-
The American Journal of Pathology Sep 2023Thyroid cancer is the most common malignant endocrine tumor. The key test to assess preoperative risk of malignancy is cytologic evaluation of fine-needle aspiration...
Thyroid cancer is the most common malignant endocrine tumor. The key test to assess preoperative risk of malignancy is cytologic evaluation of fine-needle aspiration biopsies (FNABs). The evaluation findings can often be indeterminate, leading to unnecessary surgery for benign post-surgical diagnoses. We have developed a deep-learning algorithm to analyze thyroid FNAB whole-slide images (WSIs). We show, on the largest reported data set of thyroid FNAB WSIs, clinical-grade performance in the screening of determinate cases and indications for its use as an ancillary test to disambiguate indeterminate cases. The algorithm screened and definitively classified 45.1% (130/288) of the WSIs as either benign or malignant with risk of malignancy rates of 2.7% and 94.7%, respectively. It reduced the number of indeterminate cases (N = 108) by reclassifying 21.3% (N = 23) as benign with a resultant risk of malignancy rate of 1.8%. Similar results were reproduced using a data set of consecutive FNABs collected during an entire calendar year, achieving clinically acceptable margins of error for thyroid FNAB classification.
Topics: Humans; Cytology; Deep Learning; Thyroid Neoplasms; Algorithms
PubMed: 37611969
DOI: 10.1016/j.ajpath.2023.05.011 -
Biomolecules Jul 2023Humankind is witnessing a gradual increase in cancer incidence, emphasizing the importance of early diagnosis and treatment, and follow-up clinical protocols. Oral or...
Humankind is witnessing a gradual increase in cancer incidence, emphasizing the importance of early diagnosis and treatment, and follow-up clinical protocols. Oral or mouth cancer, categorized under head and neck cancers, requires effective screening for timely detection. This study proposes a framework, OralNet, for oral cancer detection using histopathology images. The research encompasses four stages: (i) Image collection and preprocessing, gathering and preparing histopathology images for analysis; (ii) feature extraction using deep and handcrafted scheme, extracting relevant features from images using deep learning techniques and traditional methods; (iii) feature reduction artificial hummingbird algorithm (AHA) and concatenation: Reducing feature dimensionality using AHA and concatenating them serially and (iv) binary classification and performance validation with three-fold cross-validation: Classifying images as healthy or oral squamous cell carcinoma and evaluating the framework's performance using three-fold cross-validation. The current study examined whole slide biopsy images at 100× and 400× magnifications. To establish OralNet's validity, 3000 cropped and resized images were reviewed, comprising 1500 healthy and 1500 oral squamous cell carcinoma images. Experimental results using OralNet achieved an oral cancer detection accuracy exceeding 99.5%. These findings confirm the clinical significance of the proposed technique in detecting oral cancer presence in histology slides.
Topics: Humans; Carcinoma, Squamous Cell; Squamous Cell Carcinoma of Head and Neck; Mouth Neoplasms; Algorithms; Head and Neck Neoplasms
PubMed: 37509126
DOI: 10.3390/biom13071090 -
The American Journal of Pathology Dec 2023Accurate proliferation rate quantification can be used to devise an appropriate treatment for breast cancer. Pathologists use breast tissue biopsy glass slides stained...
Accurate proliferation rate quantification can be used to devise an appropriate treatment for breast cancer. Pathologists use breast tissue biopsy glass slides stained with hematoxylin and eosin to obtain grading information. However, this manual evaluation may lead to high costs and be ineffective because diagnosis depends on the facility and the pathologists' insights and experiences. Convolutional neural network acts as a computer-based observer to improve clinicians' capacity in grading breast cancer. Therefore, this study proposes a novel scheme for automatic breast cancer malignancy grading from invasive ductal carcinoma. The proposed classifiers implement multistage transfer learning incorporating domain and histopathologic transformations. Domain adaptation using pretrained models, such as InceptionResNetV2, InceptionV3, NASNet-Large, ResNet50, ResNet101, VGG19, and Xception, was applied to classify the ×40 magnification BreaKHis data set into eight classes. Subsequently, InceptionV3 and Xception, which contain the domain and histopathology pretrained weights, were determined to be the best for this study and used to categorize the Databiox database into grades 1, 2, or 3. To provide a comprehensive report, this study offered a patchless automated grading system for magnification-dependent and magnification-independent classifications. With an overall accuracy (means ± SD) of 90.17% ± 3.08% to 97.67% ± 1.09% and an F score of 0.9013 to 0.9760 for magnification-dependent classification, the classifiers in this work achieved outstanding performance. The proposed approach could be used for breast cancer grading systems in clinical settings.
Topics: Humans; Female; Neural Networks, Computer; Breast Neoplasms; Breast; Diagnosis, Computer-Assisted; Biopsy
PubMed: 37673327
DOI: 10.1016/j.ajpath.2023.07.007 -
Medical Humanities Sep 2023As a biomedical entity that has been the subject of a plethora of artistic and cultural projects, HeLa, the first immortal human cell line, calls for investigations into...
As a biomedical entity that has been the subject of a plethora of artistic and cultural projects, HeLa, the first immortal human cell line, calls for investigations into the human. Extracted and cultured from the cervical tumour of African-American woman, Henrietta Lacks, at Johns Hopkins Hospital in 1950s' Baltimore, HeLa's robust capacity to grow has ensured its role in numerous medical advances. The first part of this essay synthesises scientific, sociocultural, familial and philosophical perspectives on HeLa, while the second half applies these perspectives to a reading of a theatrical production, (2013), written and performed internationally by black British artist Adura Onashile. The discussion considers ways in which prevailing cultural narratives that situate Lacks as a victim, dispossessed of bodily agency in life and after death, might delimit productive possibilities for thinking about Lacks as a contributor to biotechnological progress, and about HeLa as a living remain. Lacks' labour in the creation of HeLa may have been unwitting, but her role in biotechnological progress is profound in that it is constitutive. Onashile's solo performance-its deft choreography moving across the subjectivities of patient, physician and family-presents the political fact of black female corporeality as part of its exploration of scientific innovation. The theatrical registers of Onashile's open up and nuance imaginings of Lacks/HeLa, moving beyond monolithic conceptions of medical research by creatively investigating Lacks' scientific contribution in the midst (and in the wake) of medical exploitation.
Topics: Female; Humans; HeLa Cells; Uterine Cervical Neoplasms; Art; Biomedical Research
PubMed: 36977571
DOI: 10.1136/medhum-2022-012524 -
Acta Neuropathologica Communications Aug 2023Cerebrovascular pathologies other than frank infarctions are commonly seen in aged brains. Here, we focus on multi-lumen vascular profiles (MVPs), which are...
Cerebrovascular pathologies other than frank infarctions are commonly seen in aged brains. Here, we focus on multi-lumen vascular profiles (MVPs), which are characterized by multiple vessel lumens enclosed in a single vascular channel. Little information exists on the prevalence, risk factors, and co-pathologies of MVPs. Therefore, we used samples and data from the University of Kentucky Alzheimer's Disease Research Center (n = 91), the University of Kentucky Pathology Department (n = 31), and the University of Pittsburgh Pathology Department (n = 4) to study MVPs. Age at death was correlated with MVP density in the frontal neocortex, Brodmann Area 9 (r = 0.51; p < 0.0001). Exploratory analyses were performed to evaluate the association between conventional vascular risk factors (e.g., hypertension, diabetes), cardiovascular diseases (e.g., heart attack, arrhythmia), and cerebrovascular disease (e.g., stroke); the only nominal association with MVP density was a self-reported history of brain trauma (Prevalence Ratio = 2.1; 95 CI 1.1-3.9, before correcting for multiple comparisons). No specific associations were detected between neuropathological (e.g., brain arteriolosclerosis) or genetic (e.g., APOE) variables and MVP density. Using a tissue clearing method called SeeDB, we provide 3-dimensional images of MVPs in brain tissue. We conclude that MVPs are an age-related brain pathology and more work is required to identify their clinical-pathological correlation and associated risk factors.
Topics: Humans; Aged; Brain; Neuropathology; Stroke; Brain Injuries, Traumatic; Aging
PubMed: 37641147
DOI: 10.1186/s40478-023-01638-2 -
Cancer Science Aug 2023Most multigene mutation tests require tissue specimens. However, cytological specimens are easily obtained in the clinical practice and provide high-quality DNA and RNA....
Most multigene mutation tests require tissue specimens. However, cytological specimens are easily obtained in the clinical practice and provide high-quality DNA and RNA. We aimed to establish a test that utilizes cytological specimens and performed a multi-institutional study to investigate the performance of MINtS, a test based on next-generation sequencing. A standard procedure for specimen isolation was defined. The specimens were considered suitable for the test if >100 ng DNA and >50 ng RNA could be extracted from them. In total, 500 specimens from 19 institutions were investigated. MINtS detected druggable mutations in 63% (136 of 222) of adenocarcinomas. Discordant results between MINtS and the companion diagnostics were observed in 14 of 310 specimens for the EGFR gene, and 6 of 339 specimens for the ALK fusion genes. Confirmation by other companion diagnostics for the EGFR mutations or the clinical response to an ALK inhibitor all supported the results obtained by MINtS. MINtS along with the isolation procedure presented in the current study will be a platform to establish multigene mutation tests that utilize cytological specimens. UMIN000040415.
Topics: Humans; Cytology; Lung Neoplasms; Mutation; Receptor Protein-Tyrosine Kinases; RNA
PubMed: 37139543
DOI: 10.1111/cas.15831 -
La Radiologia Medica Feb 2024We aimed at assessing the predictive ability of ultrasound-based radiomics combined with clinical characteristics for axillary lymph node (ALN) status in early-stage...
PURPOSE
We aimed at assessing the predictive ability of ultrasound-based radiomics combined with clinical characteristics for axillary lymph node (ALN) status in early-stage breast cancer patients and to compare performance in different peritumoral regions.
MATERIALS AND METHODS
A total of 755 patients (527 in the primary cohort and 228 in the external validation cohort) were enrolled in this study. Ultrasound images for all patients were acquired and radiomics analysis performed for intratumoral and different peritumoral regions. The MRMR and LASSO regression analyses were performed on extracted features from the primary cohort to construct a radiomics signature formula combined with clinical characteristics. Pearson's coefficient and the variance inflation factor (VIF) were performed to check the correlation and the multicollinearity among the final predictors. The best performing model was selected to develop a nomogram, which was established by performing binary logistic regression and acquiring cut-off values based on the corresponding nomogram scores of the masses.
RESULTS
Among all the radiomics models, the "Mass + Margin3mm" model exhibited the best performance. The areas under the curves (AUC) of the nomogram in the primary and external validation cohorts were 0.906 (95% confidence intervals [CI] 0.882-0.930) and 0.922 (95% CI 0.894-0.960), respectively. They both showed good calibrations. The nomogram exhibited a good ability to discriminate between positive and negative lymph nodes (AUC: 0.853 (95% CI 0.816-0.889) in primary cohort, 0.870 (95% CI 0.818-0.922) in validation cohort), and between low-volume and high-volume lymph nodes (AUC: 0.832 (95% CI 0.781-0.884) in primary cohort, 0.911 (95% CI 0.858-0.964) in validation cohort).
CONCLUSIONS
The established nomogram is a prospective clinical prediction tool for non-invasive assessment of ALN status. It has the ability to enhance the accuracy of early-stage breast cancer treatment.
Topics: Humans; Female; Breast Neoplasms; Nomograms; Lymphatic Metastasis; Prospective Studies; Radiomics; Retrospective Studies; Lymph Nodes
PubMed: 38280058
DOI: 10.1007/s11547-024-01768-0 -
Medical Image Analysis Feb 2024Artificial intelligence (AI) has a multitude of applications in cancer research and oncology. However, the training of AI systems is impeded by the limited availability...
Artificial intelligence (AI) has a multitude of applications in cancer research and oncology. However, the training of AI systems is impeded by the limited availability of large datasets due to data protection requirements and other regulatory obstacles. Federated and swarm learning represent possible solutions to this problem by collaboratively training AI models while avoiding data transfer. However, in these decentralized methods, weight updates are still transferred to the aggregation server for merging the models. This leaves the possibility for a breach of data privacy, for example by model inversion or membership inference attacks by untrusted servers. Somewhat-homomorphically-encrypted federated learning (SHEFL) is a solution to this problem because only encrypted weights are transferred, and model updates are performed in the encrypted space. Here, we demonstrate the first successful implementation of SHEFL in a range of clinically relevant tasks in cancer image analysis on multicentric datasets in radiology and histopathology. We show that SHEFL enables the training of AI models which outperform locally trained models and perform on par with models which are centrally trained. In the future, SHEFL can enable multiple institutions to co-train AI models without forsaking data governance and without ever transmitting any decryptable data to untrusted servers.
Topics: Humans; Artificial Intelligence; Learning; Neoplasms; Image Processing, Computer-Assisted; Radiology
PubMed: 38104402
DOI: 10.1016/j.media.2023.103059 -
The Journal of Molecular Diagnostics :... Dec 2023Diagnosing, selecting therapy for, and monitoring cancer in patients using a minimally invasive blood test represents a significant advance in precision medicine. Wide... (Review)
Review
Diagnosing, selecting therapy for, and monitoring cancer in patients using a minimally invasive blood test represents a significant advance in precision medicine. Wide variability exists in how circulating tumor DNA (ctDNA) assays are developed, validated, and reported in the literature, which hinders clinical adoption and may negatively impact patient care. Standardization is needed for factors affecting ctDNA assay performance and reporting, including pre-analytical variables, analytical considerations, and elements of laboratory assay reporting. The Association for Molecular Pathology Clinical Practice Committee's Liquid Biopsy Working Group (LBxWG), including organizational representation from the American Society of Clinical Oncology and the College of American Pathologists, has undertaken a full-text data extraction of 1228 ctDNA publications that describe assays performed in patients with lymphoma and solid tumor malignancies. With an emphasis on clinical assay validation, the LBxWG has developed a set of 13 best practice consensus recommendations for validating, reporting, and publishing clinical ctDNA assays. Recommendations include reporting key pre-analytical considerations and assay performance metrics; this analysis demonstrates these elements are inconsistently included in publications. The LBxWG recommendations are intended to assist clinical laboratories with validating and reporting ctDNA assays and to ensure high-quality data are included in publications. It is expected that these recommendations will need to be updated as the body of literature continues to mature.
Topics: Humans; United States; Cell-Free Nucleic Acids; Pathology, Molecular; Consensus; Pathologists; Neoplasms
PubMed: 37806433
DOI: 10.1016/j.jmoldx.2023.09.004 -
Journal of Cellular and Molecular... Oct 2023Endometrial cancer (EC) is a common gynaecological malignant tumour with unclear pathogenesis. Small nucleolar RNA (snoRNA) is involved in many biological processes,...
Endometrial cancer (EC) is a common gynaecological malignant tumour with unclear pathogenesis. Small nucleolar RNA (snoRNA) is involved in many biological processes, including those of cancers. Using the Cancer Genome Atlas (TCGA) database, the expression pattern of a snoRNA, SNORA73B, was analysed. The biological functions of SNORA73B were assessed by in vitro proliferation, apoptosis, migration, and invasion assays and in vivo by the xenograft model. RNA sequencing (RNA-seq) and RNA immunoprecipitation assays were performed to determine the relationship between SNORA73B and its target genes. High-performance liquid chromatography (HPLC) was performed to detect the pseudouridine content of the mindbomb E3 ubiquitin protein ligase 1 gene (MIB1). The stability of MIB1 mRNA was evaluated using a transcription inhibitor, actinomycin D. By performing co-immunoprecipitation assays, the change in the ubiquitin levels of the Jagged canonical Notch ligand 1 (Jag 1), caused by SNORA73B and MIB1, was identified. RNA-seq and qRT-PCR were performed to detect the alternative splicing of the regulator of the chromosome condensation 1 gene (RCC1). The TCGA database analysis showed that SNORA73B was highly expressed in EC. SNORA73B promoted cell proliferation, migration, and invasion and inhibited apoptosis. SNORA73B modified the pseudouridine content in MIB1 and increased the stability of MIB1 mRNA and protein; thus, it affected Jag 1 ubiquitination and further activated the Notch pathway. SNORA73B also affected the alternative splicing of RCC1, increasing the number of transcripts, RCC1-T2 and RCC1-T3, which promoted cell proliferation, migration, and invasion. SNORA73B can be a potential target for EC.
Topics: Female; Humans; Ubiquitin-Protein Ligases; Alternative Splicing; Pseudouridine; RNA, Small Nucleolar; Endometrial Neoplasms; RNA, Messenger; Cell Proliferation; Cell Line, Tumor; Gene Expression Regulation, Neoplastic; Nuclear Proteins; Cell Cycle Proteins; Guanine Nucleotide Exchange Factors
PubMed: 37488742
DOI: 10.1111/jcmm.17850