hazard ratio (HR) = 2.614; p = 0.001, HR = 3.976; p = 0.002) and cN2 in the tumor center showed a positive correlation with TSR (R = 0.179, p = 0.025). While CD66b stained both N1 and N2, CD15 detected most of N2. Combining both markers revealed a novel cN2, which was an independent marker of poor prognosis. The transformation from N1 to N2 predominantly occurred in the tumor center, and was associated with TSR.

High Poly(ADP-Ribose) Polymerase Expression Does Relate to Poor Survival in Solid Cancers: A Systematic Review and Meta-Analysis

Poly (ADP-ribose) polymerase (PARP) is a DNA damage repair protein, and its inhibitors have shown promising results in clinical trials. The prognostic significance of PARP is inconsistent in studies of various cancers. In the present study, we conducted a systematic review and meta-analysis to reveal the prognostic and clinicopathological significance of PARP expression in multiple solid cancers. We searched the MEDLINE, EMBASE, and Cochrane databases for relevant research articles published from 2005 to 2021. The pooled hazard ratio (HR) with confidence interval (CI) was calculated to investigate the relationship between PARP expression and survival in multiple solid cancers. In total, 10,667 patients from 31 studies were included. A significant association was found between higher PARP expression and overall survival (OS) (HR = 1.54, 95% CI = 1.34-1.76, p < 0.001), disease-free survival (DFS) (HR = 1.15, 95% CI = 1.10-1.21, p < 0.001), and progression-free survival (PFS) (HR = 1.05, 95% CI = 1.03-1.08, p < 0.001). Subgroup analyses showed that PARP overexpression was significantly related to poor OS in patients with breast cancers (HR = 1.38, 95% CI = 1.28-1.49, p < 0.001), ovary cancers (HR = 1.21, 95% CI = 1.10-1.33, p = 0.001), lung cancers (HR = 2.11, 95% CI = 1.29-3.45, p = 0.003), and liver cancers (HR = 3.29, 95% CI = 1.94-5.58, p < 0.001). Regarding ethnicity, Asian people have almost twice their worst survival rate compared to Caucasians. The pooled odds ratio analysis showed a significant relationship between higher PARP expression and larger tumour size, poor tumour differentiation, lymph node metastasis, distant metastasis, higher TNM stage and lymphovascular invasion, and positive immunoreactivity for Ki-67, BRCA1, and BRCA2. In addition, nuclear expression assessed by the QS system using Abcam and Santa Cruz Biotechnology seems to be the most commonly used and reproducible IHC method for assessing PARP expression. This meta-analysis revealed that higher PARP expression was associated with a worse OS, DFS, and PFS in patients with solid cancers. Moreover, inhibition of this pathway through its specific inhibitors may extend the survival of patients with higher PARP expression.

Prognostic significance of stem cell/ epithelial-mesenchymal transition markers in periampullary/pancreatic cancers: FGFR1 is a promising prognostic marker

BACKGROUND

Periampullary cancers (PAC) including pancreatic, ampulla of Vater (AOV), and common bile duct (CBD) cancers are highly aggressive with a lack of useful prognostic markers beyond T stage. However, T staging can be biased due to the anatomic complexity of this region. Recently, several markers related to cancer stem cells and epithelial-mesenchymal transition (EMT) such as octamer transcription factor-4 (Oct4) and fibroblast growth factor receptor 1 (FGFR1) respectively, have been proposed as new promising markers in other solid cancers. The aim of this study was to assess the expression and prognostic significance of stem cell/EMT markers in PACs.

METHODS

Formalin-fixed, paraffin-embedded tissues of surgically excised PACs from the laboratory archives from 1998 to 2014 were evaluated by immunohistochemical staining for stem cell/EMT markers using tissue microarray. The clinicopathologic parameters were documented and statistically analyzed with the immunohistochemical findings. Survival and recurrence data were collected and analyzed.

RESULTS

A total of 126 PAC cases were evaluated. The average age was 63 years, with 76 male and 50 female patient samples. Age less than 74 years, AOV cancers, lower T & N stage, lower tumor size, no lymphatic, vascular, perineural invasion and histologic well differentiation, intestinal type, no fibrosis, severe inflammation were significantly associated with the better overall survival High expression levels of FGFR1 as well as CK20, CDX2, and VEGF were significantly related to better overall survival, while other stem cell markers were not related. Similar findings were observed for tumor recurrence using disease-free survival.

CONCLUSIONS

In addition to other clinicopathologic parameters, severe fibrosis was related to frequent tumor recurrence, and high FGFR1 expression was associated with better overall survival. Histologic changes such as extensive fibrosis need to be investigated further in relation to EMT of PACs.

Identification of Torque Teno Virus/Torque Teno-Like Minivirus in the Cervical Lymph Nodes of Kikuchi-Fujimoto Lymphadenitis Patients (Histiocytic Necrotizing Lymphadenitis): A Possible Key to Idiopathic Disease

Kikuchi-Fujimoto disease (KFD) is rare, and many infectious agents have been suspected for its etiology. This report presents an interesting case of KFD found with torque teno virus/torque teno minivirus (TTV/TTMV), which closely resembles the circovirus that causes necrotizing lymphadenitis in pigs. Three Korean patients showed several enlarged lymph nodes in their neck. Quantitative polymerase chain reaction (qPCR) and subsequent DNA sequencing for TTV/TTMV using formalin-fixed paraffin-embedded tissue were performed. Histologic examination demonstrated typical features of KFD. qPCR showed successful amplification of TTV/TTMV, and DNA sequencing confirmed the results. It is the first report of TTV/TTMV presence in three patients with KFD.

Intratumoral Budding in Pretreatment Biopsies, among Tumor Microenvironmental Components, Can Predict Prognosis and Neoadjuvant Therapy Response in Colorectal Adenocarcinoma

Background and Objectives: The prediction of the prognosis and effect of neoadjuvant therapy is vital for patients with advanced or unresectable colorectal carcinoma (CRC). Materials and Methods: We investigated several tumor microenvironment factors, such as intratumoral budding (ITB), desmoplastic reaction (DR), and Klintrup-Mäkinen (KM) inflammation grade, and the tumor-stroma ratio (TSR) in pretreatment biopsy samples (PBSs) collected from patients with advanced or unresectable CRC. A total of 85 patients with 74 rectal carcinomas and 11 colon cancers treated at our hospital were enrolled; 66 patients had curative surgery and 19 patients received palliative treatment. Results: High-grade ITB was associated with recurrence (p = 0.002), death (p = 0.034), and cancer-specific death (p = 0.034). Immature DR was associated with a higher grade of clinical tumor-node-metastasis stage (cTNM) (p = 0.045), cN category (p = 0.045), and cM category (p = 0.046). The KM grade and TSR were not related to any clinicopathological factors. High-grade ITB had a significant relationship with tumor regression in patients who received curative surgery (p = 0.049). Conclusions: High-grade ITB in PBSs is a potential unfavorable prognostic factor for patients with advanced CRC. Immature DR, TSR, and KM grade could not predict prognosis or therapy response in PBSs.

A machine-learning expert-supporting system for diagnosis prediction of lymphoid neoplasms using a probabilistic decision-tree algorithm and immunohistochemistry profile database

BACKGROUND

Immunohistochemistry (IHC) has played an essential role in the diagnosis of hematolymphoid neoplasms. However, IHC interpretations can be challenging in daily practice, and exponentially expanding volumes of IHC data are making the task increasingly difficult. We therefore developed a machine-learning expert-supporting system for diagnosing lymphoid neoplasms.

METHODS

A probabilistic decision-tree algorithm based on the Bayesian theorem was used to develop mobile application software for iOS and Android platforms. We tested the software with real data from 602 training and 392 validation cases of lymphoid neoplasms and compared the precision hit rates between the training and validation datasets.

RESULTS

IHC expression data for 150 lymphoid neoplasms and 584 antibodies was gathered. The precision hit rates of 94.7% in the training data and 95.7% in the validation data for lymphomas were not statistically significant. Results in most B-cell lymphomas were excellent, and generally equivalent performance was seen in T-cell lymphomas. The primary reasons for lack of precision were atypical IHC profiles for certain cases (e.g., CD15-negative Hodgkin lymphoma), a lack of disease-specific markers, and overlapping IHC profiles of similar diseases.

CONCLUSIONS

Application of the machine-learning algorithm to diagnosis precision produced acceptable hit rates in training and validation datasets. Because of the lack of origin- or disease-specific markers in differential diagnosis, contextual information such as clinical and histological features should be taken into account to make proper use of this system in the pathologic decision-making process.

High Expression of PD-L1 Is Associated with Better Survival in Pancreatic/Periampullary Cancers and Correlates with Epithelial to Mesenchymal Transition

Periampullary cancers (PACs) are characterized by tumor-infiltrating lymphocytes (TILs), severe fibrosis, and epithelial to mesenchymal transition (EMT). The immune checkpoint marker programmed death-1 (PD-1) and its ligands 1 and 2 have gained popularity in cancers with TILs. Evidence suggests a strong relationship between immune checkpoint markers and EMT in cancers. Here, we evaluated the expression and prognostic significance of immune checkpoint and EMT markers in PAC using an automated image analyzer. Formalin-fixed, paraffin-embedded surgically excised PAC tissues from laboratory archives (1998-2014) were evaluated by immunohistochemical staining for PD-1, PD-L1, and PD-L2 in a tissue microarray. In total, 115 PAC patients (70 males and 45 females) with an average age of 63 years were analyzed. Location, gross type, size, radial resection margin, N-M stage, lymphatic invasion, vascular invasion, perineural invasion, histologically well-differentiated severe inflammation, and high PD-L1 expression were significantly associated with recurrence. Higher PD-L1 expression, but not PD-1 and PD-L2, was significantly related to better overall survival (OS) and disease-free survival (DFS). PD-L1 and PD-L2 were significantly related to EMT markers. Aside from other clinicopathologic parameters, high PD-L1 expression was significantly related to better OS and DFS of PAC patients. Moreover, immune checkpoint markers were significantly associated with EMT markers. Therefore, PD-L1 expression can be a good prognostic marker to guide future immune target-based therapies in PAC patients.

Diagnosis prediction of tumours of unknown origin using ImmunoGenius, a machine learning-based expert system for immunohistochemistry profile interpretation

BACKGROUND

Immunohistochemistry (IHC) remains the gold standard for the diagnosis of pathological diseases. This technique has been supporting pathologists in making precise decisions regarding differential diagnosis and subtyping, and in creating personalized treatment plans. However, the interpretation of IHC results presents challenges in complicated cases. Furthermore, rapidly increasing amounts of IHC data are making it even harder for pathologists to reach to definitive conclusions.

METHODS

We developed ImmunoGenius, a machine-learning-based expert system for the pathologist, to support the diagnosis of tumors of unknown origin. Based on Bayesian theorem, the most probable diagnoses can be drawn by calculating the probabilities of the IHC results in each disease. We prepared IHC profile data of 584 antibodies in 2009 neoplasms based on the relevant textbooks. We developed the reactive native mobile application for iOS and Android platform that can provide 10 most possible differential diagnoses based on the IHC input.

RESULTS

We trained the software using 562 real case data, validated it with 382 case data, tested it with 164 case data and compared the precision hit rate. Precision hit rate was 78.5, 78.0 and 89.0% in training, validation and test dataset respectively. Which showed no significant difference. The main reason for discordant precision was lack of disease-specific IHC markers and overlapping IHC profiles observed in similar diseases.

CONCLUSION

The results of this study showed a potential that the machine-learning algorithm based expert system can support the pathologic diagnosis by providing second opinion on IHC interpretation based on IHC database. Incorporation with contextual data including the clinical and histological findings might be required to elaborate the system in the future.

Validation of a Machine Learning Expert Supporting System, ImmunoGenius, Using Immunohistochemistry Results of 3000 Patients with Lymphoid Neoplasms

(1) Background: Differential diagnosis using immunohistochemistry (IHC) panels is a crucial step in the pathological diagnosis of hematolymphoid neoplasms. In this study, we evaluated the prediction accuracy of the ImmunoGenius software using nationwide data to validate its clinical utility. (2) Methods: We collected pathologically confirmed lymphoid neoplasms and their corresponding IHC results from 25 major university hospitals in Korea between 2015 and 2016. We tested ImmunoGenius using these real IHC panel data and compared the precision hit rate with previously reported diagnoses. (3) Results: We enrolled 3052 cases of lymphoid neoplasms with an average of 8.3 IHC results. The precision hit rate was 84.5% for these cases, whereas it was 95.0% for 984 in-house cases. (4) Discussion: ImmunoGenius showed excellent results in most B-cell lymphomas and generally showed equivalent performance in T-cell lymphomas. The primary reasons for inaccurate precision were atypical IHC profiles of certain cases, lack of disease-specific markers, and overlapping IHC profiles of similar diseases. We verified that the machine-learning algorithm could be applied for diagnosis precision with a generally acceptable hit rate in a nationwide dataset. Clinical and histological features should also be taken into account for the proper use of this system in the decision-making process.

Current status of cytopathology practice in Korea: impact of the coronavirus pandemic on cytopathology practice

BACKGROUND

The Continuous Quality Improvement program for cytopathology in 2020 was completed during the coronavirus pandemic. In this study, we report the result of the quality improvement program.

METHODS

Data related to cytopathology practice from each institute were collected and processed at the web-based portal. The proficiency test was conducted using glass slides and whole-slide images (WSIs). Evaluation of the adequacy of gynecology (GYN) slides from each institution and submission of case glass slides and WSIs for the next quality improvement program were performed.

RESULTS

A total of 214 institutions participated in the annual cytopathology survey in 2020. The number of entire cytopathology specimens was 8,220,650, a reduction of 19.0% from the 10,111,755 specimens evaluated in 2019. Notably, the number of respiratory cytopathology specimens, including sputum and bronchial washing/ brushing significantly decreased by 86.9% from 2019, which could be attributed to the global pandemic of coronavirus disease. The ratio of cases with atypical squamous cells to squamous intraepithelial lesions was 4.10. All participating institutions passed the proficiency test and the evaluation of adequacy of GYN slides.

CONCLUSIONS

Through the Continuous Quality Improvement program, the effect of coronavirus disease 2019 pandemic, manifesting with a reduction in the number of cytologic examinations, especially in respiratory-related specimen has been identified. The Continuous Quality Improvement Program of the Korean Society for Cytopathology can serve as the gold standard to evaluate the current status of cytopathology practice in Korea.