Morphometric Analysis of Lateral Cervical Lymph Node Metastasis in Papillary Thyroid Carcinoma Using Digital Pathology

Digital pathology uses digitized images for cancer research. We aimed to assess morphometric parameters using digital pathology for predicting recurrence in patients with papillary thyroid carcinoma (PTC) and lateral cervical lymph node (LN) metastasis. We analyzed 316 PTC patients and assessed the longest diameter and largest area of metastatic focus in LNs using a whole slide imaging scanner. In digital pathology assessment, the longest diameters and largest areas of metastatic foci in LNs were positively correlated with traditional optically measured diameters (R = 0.928 and R2 = 0.727, p < 0.001 and p < 0.001, respectively). The optimal cutoff diameter was 8.0 mm in both traditional microscopic (p = 0.009) and digital pathology (p = 0.016) evaluations, with significant differences in progression-free survival (PFS) observed at this cutoff (p = 0.006 and p = 0.002, respectively). The predictive area's cutoff was 35.6 mm2 (p = 0.005), which significantly affected PFS (p = 0.015). Using an 8.0-mm cutoff in traditional microscopic evaluation and a 35.6-mm2 cutoff in digital pathology showed comparable predictive results using the proportion of variation explained (PVE) methods (2.6% vs. 2.4%). Excluding cases with predominant cystic changes in LNs, the largest metastatic areas by digital pathology had the highest PVE at 3.9%. Furthermore, high volume of LN metastasis (p = 0.001), extranodal extension (p = 0.047), and high ratio of metastatic LNs (p = 0.006) were associated with poor prognosis. Both traditional microscopic and digital pathology evaluations effectively measured the longest diameter of metastatic foci in LNs. Moreover, digital pathology offers limited advantages in predicting PFS of patients with lateral cervical LN metastasis of PTC, especially those without predominant cystic changes in LNs.

The use of telecytology for the evaluation of thyroid nodules fine-needle aspiration biopsy specimens: a systematic review

PURPOSE

Fine needle aspiration biopsy (FNAB) is currently the gold standard for diagnosis and treatment of thyroid nodules, but the growing need for anatomic pathology services in small communities is becoming a challenge. Telecytology (TC) is defined as the electronic transmission of cytological digital images, and allows for the collection of samples, primary diagnosis, and other applications without the physical presence of a pathologist. Our aim is to systematically report, summarize, and critically analyze the most up to date applications of TC to thyroid nodules FNAB evaluation.

METHODS

We performed a systematic literature review by searching PubMed, Embase, and Cochrane Library databases. Only studies published in peer-reviewed scientific journals were included. Data were extracted using the PICO framework and critically analyzed. PRISMA guidelines were applied, and the risk of bias in the included studies was assessed using the ROBINS-I tools. The methodological quality was assessed following GRADE criteria.

RESULTS

We included 13 observational studies, resulting in a total of 3856 evaluated FNAB specimens. The majority of studies (63.6%) showed an excellent concordance rate of diagnosis via TC and conventional cytology. TC can be used to perform preliminary assessment of samples with a concordance rate ranging from 74 and 100%, showing a significant reduction of the non-diagnostic rate. Image quality was referred to as perfect or nearly perfect in most cases, regardless of telecytology technique.

CONCLUSION

Telecytology could be a valuable implementation for thyroid FNAB evaluation both for primary diagnosis and preliminary assessment of samples.

Revisiting the utility of identifying nuclear grooves as unique nuclear changes by an object detector model

BACKGROUND

Among other structures, nuclear grooves are vastly found in papillary thyroid carcinoma (PTC). Considering that the application of artificial intelligence in thyroid cytology has potential for diagnostic routine, our goal was to develop a new supervised convolutional neural network capable of identifying nuclear grooves in Diff-Quik stained whole-slide images (WSI) obtained from thyroid fineneedle aspiration.

METHODS

We selected 22 Diff-Quik stained cytological slides with cytological diagnosis of PTC and concordant histological diagnosis. Each of the slides was scanned, forming a WSI. Images that contained the region of interest were obtained, followed by pre-formatting, annotation of the nuclear grooves and data augmentation techniques. The final dataset was divided into training and validation groups in a 7:3 ratio.

RESULTS

This is the first artificial intelligence model based on object detection applied to nuclear structures in thyroid cytopathology. A total of 7,255 images were obtained from 22 WSI, totaling 7,242 annotated nuclear grooves. The best model was obtained after it was submitted 15 times with the train dataset (14th epoch), with 67% true positives, 49.8% for sensitivity and 43.1% for predictive positive value.

CONCLUSIONS

The model was able to develop a structure predictor rule, indicating that the application of an artificial intelligence model based on object detection in the identification of nuclear grooves is feasible. Associated with a reduction in interobserver variability and in time per slide, this demonstrates that nuclear evaluation constitutes one of the possibilities for refining the diagnosis through computational models.

Diagnostic Significance of Ultrasound-Guided Fine-Needle Aspiration Biopsy and on-Site Assessment by Pathologists for Thyroid Micronodules

Objective

Ultrasound-guided fine-needle aspiration biopsy (US-FNAB) is a safe and effective method for screening malignant thyroid nodules. The purpose of this study was to compare the diagnostic effectiveness of US-FNAB for nodules of different sizes.

Methods

A total of 1085 patients with thyroid nodules who underwent US-FNAB between January 2021 and July 2023 were included in the study. The patients were divided into three groups based on the maximum diameter of the nodules: there were 324 patients with thyroid nodules ≤5 mm in Group A, 537 patients with thyroid nodules between 6 mm and 10 mm in Group B, and 224 patients with thyroid nodules >10 mm in Group C. The US-FNAB satisfactory specimen rate, biopsy time and cytopathological results for the three groups were collected and compared with the postoperative pathological results.

Results

The US-FNAB satisfactory specimen rates for Groups A, B and C were 84.57% (274/324), 90.13% (484/537) and 94.64% (212/224), respectively. The average biopsy times for Groups A, B, and C were 100.84 ± 41.58 s, 91.20 ± 32.53 s, and 79.01 ± 29.62 s, respectively. In Groups A, B, and C, 103, 192 and 73 patients, respectively, underwent surgery, and the malignancy rates were 88.35%, 85.42% and 72.6%, respectively. The sensitivity, specificity, positive predictive value, and negative predictive value of US-FNAB in Group A were 78.26%, 81.82%, 97.30%, 31.03%; respectively; those in Group B were 73.78%,85.71%,96.80%, and 35.82%, respectively; and those in Group C were 75.47%, 85.00%, 93.02% and 56.67%, respectively.

Conclusion

The US-FNAB satisfactory specimen rate for thyroid nodules ≤5 mm was relatively low, but the size of nodules had no effect on the diagnostic sensitivity of US-FNAB; additionally, nodules ≤5 mm had a higher probability of malignancy. Therefore, it is necessary to perform US-FNAB for thyroid nodules with a diameter ≤5 mm with malignant signs.

Digital analyses of nuclear features can help discriminate among non-invasive follicular thyroid neoplasm with papillary-like nuclear features, papillary thyroid carcinoma follicular subtype, and follicular carcinoma in cytological specimens

BACKGROUND

As it stands, the diagnosis of non-invasive follicular thyroid neoplasm with papillary-like nuclear features (NIFTP) is primarily based on histological analysis. We hypothesised that computerised analysis of nuclear images of cytological specimens could be used to differentiate NIFTP from papillary thyroid carcinoma follicular subtype (PTCFS) and follicular carcinoma (FC), influencing patient management.

METHODS

We employed a retrospective analytical observational study based on nuclear morphometric variables of cytological material from thyroid nodules classified as PTCFS, NIFTP, or FC. Five cases of each entity were analysed. Cytological slides were photographed, and 1170 cells for each entity were analysed digitally. The captured images were evaluated (blindly) using the ImageJ software package. The morphometric evaluation included area, perimeter, width, height, and circularity. Numerical variables were expressed as mean, median, minimum, and maximum (min; max) values. Kruskal-Wallis and Dunn's tests were used with a 5% significance level.

RESULTS

Regarding nuclear analysis, all variables differed among the three groups (p < 0.001). Given the interdependence among the variables, these data indicated that nuclear size was greatest in the NIFTP group, followed by FC and PTCFS.

DISCUSSION/CONCLUSION

Our analysis of the digital images, with a focus on nuclear parameters, found significantly difference among cytological specimens from cases of NIFTP, PTCFS and FC. Thus, this tool has the potential to provide additional information that may help in the diagnosis of NIFTP, even during the preoperative period. Additional studies are needed to create protocols, evaluate the applicability of nuclear morphological and morphometric parameters-focusing on digital pathology-and create algorithms and tools to assist cytopathologists with their diagnostic routines.

Comparison of glass and digital slides for cervical cytopathology screening and interpretation

Cervical cancer is the second most common form of cancer and a leading cause of premature death among women aged 15 to 44 worldwide. In Brazil, there is a high prevalence of infection by the human papillomavirus - HPV. Digital pathology optimizes time and space for reading cervicovaginal cytology slides. We evaluated the feasibility of using whole slide images (WSI) for the routine interpretation of cytology exams. A total of 99 cases of vaginal cytology were selected from a reference laboratory in Northeastern Brazil. Three cytotechnicians participated in the study. Cellular atypia was the one that most presented concordance values. Two observers almost perfectly agreed (k = 0.86 and k = 0.84, respectively) on the negative diagnoses. The performance of the evaluators for NILM (negative for intraepithelial lesion and malignancy) showed high reproducibility and sensitivity in the digital slides, mainly between evaluators A and C. In contrast, the microbiology group showed disagreement between the diagnoses by digital slides and the standard- gold. The concordance between the digital diagnoses and the gold standard for ASCUS was 89%. In the inflammatory category, Spearman's test showed similar results between raters A, B, and C (rs = 0.47, rs = 0.41, and rs = 0.47, respectively). This study reports the diagnostic validation using digital slides in view of the need to optimize the cytology visualization process. Our experience shows good diagnostic agreement between digital and optical microscopy in several analyzed categories, but mainly in relation to cellular atypia and inflammatory processes.

Effects of Rapid On-Site Evaluation on Diagnostic Accuracy of Thyroid Fine-Needle Aspiration

INTRODUCTION

Fine-needle aspiration (FNA) is well-established for the evaluation of suspicious thyroid nodules. However, a significant proportion is nondiagnostic. Rapid on-site evaluation (ROSE) has been proposed to improve the overall adequacy of FNA.

METHODS

Retrospective cohort study comparing adequacy of thyroid FNA findings pre- and postimplementation of ROSE at a tertiary center in Switzerland. Patients undergoing thyroid FNA from January 2016 to December 2019 were included. The primary outcome was the rate of nondiagnostic findings (Bethesda System for Reporting Thyroid Cytopathology category I).

RESULTS

In total, 410 thyroid nodule FNAs were performed. Of those, 309 with standard FNA and 101 with ROSE. The majority of patients were female (71%), with a median age of 56 years (IQR 46-68) and a nodule diameter of 1.9 cm (IQR 1.2-2.9). Implementation of ROSE led to a decrease in nondiagnostic findings from 41.1% to 23.8%, with an odds ratio of 0.42 (95% CI: 0.24-0.72; p = 0.002). Implementation of ROSE was associated with significantly higher rates of Bethesda category III (27.7% vs. 19.1%), category IV (15.8% vs. 5.5%), and Bethesda category VI (6.9% vs. 2.3%). Repeated FNA was performed in 29.1% before and 20.8% after implementation of ROSE (p = 0.18). The mean number of FNA per nodule was reduced from 1.4 (0.6) to 1.2 (0.4) with ROSE (p = 0.04).

CONCLUSIONS

Implementation of ROSE of thyroid nodule specimen improved diagnostic adequacy of FNA, reducing nondiagnostic findings. However, due to increased equivocal findings (Bethesda category III), there was no significant reduction of repeat FNA.

Application and performance of artificial intelligence technology in cytopathology

Deep learning algorithms and artificial intelligence (AI) are making great progress in their capacity to evaluate and interpret image data recent advancements in computer vision and machine learning. The first use of AI in a pathology lab was in cytopathology, when a computer-assisted Pap test screening was created. Initially designed to diagnose rather than screen, there was a lot of disagreement concerning their wide use to clinical specimens. However, whole-slide imaging of both gynaecological and non-gynaecological histopathology have been the subject of recent AI work. An overview of the literature on AI in cytopathology is provided in this brief review. To be more precise, it intends to emphasize the relevance of applications of AI algorithms to gynaecological and non-gynaecologic cytology. Between January 2000 and December 2021, a search on artificial intelligence in cytopathology was conducted in several well-known databases, including PubMed, Web of Science, Scopus, Embase, and Google Scholar. Only full-text papers that could be accessed online were evaluated.

Whole-Slide Imaging: Updates and Applications in Papillary Thyroid Carcinoma

Whole-slide imaging (WSI) has wide spectrum of application in histopathology, especially in the study of cancer including papillary thyroid carcinoma. The main applications of WSI system include research, teaching, and assessment and recently pathology practices. The other major advantages of WSI over histological sections on glass slides are easier storage and sharing of information as well as adaptation of use in artificial intelligence. The applications of WSI depend on factors such as volume of services requiring WSI, physical factors (computer server, bandwidth limitation of networks, storages requirements for data), adaption of the WSI images with the laboratory workflow, personnel (IT expert, pathologist, technicians) adaptation to the WSI workflow, validation studies, ethics, and cost efficiency of the application(s).

Use of Machine Learning-Based Software for the Screening of Thyroid Cytopathology Whole Slide Images

CONTEXT.—

The use of whole slide images (WSIs) in diagnostic pathology presents special challenges for the cytopathologist. Informative areas on a direct smear from a thyroid fine-needle aspiration biopsy (FNAB) smear may be spread across a large area comprising blood and dead space. Manually navigating through these areas makes screening and evaluation of FNA smears on a digital platform time-consuming and laborious. We designed a machine learning algorithm that can identify regions of interest (ROIs) on thyroid fine-needle aspiration biopsy WSIs.

OBJECTIVE.—

To evaluate the ability of the machine learning algorithm and screening software to identify and screen for a subset of informative ROIs on a thyroid FNA WSI that can be used for final diagnosis.

DESIGN.—

A representative slide from each of 109 consecutive thyroid fine-needle aspiration biopsies was scanned. A cytopathologist reviewed each WSI and recorded a diagnosis. The machine learning algorithm screened and selected a subset of 100 ROIs from each WSI to present as an image gallery to the same cytopathologist after a washout period of 117 days.

RESULTS.—

Concordance between the diagnoses using WSIs and those using the machine learning algorithm-generated ROI image gallery was evaluated using pairwise weighted κ statistics. Almost perfect concordance was seen between the 2 methods with a κ score of 0.924.

CONCLUSIONS.—

Our results show the potential of the screening software as an effective screening tool with the potential to reduce cytopathologist workloads.

Computer-assisted image analysis of cytological specimens clarify the correlation between nuclear size and intranuclear cytoplasmic inclusions regardless of BRAFV600E mutation in papillary thyroid carcinoma

OBJECTIVE

The morphological features of nuclei in cytological and histological specimens were compared and examined for the presence of BRAFV600E mutation and the appearance rate of intranuclear cytoplasmic inclusions (NI).

METHODS

BRAFV600E mutation was identified using a mutation-specific antibody (clone; VE1) in 103 thyroid papillary carcinoma cases at Gunma University Hospital. The nuclear area, perimeter, and roundness of the corresponding cytological specimens and haematoxylin and eosin-stained specimens were analysed using image analysis software, and the appearance rate of NI was calculated and compared.

RESULTS

BRAFV600E mutation was detected in 71 (69%) cases. The appearance rate of NI was significantly higher in the BRAFV600E mutation-positive group in cytological and histological specimens (P = .0070 and .0184, respectively). Significant differences were observed between the BRAFV600E mutation-negative and -positive groups in the average nuclear area and average nuclear perimeter in cytological specimens (P = .0137 and .0152, respectively). In addition, nuclear enlargement was correlated with the appearance rate of NI regardless of the presence of BRAFV600E mutation in cytological specimens. In the BRAFV600E mutation-negative group, the nuclear area and perimeter were significantly smaller in the lymph node metastasis-positive cases (P = .0182 and .0260, respectively).

CONCLUSION

This study found that the appearance rate of NI was positively correlated with the nuclear area and perimeter and negatively correlated with nuclear roundness in cytological specimens. Furthermore, these results were observed regardless of the existence of BRAFV600E mutation. These results have never been previously reported and clearly demonstrate the usefulness of cytological specimens in computer-assisted image analysis.

Applications of Computational Pathology in Head and Neck Cytopathology

BACKGROUND

The application of computational technology to head and neck cytology material has been explored experimentally in several areas with a variety of potential applications.

SUMMARY

This review summarizes the application of these techniques to the diagnosis of thyroid, salivary gland, and other head and neck fine-needle aspiration specimens. Current limitations and potential future applications in diagnosis are discussed along with the possibilities for therapeutic applications of computational methodology. Key Message: Particularly promising applications include resolving diagnostic uncertainty in indeterminate thyroid aspirates and assessing the tumor microenvironment in response to immunotherapy for squamous cell carcinoma.

Artificial Intelligence for Personalized Medicine in Thyroid Cancer: Current Status and Future Perspectives

Thyroid cancers (TC) have increasingly been detected following advances in diagnostic methods. Risk stratification guided by refined information becomes a crucial step toward the goal of personalized medicine. The diagnosis of TC mainly relies on imaging analysis, but visual examination may not reveal much information and not enable comprehensive analysis. Artificial intelligence (AI) is a technology used to extract and quantify key image information by simulating complex human functions. This latent, precise information contributes to stratify TC on the distinct risk and drives tailored management to transit from the surface (population-based) to a point (individual-based). In this review, we started with several challenges regarding personalized care in TC, for example, inconsistent rating ability of ultrasound physicians, uncertainty in cytopathological diagnosis, difficulty in discriminating follicular neoplasms, and inaccurate prognostication. We then analyzed and summarized the advances of AI to extract and analyze morphological, textural, and molecular features to reveal the ground truth of TC. Consequently, their combination with AI technology will make individual medical strategies possible.

Impact of image analysis and artificial intelligence in thyroid pathology, with particular reference to cytological aspects

OBJECTIVE

Thyroid pathology has great potential for automated/artificial intelligence algorithm application as the incidence of thyroid nodules is increasing and the indeterminate interpretation rate of fine-needle aspiration remains relatively high. The aim of the study is to review the published literature on automated image analysis and artificial intelligence applications to thyroid pathology with whole-slide imaging.

METHODS

Systematic search was carried out in electronic databases. Studies dealing with thyroid pathology and use of automated algorithms applied to whole-slide imaging were included. Quality of studies was assessed with a modified QUADAS-2 tool.

RESULTS

Of 919 retrieved articles, 19 were included. The main themes addressed were the comparison of automated assessment of immunohistochemical staining with manual pathologist's assessment, quantification of differences in cellular and nuclear parameters among tumour entities, and discrimination between benign and malignant nodules. Correlation coefficients with manual assessment were higher than 0.76 and diagnostic performance of automated models was comparable with an expert pathologist diagnosis. Computational difficulties were related to the large size of whole-slide images.

CONCLUSIONS

Overall, the results are promising and it is likely that, with the resolution of technical issues, the application of automated algorithms in thyroid pathology will increase and be adopted following suitable validation studies.

Current state of whole slide imaging use in cytopathology: Pros and pitfalls

Whole slide imaging (WSI) allows generation of large whole slide images and their navigation with zoom in and out like a true virtual microscope. It has become widely used in surgical pathology for many purposes, such as education and training, research activity, teleconsultation, and primary diagnosis. However, in cytopathology, the use of WSI has been lagging behind histology, mainly due to the cytological specimen's characteristics, as groups of cells of different thickness are distributed throughout the slide. To allow the same focusing capability of light microscope, slides have to be scanned at multiple focal planes, at the cost of longer scan times and larger file size. These are the main technical pitfalls of WSI for cytopathology, partly overcome by solutions like liquid-based preparations. Validation studies for the use in primary diagnosis are less numerous and more heterogeneous than in surgical pathology. WSI has been proved effective for training students and successfully used in proficiency testing, allowing the creation of digital cytology atlases. Longer scan times are also a barrier for use in rapid on-site evaluation, but WSI retains its advantages of easy sharing of images for consultation, multiple simultaneous viewing in different locations, the possibility of unlimited annotations and easy integration with medical records. Moreover, digital slides set the laboratory free from reliance on a physical glass slide, with no more concern of fading of stain or slide breakage. Costs are still a problem for small institutions, but WSI can also represent the beginning of a more efficient way of working.

Application of a machine learning algorithm to predict malignancy in thyroid cytopathology

BACKGROUND

The Bethesda System for Reporting Thyroid Cytopathology (TBSRTC) comprises 6 categories used for the diagnosis of thyroid fine-needle aspiration biopsy (FNAB). Each category has an associated risk of malignancy, which is important in the management of a thyroid nodule. More accurate predictions of malignancy may help to reduce unnecessary surgery. A machine learning algorithm (MLA) was developed to evaluate thyroid FNAB via whole slide images (WSIs) to predict malignancy.

METHODS

Files were searched for all thyroidectomy specimens with preceding FNAB over 8 years. All cytologic and surgical pathology diagnoses were recorded and correlated for each nodule. One representative slide from each case was scanned to create a WSI. An MLA was designed to identify follicular cells and predict the malignancy of the final pathology. The test set comprised cases blindly reviewed by a cytopathologist who assigned a TBSRTC category. The area under the receiver operating characteristic curve was used to assess the MLA performance.

RESULTS

Nine hundred eight FNABs met the criteria. The MLA predicted malignancy with a sensitivity and specificity of 92.0% and 90.5%, respectively. The areas under the curve for the prediction of malignancy by the cytopathologist and the MLA were 0.931 and 0.932, respectively.

CONCLUSIONS

The performance of the MLA in predicting thyroid malignancy from FNAB WSIs is comparable to the performance of an expert cytopathologist. When the MLA and electronic medical record diagnoses are combined, the performance is superior to the performance of either alone. An MLA may be used as an adjunct to FNAB to assist in refining the indeterminate categories.