Unveiling the Power of Model-Agnostic Multiscale Analysis for Enhancing Artificial Intelligence Models in Breast Cancer Histopathology Images.

Developing AI models for digital pathology has traditionally relied on single-scale analysis of histopathology slides. However, a whole slide image is a rich digital representation of the tissue, captured at various magnification levels. Limiting our analysis to a single scale overlooks critical information, spanning from intricate high-resolution cellular details to broad low-resolution tissue structures. In this study, we propose a model-agnostic multiresolution feature aggregation framework tailored for the analysis of histopathology slides in the context of breast cancer, on a multicohort dataset of 2038 patient samples. We have adapted 9 state-of-the-art multiple instance learning models on our multi-scale methodology and evaluated their performance on grade prediction, TP53 mutation status prediction and survival prediction. The results prove the dominance of the multiresolution methodology, and specifically, concatenating or linearly transforming via a learnable layer the feature vectors of image patches from a high (20x) and low (10x) magnification factors achieve improved performance for all prediction tasks across domain-specific and imagenet-based features. On the contrary, the performance of uniresolution baseline models was not consistent across domain-specific and imagenet-based features. Moreover, we shed light on the inherent inconsistencies observed in models trained on whole-tissue-sections when validated against biopsy-based datasets. Despite these challenges, our findings underscore the superiority of multiresolution analysis over uniresolution methods. Finally, cross-scale analysis also benefits the explainability aspects of attention-based architectures, since one can extract attention maps at the tissue- and cell-levels, improving the interpretation of the model's decision. The code and results of this study can be found at github.com/tsikup/multiresolution_histopathology.

PIK3CA mutations in endocrine-resistant breast cancer.

Around 75% of breast cancer (BC) patients have tumors expressing the predictive biomarker estrogen receptor α (ER) and are offered endocrine therapy. One-third eventually develop endocrine resistance, a majority with retained ER expression. Mutations in the phosphatidylinositol bisphosphate 3-kinase (PI3K) catalytic subunit encoded by PIK3CA is a proposed resistance mechanism and a pharmacological target in the clinical setting. Here we explore the frequency of PIK3CA mutations in endocrine-resistant BC before and during treatment and correlate to clinical features. Patients with ER-positive (ER +), human epidermal growth factor receptor 2 (HER2)-negative primary BC with an ER + relapse within 5 years of ongoing endocrine therapy were retrospectively assessed. Tissue was collected from primary tumors (n = 58), relapse tumors (n = 54), and tumor-free lymph nodes (germline controls, n = 62). Extracted DNA was analyzed through panel sequencing. Somatic mutations were observed in 50% (31/62) of the patients, of which 29% occurred outside hotspot regions. The presence of PIK3CA mutations was significantly associated with nodal involvement and mutations were more frequent in relapse than primary tumors. Our study shows the different PIK3CA mutations in endocrine-resistant BC and their fluctuations during therapy. These results may aid investigations of response prediction, facilitating research deciphering the mechanisms of endocrine resistance.

Longitudinal molecular profiling elucidates immunometabolism dynamics in breast cancer.

Although metabolic reprogramming within tumor cells and tumor microenvironment (TME) is well described in breast cancer, little is known about how the interplay of immune state and cancer metabolism evolves during treatment. Here, we characterize the immunometabolic profiles of tumor tissue samples longitudinally collected from individuals with breast cancer before, during and after neoadjuvant chemotherapy (NAC) using proteomics, genomics and histopathology. We show that the pre-, on-treatment and dynamic changes of the immune state, tumor metabolic proteins and tumor cell gene expression profiling-based metabolic phenotype are associated with treatment response. Single-cell/nucleus RNA sequencing revealed distinct tumor and immune cell states in metabolism between cold and hot tumors. Potential drivers of NAC based on above analyses were validated in vitro. In summary, the study shows that the interaction of tumor-intrinsic metabolic states and TME is associated with treatment outcome, supporting the concept of targeting tumor metabolism for immunoregulation.

Prognostic impact of HER2 biomarker levels in trastuzumab-treated early HER2-positive breast cancer.

BACKGROUND: Overexpression of human epidermal growth factor receptor 2 (HER2) caused by HER2 gene amplification is a driver in breast cancer tumorigenesis. We aimed to investigate the prognostic significance of manual scoring and digital image analysis (DIA) algorithm assessment of HER2 copy numbers and HER2/CEP17 ratios, along with ERBB2 mRNA levels among early-stage HER2-positive breast cancer patients treated with trastuzumab. METHODS: This retrospective study comprised 371 early HER2-positive breast cancer patients treated with adjuvant trastuzumab, with HER2 re-testing performed on whole tumor sections. Digitized tumor tissue slides were manually scored and assessed with uPath HER2 Dual ISH image analysis, breast algorithm. Targeted ERBB2 mRNA levels were assessed by the Xpert® Breast Cancer STRAT4 Assay. HER2 copy number and HER2/CEP17 ratio from in situ hybridization assessment, along with ERBB2 mRNA levels, were explored in relation to recurrence-free survival (RFS). RESULTS: The analysis showed that patients with tumors with the highest and lowest manually counted HER2 copy number levels had worse RFS than those with intermediate levels (HR = 2.7, CI 1.4-5.3, p = 0.003 and HR = 2.1, CI 1.1-3.9, p = 0.03, respectively). A similar trend was observed for HER2/CEP17 ratio, and the DIA algorithm confirmed the results. Moreover, patients with tumors with the highest and the lowest values of ERBB2 mRNA had a significantly worse prognosis (HR = 2.7, CI 1.4-5.1, p = 0.003 and HR = 2.8, CI 1.4-5.5, p = 0.004, respectively) compared to those with intermediate levels. CONCLUSIONS: Our findings suggest that the association between any of the three HER2 biomarkers and RFS was nonlinear. Patients with tumors with the highest levels of HER2 gene amplification or ERBB2 mRNA were associated with a worse prognosis than those with intermediate levels, which is of importance to investigate in future clinical trials studying HER2-targeted therapy.

Cross-Cultural Adaptation, Reliability, and Validity of the Greek Version of the Fremantle Shoulder Awareness Questionnaire (FreSHAQ-GR) in Patients with Chronic Shoulder Pain.

Shoulder pain is one of the most common musculoskeletal pathologies. The association of shoulder pain with impaired proprioception and bodily self-awareness called for the cross-cultural adaptation of the Fremantle Shoulder Awareness Questionnaire (FreSHAQ) in Greek. The FreSHAQ is a relatively new self-assessment 9-item scale for impaired somatic awareness in patients with shoulder pain. The study included 100 participants (54 women) between 20 and 80 years old, with chronic shoulder pain (>3 months duration), able to comprehend and speak Greek, and recruited from three private physical therapy practices. The cross-cultural adaptation process was followed to develop the Greek version of the questionnaire (FreSHAQ-GR). Further testing for construct validity via exploratory factor analysis and correlational analysis with the Shoulder Pain and Disability Index (SPADI), the Quick Disabilities of Arm, Shoulder, and Hand (QuickDASH), the pain catastrophizing scale (PCS), a pain intensity visual analog scale (PI-VAS) in the last week, and the demographic characteristics followed. Test-retest reliability of the FreSHAQ-GR in a sub-sample of patients (n = 25) was tested upon repeated completion of the scale after a week and examined with the intraclass correlation coefficient (ICC2,1), the standard error of the measurement (SEM), and the minimum detectable change (MDC) indices. The internal consistency of the scale was also assessed. The factor analysis revealed that the scale comprised one factor relevant to shoulder awareness but with fewer items (first 6 items) than the original FreSHAQ. Additionally, the FreSHAQ-GR (both versions) demonstrated statistically significant correlations (Spearman's r) with participants' age (r = -0.31/-0.29, p < 0.002), the PI-VAS (r = 0.54/0.52, p < 0.001), the SPADI total score and both of its subscales (r = 0.39-0.45/0.34-0.39, p < 0.001), the QuickDASH (r = 0.37/0.34, p < 0.001), and the PCS (r = 0.50/0.40, p < 0.001). The questionnaire's internal consistency (Cronbach's α/McDonald's ω) was 0.833/0.827 for the 9-item and 0.871/0.867 for the 6-item FreSHAQ-GR. Test-retest reliability was excellent for both versions of the FreSHAQ-GR (ICC2,1(95% CI) = 0.97/0.98 (0.91-0.99/0.94-0.99), and with a low error margin (SEM = 1.31/0.94 and MDC95% = 3.63/2.61). Therefore, the FreSHAQ-GR (6-item version) possesses very good measurement properties and can be administered to Greek-speaking patients with chronic shoulder pain.

Breast cancer patient-derived whole-tumor cell culture model for efficient drug profiling and treatment response prediction.

Breast cancer (BC) is a complex disease comprising multiple distinct subtypes with different genetic features and pathological characteristics. Although a large number of antineoplastic compounds have been approved for clinical use, patient-to-patient variability in drug response is frequently observed, highlighting the need for efficient treatment prediction for individualized therapy. Several patient-derived models have been established lately for the prediction of drug response. However, each of these models has its limitations that impede their clinical application. Here, we report that the whole-tumor cell culture (WTC) ex vivo model could be stably established from all breast tumors with a high success rate (98 out of 116), and it could reassemble the parental tumors with the endogenous microenvironment. We observed strong clinical associations and predictive values from the investigation of a broad range of BC therapies with WTCs derived from a patient cohort. The accuracy was further supported by the correlation between WTC-based test results and patients' clinical responses in a separate validation study, where the neoadjuvant treatment regimens of 15 BC patients were mimicked. Collectively, the WTC model allows us to accomplish personalized drug testing within 10 d, even for small-sized tumors, highlighting its potential for individualized BC therapy. Furthermore, coupled with genomic and transcriptomic analyses, WTC-based testing can also help to stratify specific patient groups for assignment into appropriate clinical trials, as well as validate potential biomarkers during drug development.

PD-1 protein and gene expression as prognostic factors in early breast cancer.

BACKGROUND: There is a paucity of data on the prognostic value of programmed cell death protein 1 (PD-1) protein and gene expression in early breast cancer (BC) and the present study's aim was to comprehensively investigate it. METHODS: The study consisted of three parts: a correlative analysis of PD-1 protein and gene expression from an original patient cohort of 564 patients with early BC; a systematic review and trial-level meta-analysis on the association between PD-1 protein expression and disease-free survival/overall survival (OS) in early BC; and a pooled gene expression analysis from publicly available transcriptomic datasets regarding PDCD1 expression. RESULTS: In the study cohort, PD-1 protein, but not gene expression, was associated with improved OS (HRadj=0.73, 95% CI 0.55 to 0.97, p=0.027 and HRadj=0.88, 95% CI 0.68 to 1.13, p=0.312, respectively). In the trial-level meta-analysis, PD-1 protein expression was not found to be statistically significantly associated with outcomes in the overall population. Finally, in the pooled gene expression analysis, higher PDCD1 expression was associated with better OS in multivariable analysis in the entire population (HRadj=0.89, 95% CI 0.80 to 0.99, p=0.025) and in basal-like tumours. CONCLUSIONS: PD-1 protein and gene expression seem to be promising prognostic factors in early BC. Standardisation of detection and assessment methods is of utmost importance.

MYC as a driver of stochastic chromatin networks: implications for the fitness of cancer cells.

The relationship between stochastic transcriptional bursts and dynamic 3D chromatin states is not well understood. Using an innovated, ultra-sensitive technique, we address here enigmatic features underlying the communications between MYC and its enhancers in relation to the transcriptional process. MYC thus interacts with its flanking enhancers in a mutually exclusive manner documenting that enhancer hubs impinging on MYC detected in large cell populations likely do not exist in single cells. Dynamic encounters with pathologically activated enhancers responsive to a range of environmental cues, involved <10% of active MYC alleles at any given time in colon cancer cells. Being the most central node of the chromatin network, MYC itself likely drives its communications with flanking enhancers, rather than vice versa. We submit that these features underlie an acquired ability of MYC to become dynamically activated in response to a diverse range of environmental cues encountered by the cell during the neoplastic process.

Author Correction: WNT signaling and AHCTF1 promote oncogenic MYC expression through super-enhancer-mediated gene gating.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Programmed death-ligand 1 gene expression is a prognostic marker in early breast cancer and provides additional prognostic value to 21-gene and 70-gene signatures in estrogen receptor-positive disease.

Gene and protein expression of programmed death-ligand 1 (PD-L1) are prognostic in early breast cancer (BC), but their prognostic information is inconsistent at least in some biological subgroups. The validated prognostic gene signatures (GS) in BC are mainly based on proliferation and estrogen receptor (ER)-related genes. Here, we aimed to explore the prognostic capacity of PD-L1 expression at the protein vs mRNA levels and to investigate the prognostic information that PD-L1 can potentially add to routinely used GS. Gene expression data were derived from two early BC cohorts (cohort 1: 562 patients; cohort 2: 1081 patients). Tissue microarrays from cohort 1 were immunohistochemically (IHC) stained for PD-L1 using the SP263 clone. GS scores (21-gene, 70-gene) were calculated, and likelihood-ratio (LR) tests and concordance indices were used to evaluate the additional prognostic information for each signature. The immune cell composition was also evaluated using the CIBERSORT in silico tool. PD-L1 gene and protein expressions were independently associated with better prognosis. In ER+/HER2- patients, PD-L1 gene expression provided significant additional prognostic information beyond that of both 21-GS [LR-Δχ2  = 15.289 and LR-Δχ2  = 8.812, P < 0.01 for distant metastasis-free interval (DMFI) in cohorts 1 and 2, respectively] and 70-GS score alone (LR-Δχ2  = 18.198 and LR-Δχ2  = 8.467, P < 0.01 for DMFI in cohorts 1 and 2, respectively). PD-L1 expression was correlated with IHC-determined CD3+ cells (r = 0.41, P < 0.001) and with CD8+ (r = 0.62, P < 0.001) and CD4+ memory activated (r = 0.66, P < 0.001) but not with memory resting (r = -0.063, P = 0.14) or regulatory (r = -0.12, P < 0.01) T cells in silico. PD-L1 gene expression represents a promising favorable prognostic marker and can provide additional prognostic value to 21- and 70-gene scores in ER+/HER2- BC.

WNT signaling and AHCTF1 promote oncogenic MYC expression through super-enhancer-mediated gene gating.

WNT signaling activates MYC expression in cancer cells. Here we report that this involves an oncogenic super-enhancer-mediated tethering of active MYC alleles to nuclear pores to increase transcript export rates. As the decay of MYC transcripts is more rapid in the nucleus than in the cytoplasm, the oncogenic super-enhancer-facilitated export of nuclear MYC transcripts expedites their escape from the nuclear degradation system in colon cancer cells. The net sum of this process, as supported by computer modeling, is greater cytoplasmic MYC messenger RNA levels in colon cancer cells than in wild type cells. The cancer-cell-specific gating of MYC is regulated by AHCTF1 (also known as ELYS), which connects nucleoporins to the oncogenic super-enhancer via ß-catenin. We conclude that WNT signaling collaborates with chromatin architecture to post-transcriptionally dysregulate the expression of a canonical cancer driver.

STAT3 Activity Promotes Programmed-Death Ligand 1 Expression and Suppresses Immune Responses in Breast Cancer.

Signal transducer and activator of transcription 3 (STAT3) is an oncogene and multifaceted transcription factor involved in multiple cellular functions. Its role in modifying anti-tumor immunity has been recently recognized. In this study, the biologic effects of STAT3 on immune checkpoint expression and anti-tumor responses were investigated in breast cancer (BC). A transcriptional signature of phosphorylated STAT3 was positively correlated with PD-L1 expression in two independent cohorts of early BC. Pharmacologic inhibition and gene silencing of STAT3 led to decreased Programmed Death Ligand 1 (PD-L1) expression levels in vitro, and resulted as well in reduction of tumor growth and decreased metastatic dissemination in a mammary carcinoma mouse model. The hampering of tumor progression was correlated to an anti-tumoral macrophage phenotype and accumulation of natural-killer cells, but also in reduced accrual of cytotoxic lymphocytes. In human BC, pro-tumoral macrophages correlated to PD-L1 expression, proliferation status and higher grade of malignancy, indicating a subset of patients with immunosuppressive properties. In conclusion, this study provides evidence for STAT3-mediated regulation of PD-L1 and modulation of immune microenvironment in BC.

The long-term prognostic and predictive capacity of cyclin D1 gene amplification in 2305 breast tumours.

BACKGROUND: Use of cyclin D1 (CCND1) gene amplification as a breast cancer biomarker has been hampered by conflicting assessments of the relationship between cyclin D1 protein levels and patient survival. Here, we aimed to clarify its prognostic and treatment predictive potential through comprehensive long-term survival analyses. METHODS: CCND1 amplification was assessed using SNP arrays from two cohorts of 1965 and 340 patients with matching gene expression array and clinical follow-up data of over 15 years. Kaplan-Meier and multivariable Cox regression analyses were used to determine survival differences between CCND1 amplified vs. non-amplified tumours in clinically relevant patient sets, within PAM50 subtypes and within treatment-specific subgroups. Boxplots and differential gene expression analyses were performed to assess differences between amplified vs. non-amplified tumours within PAM50 subtypes. RESULTS: When combining both cohorts, worse survival was found for patients with CCND1-amplified tumours in luminal A (HR = 1.68; 95% CI, 1.15-2.46), luminal B (1.37; 1.01-1.86) and ER+/LN-/HER2- (1.66; 1.14-2.41) subgroups. In gene expression analysis, CCND1-amplified luminal A tumours showed increased proliferation (P < 0.001) and decreased progesterone (P = 0.002) levels along with a large overlap in differentially expressed genes when comparing luminal A and B-amplified vs. non-amplified tumours. CONCLUSIONS: Our results indicate that CCND1 amplification is associated with worse 15-year survival in ER+/LN-/HER2-, luminal A and luminal B patients. Moreover, luminal A CCND1-amplified tumours display gene expression changes consistent with a more aggressive phenotype. These novel findings highlight the potential of CCND1 to identify patients that could benefit from long-term treatment strategies.

Correlation Between Contrast-enhanced Ultrasound Characteristics (Qualitative and Quantitative) and Pathological Prognostic Factors in Breast Cancer.

BACKGROUND/AIM: Little is known about the correlation between contrast-enhanced ultrasound (CEUS) characteristics and pathological prognostic factors in breast cancer. The aim of this study was to explore the correlation between CEUS characteristics and pathological prognostic factors. PATIENTS AND METHODS: A retrospective study with 34 malignant breast lesions was conducted. CEUS characteristics included qualitative characteristics (e.g. lesion's enhancement degree and order, internal lesion homogeneity etc.) and quantitative characteristics (e.g. peak intensity, time to peak etc.). Also, pathological prognostic factors were included (e.g. tumor grade, estrogen receptor status etc.). RESULTS: Blurred lesion margins were observed more often in tumors of high histological grade (p=0.01) and in estrogen receptor-negative tumors (p=0.049). Furthermore, perilesional enhancement was associated with positive Ki-67 expression (p=0.049), while heterogeneous internal sentinel lymph node enhancement was associated with malignant infiltration of the node (p=0.002). CONCLUSION: CEUS has the potential to provide a prevision of pathological prognostic factors in malignant breast lesions, helping in the better early patient management.

PARP1- and CTCF-Mediated Interactions between Active and Repressed Chromatin at the Lamina Promote Oscillating Transcription.

Transcriptionally active and inactive chromatin domains tend to segregate into separate sub-nuclear compartments to maintain stable expression patterns. However, here we uncovered an inter-chromosomal network connecting active loci enriched in circadian genes to repressed lamina-associated domains (LADs). The interactome is regulated by PARP1 and its co-factor CTCF. They not only mediate chromatin fiber interactions but also promote the recruitment of circadian genes to the lamina. Synchronization of the circadian rhythm by serum shock induces oscillations in PARP1-CTCF interactions, which is accompanied by oscillating recruitment of circadian loci to the lamina, followed by the acquisition of repressive H3K9me2 marks and transcriptional attenuation. Furthermore, depletion of H3K9me2/3, inhibition of PARP activity by olaparib, or downregulation of PARP1 or CTCF expression counteracts both recruitment to the envelope and circadian transcription. PARP1- and CTCF-regulated contacts between circadian loci and the repressive chromatin environment at the lamina therefore mediate circadian transcriptional plasticity.

Robust carotid artery recognition in longitudinal B-mode ultrasound images.

Automatic segmentation of the arterial lumen from ultrasound images is an important task in clinical diagnosis. Carotid artery recognition, the first task in lumen segmentation, should be performed in a fully automated, fast, and reliable way to further facilitate the low-level task of arterial delineation. In this paper, a user-independent, real-time algorithm is introduced for carotid artery localization in longitudinal B-mode ultrasound images. The proposed technique acts directly on the raw image, and exploits basic statistics along with anatomical knowledge. The method's evaluation and parameter value optimization were performed on a threefold cross validation basis. In addition, the introduced algorithm was systematically compared with another algorithm for common carotid artery recognition in B-mode scans, separately for multi-frame and single-frame data. The data sets used included 2,149 images from 100 subjects taken from three different institutions and covering a wide range of possible lumen and surrounding tissue representations. Using the optimized values, the carotid artery was recognized in all the processed images in both multi-frame and single-frame data. Thus, the introduced technique will further reinforce automatic segmentation in longitudinal B-mode ultrasound images.

A composite framework for the statistical analysis of epidemiological DNA methylation data with the Infinium Human Methylation 450K BeadChip.

High-throughput DNA methylation profiling exploits microarray technologies thus providing a wealth of data, which however solicits rigorous, generic, and analytical pipelines for an efficient systems level analysis and interpretation. In this study, we utilize the Illumina's Infinium Human Methylation 450K BeadChip platform in an epidemiological cohort, targeting to associate interesting methylation patterns with breast cancer predisposition. The computational framework proposed here extends the--established in transcriptomic microarrays--logarithmic ratio of the methylated versus the unmethylated signal intensities, quoted as M-value. Moreover, intensity-based correction of the M-signal distribution is introduced in order to correct for batch effects and probe-specific errors in intensity measurements. This is accomplished through the estimation of intensity-related error measures from quality control samples included in each chip. Moreover, robust statistical measures exploiting the coefficient variation of DNA methylation measurements between control and case samples alleviate the impact of technical variation. The results presented here are juxtaposed to those derived by applying classical preprocessing and statistical selection methodologies. Overall, in comparison to traditional approaches, the superior performance of the proposed framework in terms of technical bias correction, along with its generic character, support its suitability for various microarray technologies.

Elucidating the identity of resistance mechanisms to prednisolone exposure in acute lymphoblastic leukemia cells through transcriptomic analysis: A computational approach.

BACKGROUND: It has been shown previously that glucocorticoids exert a dual mechanism of action, entailing cytotoxic, mitogenic as well as cell proliferative and anti-apoptotic responses, in a dose-dependent manner on CCRF-CEM cells at 72 h. Early gene expression response implies a dose-dependent dual mechanism of action of prednisolone too, something reflected on cell state upon 72 h of treatment. METHODS: In this work, a generic, computational microarray data analysis framework is proposed, in order to examine the hypothesis, whether CCRF-CEM cells exhibit an intrinsic or acquired mechanism of resistance and investigate the molecular imprint of this, upon prednisolone treatment. The experimental design enables the examination of both the dose (0 nM, 10 nM, 22 uM, 700 uM) effect of glucocorticoid exposure and the dynamics (early and late, namely 4 h, 72 h) of the molecular response of the cells at the transcriptomic layer. RESULTS: In this work, we demonstrated that CCRF-CEM cells may attain a mixed mechanism of response to glucocorticoids, however, with a clear preference towards an intrinsic mechanism of resistance. Specifically, at 4 h, prednisolone appeared to down-regulate apoptotic genes. Also, low and high prednisolone concentrations up-regulates genes related to metabolism and signal-transduction in both time points, thus favoring cell proliferative actions. In addition, regulation of NF-κB-related genes implies an inherent mechanism of resistance through the established link of NF-κB inflammatory role and GC-induced resistance. The analysis framework applied here highlights prednisolone-activated regulatory mechanisms through identification of early responding sets of genes. On the other hand, study of the prolonged exposure to glucocorticoids (72 h exposure) highlights the effect of homeostatic feedback mechanisms of the treated cells. CONCLUSIONS: Overall, it appears that CCRF-CEM cells in this study exhibit a diversified, combined pattern of intrinsic and acquired resistance to prednisolone, with a tendency towards inherent resistant characteristics, through activation of different molecular courses of action.

Using the Hough transform to segment ultrasound images of longitudinal and transverse sections of the carotid artery.

Automatic segmentation of the arterial lumen from ultrasound images is an important task in clinical diagnosis. In this paper, the Hough transform (HT) was used to automatically extract straight lines and circles from sequences of B-mode ultrasound images of longitudinal and transverse sections, respectively, of the carotid artery. In 10 normal subjects, the specificity and accuracy of HT-based segmentation were on average higher than 0.96 for both sections, whereas the sensitivity was higher than 0.96 in longitudinal and higher than 0.82 in transverse sections. The intima-media thickness (IMT) was also estimated from images of longitudinal sections; the corresponding validation parameters were generally higher than 0.90. To further validate the results, arterial distension waveforms (ADW) were estimated from sequences of images using the HT technique as well as motion analysis using block matching (BM). In longitudinal sections, diastolic and systolic diameters and relative diameter changes using HT and BM were not significantly different. In transverse sections, diastolic and systolic diameters were significantly lower using the HT technique; the differences were <7%. Relative diameter changes in transverse sections were not significantly different from BM-estimated ones. The HT technique was also applied to four subjects with atherosclerosis, in which sensitivity, specificity and accuracy were comparable to those of normal subjects; the low values of sensitivity in transverse sections may reflect departure from the circular model because of the presence of plaque. In conclusion, the HT technique provides a reliable way to segment ultrasound images of the carotid artery and can be used in clinical practice to estimate indices of arterial wall physiology, such as the IMT and the ADW.